nL eure. ' + +yecg dWeh) bo ba meee lpia pret eeererers rent cal tae es SO hi bie os DL et hie \ . ; Ly bre TRiiibablissstues tee Have Sectd te rate Fe Sete teesti ry oes be rts is eg aia oad 3 Has > ¥h — Pot Re sik Ree Neted LLL Ee Poe Al et =: ee gots ph gaye 5. Sehr at 4 ee ee Ho, Seared ee Wee 2 bresletcleteeetnsenns hpn es bribtetetel roots ot Weve Peed el oa iteeetiat sue : sia lg Lad re - :: Aa ae eeebehscie se Fre gee tt hechadahaeas eee Fett od eer ne 4 bebe oe rt *hSi sesh ep etee thre bres. Wee ae TEAS A it We Oe RIN AE OF EXPERIMENTAL ZOOLOGY EDITED BY WILLIAM E. CASTLE FRANK R. LILLIE Harvard University University of Chicago EDWIN G. CONKLIN JACQUES LOEB Princeton University University of California CHARLES B. DAVENPORT THOMAS H. MORGAN Carnegie Institution Columbia University HORACE JAYNE GEORGE H. PARKER The Wistar Institute Harvard University HERBERT S. JENNINGS CHARLES O. WHITMAN Johns Hopkins University : University of Chicago EDMUND B. WILSON, Columbia University and ROSS G. HARRISON, Yale University Managing Editor VOLUME VI PUBLISHED EIGHT TIMES A YEAR BY THE WISTAR INSTITUTE OF ANATOMY AND BIOLOGY 36th STREET AND WOODLAND AVENUE PHILADELPHIA, PA, i oon Py ie he ie : , CONTENTS No. 1—January, 1909 ADA SPRINGER A Study of Growth in the Salamander, Diemyctyles viridescens ....... EpmunpD B. WILSON Studies on Chromosomes. IV. The “Accessory” Chromosome in Syro- mastes and Pyrrochoris with a Comparative Review of the Types of Sexual Differences of the Chromosome Groups. With Two Plates and. “Two Figures mithe exts.i40). . chs aces. - sine eee ae N. M. STEVENS - Further Studies on the Chromosomes of the Coleoptera. With Four PLACES reece ie ods gid wip piel eivi 0/8 Ar vrape camel nisumto we etadGNt one neem An Unpaired Heterochromosome in the Aphids. With Two Plates..... Davip Day WHITNEY The Effect of a Centrifugal Force upon the Development and Sex of Parthenogenetic Eggs of Hydatina Senta. With One Plate........ 2 Observations on the Maturation Stages of the Parthenogenetic and Sexual Eggs of Hydatina Senta. With Five Figures in the Text.......... No. 2—February, 1909 Epmunp B. Wi1son Studies on Chromosomes. V. The Chromosomes of Metapodius. A Contribution to the Hypothesis of the Genetic Continuity of Chromo- somes. With One Plate and Thirteen Figures in the Text.......... MERKEL Henry JAcoss J. F. McCLenpon Panozoan studies: Wath Two Plates: vc s3es Saad. sae Ge enews Cuar_es R. StocKarD The Development of Artificially Produced Cyclopean Fish—‘‘ The Mag- 69 125 fois) 147 207 265 nesium Embryo.” With One Plate and Sixty-three Text Figures.... 285 No. 3—May, 1909 RAYMOND PEARL Studies on the Physiology of Reproduction in the Domestic Fowl. I. Regulations in the Morphogenetic Activity of the Oviduct.......... 341 O. C. GLasER anD C. M. Sparrow ‘The Phystolazy of Nematocysts...¢ 2. aes oe oe Poe bie <2 eee e 361 LoutsE Hoyt GreGory Observations on the Life History of Tillina Magna. With Two Plates, hree Fieures and! Six/Diaprams inthe: Dext’. 2 2). ose. 5-0 eee 383 CHarLEs R. STOCKARD Studies of Tissue Growth. II. Functional Activity, Form of Regula- tion, Level of the Cut, and Degree of Injury as Factors in Determin- ing the Rate of Regeneration. ‘The Reaction of Regenerating Tissue in the Old Body. With One Plate and Eight Figures in Text....... 433 No. 4—July, 1909 C. M. CuiLp Factors of Form Regulation in Harenactis Attenuata. I. Wound Reac- tion and Restitution in General and the Regional Factors in Oral Restitution. With Twenty-Four .Figutes..... 020.05 1.c0\sae)i ae 47 R. W. HEGNER The Effects of Centrifugal Force upon the Eggs of Some Chrysomelid Beetles. With ‘Twenty-Four Figures......... 02.0 .«1ci:.sie an eeeeiee 507 WriiuiaM REIFF Contributions to Experimental Entomology. I. Junonia Ccenia Hiibner 553 II. Two Cases of Anabiosis in Actias Selene Hubner...........-.. 565 J. Frank DanieL Adaptation and Immunity of the Lower Organisms to Ethyl Alcohol... . 579° A STUDY OF GROWTH IN THE SALAMANDER, DIEMYCTYLUS «VIRIDESCENS ADA SPRINGER Part I if” Thatiigoye taal HR eca ee Oe a eras ert sorte Tart code orason hGenooaecomDaduoobods sboedd i IUD ING omalliehe Ciara Ngseqdeomeren es ToD OCOD AMON As 2H bE Ohman Anco oobonomnmaodqeodtoen 4 Tiieseie rate OhistarvatlOn cs .sccraeu istered cieiere @ aro ce ecco eee ec alee aeeteyscelete ariel elesensbevetatoys 5 INS WBS Cot ALGAE INNO oie sdomind anndhoauodaddoctimoeocoadacayseauuasanoc 6 Vie he etiectolmyunyompheraterot prow pln masters icte)-teteleuater sae iedeteteiets) aaa l-tatel-ta-teicttt = 7 VI _ Rate of decrease during starvation in normal animals and in those with the tailremoved .... 12 Willielate ot prowth\atternstanvationmen sacs iris tat orate ope os oi fees ote eter eee ae 12 VIII Effect of temperature on the rate of growth and on the rate of starvation ............. eee 13 exer Relationlohprowthitommo witim pte myertiereeiie tee hetetcie lst bert lane igeceret ete etee rete ert ae ere ae 15 Part It I To what factors are due irregularities in rate of growth........0... 000s eeeee ees eeeeees 16 ig eMWales anid fem ales...) dea: cuckoo isles: ai Saar apehebere eee mastitis ead Saree pect at Rees ae 16 Bommitial wetehtof-animals we etattir. cee iayope snot ueter seed enct os co oA epee a ee ae ee 18 MGeutiMenceoktentperature onthe mate Of erowwit leer cteleterete etal cne tei aietet ete rhete tote totes ieeereten te 21 III Results of Part I interpreted in the light of the facts obtained in Part IT ................... 25 Vem Uheoreti cal GiscussiOn’ coe sob oe et tke ccactt ne Beas Se SR ee 29 WR Sinton AE Ae re A OE At ane AISStoIe MAC EAR MTOM an anarion spe bor aac rs rEg ost: 30 PART I INTRODUCTION For the purpose of studying the rate of growth and the external factors that influence its rate, the spotted salamander, Diemy- ctylus viridescens, has several advantages. It lives well in confinement, it takes food readily from the hand, and can be fed on pieces of beef; it withstands injury and can live for a long period without food. The factors taken into consideration in this paper are four: (1) food, (2) starvation, (3) injury and (4) temperature; the re- sults being based upon experiments carried on between October, 1906, and May, 1908. Tue JourNAL or ExprrIMENTAL ZOOLOGY, VOL. VI, NO. I. 2 Ada Springer The total number of individuals used in this and the subsequent investigation was about three hundred and seventy-five. All of the salamanders were adults, at least two and one-half or three years old, for, according to Gaget “the autumn of the third or the spring of the fourth year after hatching when two and one- half or three years old, the animal enters the water and assumes an aquatic life.” All the animals used in the experiments, with the exception of six sets in Part I used to determine the effects of temperature, were taken from the same environment. The animals were kept in sets consisting of six, ten and five individuals, respectively. For convenience the various sets are designated by capital letters, A, B, C; and when one set is a duplicate of another this fact is shown by designating those sets by capitals with the addition of the numerals, A‘, A’, etc. The animals were kept in flat dishes containing from one and one-half to two inches of water, the room temperature varying from 20° Coosa tC. Pieces of raw beef of a definite size were used for food. Although an attempt was made to keep the pieces uniform, the size varied somewhat; that is, both the size of the pieces given to different animals at the same feeding and also the size of the pieces given to one animal at different feedings. A comparison of records, taken at random, of the amount of beef fed to the individuals at the same and at different feeding periods (the amount fed to the individual being determined by taking the average of the weight of beef given to the set), shows the follow- ing variations. The weight is expressed in milligrams. Three feeding periods: (1) 44—58—47—52—55; average for the period, 51 mg. (2) 61—52—61—59—63—56—_68—65—_5 360; average, 59 mg. (3) 49—49—45—55; average, 49 mg. During the first, second and possibly the third week after the beginning of the experiments the average amount of beef fed to the individuals was lower than the averages given above, because 1 The life of the vermilion-spotted newt (Diemyctylus viridescens). A Study of Growth 3 there was difficulty in feeding certain animals. After this time the average amount taken by each animal was 153 mg. in the three feeding periods during the week. The weights of the salamanders were taken once a week, there intervening a period of from forty-eight to fifty hours, on the average, between the last feeding and the time of weighing; that is, from Saturday until Tuesday. The animals were dried with a cloth so as to get rid of as much of the moisture as possible, and weighed in a beaker of known weight. ‘The weights were recorded in grams. Possible errors in the comparative weights deserve consideration, for in all of the weights taken there were opportunities for positive and negative error: (1) Error as a result of undigested matter in the alimentary tract at the time of weighing. (2) Error from not drying the animals uniformly at the same and at different weighing periods. (3) Error from lack of uniform feeding, that is, Varying amounts of food. (4) Some individuals failed to eat: this error, however, is lessened by the fact that where a certain number of individuals failed to eat, the same number in the con- trol for that set were intentionally not fed. (5) Finally, correc- tions must be made for the death of one or more animals, which may shift the average of the set. According to Minot,? increase of weight or growth depends upon two factors: (1) Upon the amount of body substance, the growing material present at a given time; (2) upon the rapidity with which that amount increases. The rate of growth may be expressed as a fraction of the weight added during a given period. Minot suggests taking the mathematical mean between the weights at the beginning, and at the end of a period, in order to furnish a basis from which to calculate the fraction, or the percentage increment. ‘The rate of growth may in this way be expressed more accurately, for if during a period there is a definite incre- ment, at the beginning of that period there is less material to increase than at the end. New material is being added daily, which in turn increases. By taking the mean, the actual amount of growing material at a definite period is more accurately ex- *Senescence and Rejuvenescence. 4 Ada Springer pressed. ‘The same principle holds in case of the rate of star- vation. During a period there is a definite loss, but at the beginning of that period there is more material to lose weight than at the end, so the actual amount of decreasing material will lie as a mean between the first and last weights of the period. The rate of increment and loss is best expressed for purposes of comparison in percentages, as absolute increments do not show relations between rates; for example, the absolute increment of a larger individual may be greater than that of a smaller one, yet the rate of growth in both may be the same, or even greater in the small one. ‘This is shown in many of the tables, where the absolute increment and the percentage increment are both given. The percentage was calculated from the average weights of the sets. II THE NORMAL RATE OF GROWTH Increase in weight in Diemyctylus viridescens is due, as Mr. Morgulis* has discovered, not to storage of fat in the various parts of the body, but to a uniform increase in the size of many of the organs, e. g., the skin, muscles, liver, ovaries, etc. ‘This type of growth seems, therefore, comparable to growth from a young to an adult stage. Conversely in starvation there is a decrease in the size of the organs. Percentage increment decreases as the maximum weight 1s approached; that is, the rate of growth becomes slower the nearer the animal approaches its upper limit of weight. In Set At (Table I), where the individuals had been fed 153 mg. (average) of the beef a week, there was a tendency for the percentage increment of weight to decrease as the animal approached the maximum or limit of weight. It might be supposed that a stage would be reached where this amount of food proved insufficient to produce an increase, and in conse- quence a state of equilibrium would be established between the body material and the amount of food, around which equilibrium there would be a fluctuation between loss and gain. If now an a Unpublished work of S. Morgulis. A Study of Growth 5 increase in the amount of food be given, this maximum should be pushed higher. Unfortunately the records in the tables do not show conclusively that this is actually the case, but there is a certain indication which may be interpreted as bearing on this question. The uniform loss in many of the sets after February 19 and March 5, lasting for about five or six weeks, cannot be attributed to the fact that the maximum had been reached, although it might have been that they were approaching the maximum. ‘This seems the case because all of the sets had not been fed as much as 153 mg. of beef for eighteen weeks, as had been Set A'(Table I), so that they cannot be supposed to have reached their limit of size at the same time. [he general loss was probably due to some general factor which affected many of the sets alike. It may have been due to the breeding season. After five weeks of loss, Set At was fed three times, 153 mg. a week (459 mg.), and after a week’s time regained almost the entire loss. The weekly percentage increments do not always show a decrease as a limit is approached, because of the varied fluctua- tions from week to week. By taking definite periods, however, such as eighteen weeks in the case of Set A! (Table I), the per- centage increment or rate during the first nine weeks, 23.3, will be found greater than during the second half, which was 21.5. The absolute increment, however, during the first half was less (545 mg.) than that during the last half of the period, which was 629 mg. In Set C? (Table III), consisting also of six individuals, which had been fed 102 mg. of beef a week during a period of eighteen weeks, the rate during the first half was 14.5 per cent., and that during the second half was g.1 per cent. In other sets the differ- ence has also proved constant, as in Set Dt (Yable IV), and Set Be (able VI). Dt tHE RAPES Oh SSlAR VAdION Set B' (Table II), consisting of six individuals, was starved for seventeen weeks, at the end of which time one individual died, indicating that the minimum limit of weight for the set was nearly reached. 6 Ada Springer The weights show that where an individual approaches the minimum weight there is a decrease in percentage of loss. In Set B* during the first half of the period, eight weeks, the rate of loss was 41 per cent, while during the last half, or nine weeks, the rate of loss was 35.2 per cent. IV EFFECTS OF VARYING AMOUNTS OF FOOD Percentage increment, or rate of growth, is directly proportionate to the amount of substance taken into the body. This is shown to be true by a comparison of the following sets. Sets Ct and C? (Table IID), consisting of six individuals respectively, were fed on the average 102 mg. of beef a week; Sets A! and A? (Table I), consisting of six individuals also, were fed 153 mg. of beef a week. After a period of nine weeks the percentage increments were as follows: CGI CHER RES Sanne UE CHTe DOBDOUo Or condben dem uicda dan cuasa Soon odocDoobS Shon 18.4 PSOE C2 racicteise shelshstsieysteree stots elo es eiatshelore\eneheheloleyoeretepereieseia/e/oletehel Nels) Pepainiarete kere stetetonal 14.5 Average of the tw0 Sets +e ce «20 seco elise mnie eo tele 16.4 per cent GLAD S poin s nietelelover sneiase ras eiceate) istererstaiet ove eter Seated liee Mehetefave cl hegetielele oleh ste) eka aasks agers 23)-3 SG) aaS Hee ene DOD ose COpOOmIaeO dn TE soda ne Ono potas CB Aaarmads Ome 29.8 Averapeiof thei two) Sets) cjeyate slater ete cibere icicle level] iolers/alnlelein)=ts 26.5 per cent After the ninth week, December 18, the individuals of Set Ct were fed 153 mg. of beef a week, while those in Set C? were, as before, fed 102 mg. a week. After another period of nine weeks the relations were as follows: Increment per cent Clio OLR enon connec eaaaad CnU SPR rae = soacape ord ToguoduoMedooouiddd. dota 34-9 See (CHetcgocevenont odog une sdonso osc doDoosaEmdonoD GOIOOdUGoU soos oHdOON Her 9.1 This would seem to show that 102 mg. of beef per week was not the maximum amount of food which could be assimilated by the digestive tract. After a period of eighteen weeks Set A‘ had gained 44.2 per cent of the original weight, while C? had gained but 23.5 per cent. Comparing Set At (Table I), in which the animals were fed 153 mg. of beef a week, and Set Bt (Table II), in which the A Study of Growth 7 animals were starved, it was found that after a period of eighteen weeks the former had gained 44.2 per cent of the initial weight, while the latter had lost 73.6 per cent of the initial weight. VY EFFECT OF INJURY ON THE RATE OF GROWTH To determine the effects of injury, the tails of the salamanders were cut off, either once at the base, or six or even nine successive times, a very small piece each time. On October 23 the tails of six salamanders were cut at the base, the animals being weighed before and after the operation, enough time being left between the weighings so that the cut surface had ceased to bleed. ‘The animals were then weighed once a week for five weeks. Control normal animals were also weighed, and kept under the same conditions as the injured ones. The feeding was, as in the experiments above, about 153 mg. of beef a week. At the end of a period of five weeks the results were as follows: ‘The injured set D' (Table IV) had gained (average) 19.5 per cent of the original weight after the tails were cut, while the percentage gained in the two normal control sets A’ and A? (Table I) was 14.3 and 20.3, respectively, the average between the two control sets being 17.3 per cent. After the first week the individuals of the injured set had almost made up the weight lost by cutting their tails. Another set, started November 6, also showed this to be the case; the first week after the injury the percentage increase was 13.6, the entire weight of the tails being almost made up. The experiment described above was repeated November 13, the number of animals used being somewhat larger. Six sets consisting of ten individuals each were used, the animals in three of the sets being injured by cutting off the tails at the base. Each set was controlled by a set of normal animals with tails intact. The injured and the normal control sets in every case were fed the same amount of beef, 153 mg. (average), and were kept under identical conditions. At the end of five weeks the results in percentage increments were as follows: 8 Ada Springer The injured sets Normal control sets (i) Seth (MableVa)iae sat tapactien nr 147 1 = SetGl (Table) 2.9.-0e- den eee 11.4 (@)ESeth2 (able) se sawiseien os serve stoners 10.5 SetG? (Mable) seasus see secre Oss Sh (GQ) eScpks (able) teenies reer 1472.) «SevG* (WableV) 5. ne esc eee eee 6a The average for the injured sets is 13.1 per cent increment, and for the two normal control sets, in which there was a gain, 8.9 per cent increment. After a period of fourteen weeks Set E* had gained 47.7 per cent, while the normal set had gained 44.6 per cent. Set E* (Table V), in which the new tails had been regenerating for five weeks, was divided, December 18, into two sets of five individuals each. In one set, E* (Table VII), the animals were again injured by cutting off the regenerating stump, together with a small piece of the old material at the base of the tail; in the other set, E'® (Table VII), the regenerating stumps were left intact. This latter set, together with the normal intact Set G! (Table V), were used as controls. ‘The results after a period of four weeks were as follows. Increment per cent See Hi injuredianimals. \4 saice ees (ae tae sie = eisteles meat ene ca (average) Dito Set EIb, animals with regenerating stumps........--- +. 222s sees eee cece cece eee 5-4 SeeGimormalintact amirvals ster sayebetats cietara)soare/ ote otek etete elite aed atitaatYoto Vk Peet tot 11.4 The same experiment was duplicated by dividing Set E* (Table V) in the same way as Set E', the two being equivalent sets. ‘The results were correspondingly the same (Table VIII): Increment average per cent Sepb?*sinjlred animalsac.- om. vase veer oes nerine rs ate cele psy iat eee aN a Set Eb, animals with regenerating stumps........----2+- eee e eee eee eee ee eee 18.2 SetiG) normalantactamimal sey wet eiersec eet eeeiaratetiaie ley siebettrccee ite een ee Tne 4 The percentage increment, or the rate of growth in the injured set in both cases, was greater than in the normal control or in those with regenerating stumps. Three other experiments, also started on December 18, resulted in similar comparative percentage increments. ‘Iwo sets which A Study of Growth fe) had been starved for five weeks were taken. In the one the tails were cut at the base, H*® (Table IX); in the other, Set H? (Table IX), the tails were left intact. After feeding both sets 153 mg. (average) of beef a week, for four weeks, the percentage increment in the injured set was 31.6, while that of the normal control only 20:2. Set F! (Table XIV), in which the tails had been cut off at the base, and which had been starved for five weeks, was divided into two sets of five individuals each. In Set F* (Table X) the animals were again injured by cutting off the regenerating stumps, together with a small piece of the old material of the tail. In the other set F'” (Table X), the regenerating tails were left intact, and this set together with Set H? (Table IX), the normal animals, served as controls. ‘The results show the following percentage increments in the three cases, the animals being fed 153 mg. of beef (average) for four weeks: Increment per cent Seb nyuredianinaal ss 7,-0.5, accisitieccravs sage wine aes kore oes aiaie oer sicbote saa eee 21.8 Set F1, with Ine{YSA IMAI aviaN SINK onogdoanbedoo oc odovemDIOSEZECSooDDTUS 23.4 SetH=nonmalintactanimalsic nani ccc csi srereserseretel ele ure = tue ieieves sehesasne sections 20.3 This experiment was carried further by dividing Set F* (Table XIV) as F' had been. The animals in Set F* (Table XI) were injured by cutting the regenerating stumps as before. ‘The results, after feeding these sets as before for four weeks, were similar to to those above, viz: Increment per cent Sate betaiing une dtaniinals Mesrayecictsesee casei veers et Grae area ie te aca oe eae 28. J Set Fb, with IaH aval Gaevenjogmiwlsrsaneseacwe 4 dade sndosohoseAcowsmeAsna see 27.8 Segllasnornmalintact-animal’S....cmec 2 A siaaratto nineties one iaeiro ion el rer eaten 20.3 After five weeks’ starvation and subsequent feeding 153 mg. of beef a week for four weeks, the percentage increments in Sere F? and H? (Table XV) were as follows: Increment per cent Seu be withire cen eratinieys tum pSrrttty statck aie ravi ete sere ite ariel venatt tane eae on eee net ee 27 Sepbesnormaliintacthanimials:taa: cin se neret-vccst Warsreye estate eal etclcteee uke ee eres o 20.3 10 Ada Springer After a period of eleven weeks the results were as follows: Increment per cent SEE oo Sogo doa0 CONDOS Dod ES OOO Pade GeO ZOO NDDUSCODuOOOODGOGOMOONSS 55-5 Saal s CaS aan Cee a arate te GO EER CP A Ree ceo MoO do onmaacoce 48.9 In every experiment, with the exception of Set F!> (Table X), the percentage increment was greater in the injured animals. The degree of difference in some cases was very great, while in others It was but comparatively slight. In the majority of cases, as may be seen by a comparison of the tables, the percentage increment of the injured animals was especially great during the first week after the injury; in some cases the loss of the tail being almost made up. As a basis for comparison of the rate for the intact and for the injured animals, short periods of four or five weeks were taken. During this time the tails of the injured animals had begun to regenerate, and had added new material to the old. The average actual weight of the new material added during this period was from 42 to 50 mg.; and taking into consideration indi- vidual variation in this regard, in no case can the greater percentage increment in the injured animal be said to be due to the added weight of the new material; the difference in all cases being too great to be attributed to such a cause. The results of the experiments cited above are in accordance with those by Professor Morgan. As expressed by him,! “The greater percentage increment in the injured animals may be due to the influence of the regenerating tail on the growth of the rest of the body, or, if not due to accidental factors, it may be that the changes taking place at the cut surface incite the digestive tract to greater activity or the cells of the body to greater assimilation. In this way the injured animals would gain proportionately more body weight.” ‘This question is treated again in the light of additional experiments in Part II. To determine the effects of successive cuts in three sets, I', I’, I%, (Table XII), consisting of six individuals each, the tails were cut at the tips, the weight being taken before and after the operation. 4 Jour, Experimental Zodlogy for December, 1906, The Physiology of Regeneration. A Study of Growth II During five successive weeks the remainder of the tail was cut five successive times. As a control set the records of a corre- sponding number of weeks for Sets At and A? (Table I) were taken. ‘These cannot, however, be considered strictly equivalent, for the experiments were not started on the same dates. The comparisons are therefore not very satisfactory. The animals were fed as before. Four weeks after the last cut was made, a time corresponding to that in the cases where the tails were cut at the base, the percentage increments were as follows (because of the great variation in the results of the different sets averages in the two groups of sets were taken): In the three injured sets: Setiit (Mablesxui): ermeam nae niaits oa tiste ener cited reeset cee aera me 42.7 Seel2|((ia ble wXaih)o igen Sita cutest semen Gael eaten rasa tas eee creeper ees 24.5 Seone2 (GTvalb Le eXuell) | ove stare erat oenlstatsts vei av asolae tro aes tears a telSyefosas cf aperere eanon era ewe Pehl INTENRS “o 0 oda Uap dobaebe Souoos GuascboouendggcbDOr 30.2 per cent In the normal control sets: SetenU (Glial len) ye cearaie-areraveje taretetere si ateve oltovi ess Mespeasrore aie eae he eae ore aaa eR oes 23.3 Seine (Mab lens) strte ee ricsese te ae siavesuie tm ate STE Ie sre elole d Municlee’S « Mharssnpe mermtmrateve mame ater 29.8 INV CLARE daictstcrerdeve chloe tp sseerse mace acicresevesint sisters siete 26.5 per cent On December 18 a similar experiment was started. Sets G* and G’, consisting of ten individuals each, after having been well fed for five weeks, were divided into two sets of five individuals, respectively, viz: Sets G* and G® (Table XIII), and G* and G*® (Table XIII). In set G* the tails were cut nine successive times; five weeks after the last cut the percentage increment was 48.7. In Set G” the tails were cut six successive times; four weeks after the last cut the percentage increment was 52.3. In Set G* the weight was so far below the average that the results are of no value, and G? became infected. For Set G* there was no set which could be taken as a comparative control; but comparing the thirteen weeks with a corresponding period of any normal intact set the percentage increment was considerably higher. In fact, the percentage in these two cases, except those obtained after a period of starvation, were the highest percentages obtained in any of the experiments. 12 Ada Springer There are, perhaps, not sufficient data upon which to determine definitely whether or not the percentage increment is greater when small pieces of the tail are cut off successively and when it is cut only once at the base; but there are indications from the weekly records and from the end results showing that the successive cuts produce a greater increment than does the single cut at the base. This would seem to indicate that the increased rate of growth is the direct response to the cut without regard to the regenerating mass; but there are other factors yet to be considered that show the results can not be so simply interpreted. VI RATE OF DECREASE DURING STARVATION IN NORMAL ANIMALS AND IN THOSE WITH THE TAIL REMOVED Six sets, consisting of ten individuals each, were taken November 13. In three sets the tails of the animals were cut off at the base, in the other three sets they were left intact. After four weeks’ starvation a comparison shows the following percentage incre- ments: 1 The injured animals 2 The normal controls SepRu (Mable Xl Vine. ie once echoes ele 6 SetHt (PablesX0Vi) sss oot eee 6.5 Seal GR Inly dN) Peendepanebo no cadancc 4.6 Seti? (Lable Xan) ies ic... ce bs ve creep OnG Set BS Ga blewXaivy) srs. sew eee te store oye 8.4 SetiH? (able XGW))25 sce. as oie seers 0.2 The average for the three injured sets is 26.3 per cent, while for the normal it is 29.06 per cent. VII RATE OF GROWTH AFTER STARVATION Minot states that any irregularity of growth in his guinea pigs tends to be followed by an opposite compensating irregularity. “Each individual appears to be striving to reach a particular size. If growth ceases because of any factor which deprives the individual of the normal conditions, as sickness, when the normal conditions are again brought about there is a tendency, by accel- eration of the rate of growth, to make up the loss.” Starvation, although not equivalent to sickness, may be com- pared with it, in that starvation is a factor which deprives the A Study of Growth 13 individual of normal conditions. In animals that had been starved for five weeks, and after that time were fed 153 mg. (average) of beef, there was a marked acceleration of growth. Four sets were starved for five weeks, after which they were fed 153 mg. (average) of beef. Four sets were taken as controls which had been well fed for five weeks, after which time the same amount, I53 mg., was given to each animal. The condition of the animals at the beginning of the experiment was identical. After four weeks the differences in percentage increments were as follows: : I After five weeks’ starvation 2 After five weeks’ of feeding Sennen (Mable XM )tycceiscea vate BOR Set iGU (able nVal) ren seco ork cee: 11.4 Sethe (@ableoXO ie aeweteoss Asse | 28 See Ei2) (lables Vill) te nesseeee sie eee Sepery ei ablemX) 184 ay. o ss scares 4 Set BP (Pable Villain court 5.3 Seeykso (habla: XM) os. ccet coos cect AATee ee aSee BPN VIN) Os aie epee Meee eet. 18.2 The percentage increment after the period of starvation in every case proved to be the greater. In many other sets this isalso shown, as Sets A! (Table I); B? (Table II); D? (Table IV). Throughout the experiments another fact comes to light that appears to be quite general, namely, that a very high percentage of gain during one week is followed the next by a low percentage, and in many cases even by a loss. This is most striking after feeding individuals that have been starved; the first percentage is very high, while those immediately following them are low. VIL EERE CT OR EVR EVA URE It is generally known that the rate of growth is accelerated by warmth and retarded by cold. ‘This has been determined for the growing embryo and for regenerating parts. With the view to determine the effect of temperature above and below the normal (the normal in this case, being considered the room temperature), upon the rate of growth and the rate of starvation in adult salamanders, six sets, consisting of six individuals each, were kept under three conditions of temperature, as follows: Two sets were kept at the normal temperature of the room, averaging 22° C. The highest temperature recorded during the 14 Ada Springer period of seven weeks between March 12 and April 16 was 24° C., the lowest 19° C. One set was fed three times a week, the other starved. Two sets were kept at a temperature lower than that of the room, viz: 11° C. on the average. The highest temperature recorded was 15° C., the lowest 6° C. One set was fed three times a week, the other starved. ‘Two sets were kept at a temperature of 28° C. on the average. The highest temperature recorded was 31°, the lowest 25.5° C. As before, one set was fed and the other starved. The quantity of beef given to each individual averaged some- what lower than 153 mg. per week, the variation being from one- half to two-thirds of that amount. This was due to the fact that the individuals under conditions of low temperature proved diff- cult to feed, and in almost all cases it was found necessary to put the food into the mouth. This quantity was necessarily taken as a basis for the other sets. A comparison of the records in Table XVI shows that the relative rate of growth cannot be determined, because the quantity of beef taken by the individuals in Set C, Table XVI (those under conditions of low temperature), was only enough to preserve the equilibrium. ‘There were fluctuations between gain and loss, but the results after seven weeks’ time show there was a gain of only 1.04 per cent of the original weight. ‘The same amount of food was inadequate in the case of A and B to preserve the equilibrium, and a steady loss was the result. The percentage of loss was greater in Set A, where the temperature was higher than in Set B; in the one case 31 per cent, and in the other 23.08 per cent. On the other hand, the relative rate of loss can be determined in starving individuals. This is found to be highest at a high temperature (Table XVI), Set D, 58.5 per cent; it becomes lower at the normal temperature, Set E 41.4 per cent; and lowest when the temperature is below normal, Set F, 18.3 per cent. Records were kept of individual salamanders which had been subjected to these conditions of temperature. Previous to the experiment the animals had been fed three times per week since A Study of Growth 15 November 13. ‘The records from these weighings show the same general results, namely, that the rate of starvation was lowest under conditions of low temperature, and that the quantity of beef necessary to preserve equilibrium in case of the individual at a low temperature was not sufficient for those at a higher temperature. IX RELATION OF GROWTH TO MOULTING There seems to be a more or less definite relation between the quantity of food and the frequency of shedding the skin. Al- though there is great individual variation in this regard, there is a certain uniformity which warrants the following conclusion, based upon the records of thirty-six individuals. The greater the quantity of beef fed to the individual the more frequent is the period of shedding. ‘This was determined by a comparison of the records of the individuals in Set At (Table 1) and Set C? (Table III); the quantity of beef fed in the one case being greater than in the other. The average record of moulting in Set A? was seven, and in Set C? four, during the period from October to April. ‘The individual records, showing the greatest frequency; give ten moultings. Four moultings were recorded in the case of starving individuals during the period. PART II With the view to determine some of the causes of variation in the results of the experiments carried on during the winter of 1906-07 (Part I), a new series was started in October, 1907. Greater precautions were taken to diminish the possibilities of error. The animals were carefully dried and weighed in a covered dish, so as to prevent evaporation during weighing. Greater care was taken to preserve the average size of the pieces of beef. The average size of the pieces given to each individual at the beginning of the experiment was maintained throughout the later feedings, and was not, as in the first series (that of 1906-07), gradually increased during the first few weeks. Inthe first series of experiments the average size of the pieces of beef 16 Ada Springer after the first week or two, was about 50 mg., or 153 mg. per week; while in the present series the amount was twice as large. The average weight of each piece was 105 mg., or 315 mg. per week. The feeding in this series was regular from the beginning, which was not the case in the first series. A record of each individual was kept, together with the niurHee of pieces of beef consumed, so that at the end of a definite period the amount of food taken by each individual was known. Of course there was a chance for error in cases where the beef was later rejected by the animal. It will be well to state that if a certain individual in a set was not disposed to eat for several periods, the corresponding control individual was not fed at these times; thus keeping the conditions in each set as even as possible. I TO WHAT FACTORS ARE DUE IRREGULARITIES IN RATE OF GROWTH It has been found that animals kept under almost identical conditions of temperature, of food given and of depth of water; varied considerably in the rate of increment. As to the causes of these variations, several questions arose: (a) Does sex account for the variation; do females gain faster than males? (b) Does the initial weight of the animal affect the percentage increment? (c) Do the periods of moulting affect the rate ? (a) Males and females. Two sets, one consisting of ten males, the other of ten females, were placed under identical conditions of temperature and food, October 22. The average amount of beef taken by each individual per week was 315 mg. In respect to feeding, these two sets proved to be the most satisfactory of any during the experiments—for, with the possible exception of once or twice, the feeding was absolutely regular. After a period of ten weeks the average percentage increment was as follows: Per cent Set Au@Rable 2avilili) males cx.t14..cpreae eee eee eee cee A ere eRe eee 29.7 Set Bi((Dable: kVail) females: cr 2 eis foe eters Raye eregeiechd ote Pare eee 42.1 A Study of Growth 17 Thus the percentage increment of the females was considerably higher than that of the males. In order to determine individual variation within a set, five males and five females from Sets A and B respectively were selected at random. ‘The individual feeding record, together with that of the weights, was kept. The following table shows the initial weight of each individual with its percentage increment during the period of ten weeks. Average amount of beef per week was 315 mg. MALES (TABLE xvil1) FEMALES (TABLES XIX) In, wt. Increment In, wt. Increment grams per cent grams per cent INDY AC) eae eaana ae oeces ols 38.9 INioG@) eatin cic sens W237 64.8 HNO (5) Sic letes chet eveserotsns o> 1.885 28.8 INOH(G)) weseen ates. 1.387 54-2 EN OEA(A)) ret crete tetenets fer taj eh 2.509 26.8 INiopH(G) bendeeaaaooeDeT 1.447 Aijatt IN OE (1D) Brersietststovevels, sicker 16 2.627 28 .o1 INOW (woe cwhiaasaeesere 1.927 35.6 UNV orse (2) es setevarersts roe o/s 2.787 20.1 INOS (Q) iaeoncrte inne sesnars D222) 22 7 From the above table it may be seen that while the percentage increment of several of the individual males is greater than that of some of the females, in general the percentage in the females is greater. The highest percentage recorded for a female for the period of ten weeks is 64.8 per cent while that of a male was Be O) pet cent. Records of individuals taken from the sets kept under high temperatures are not comparable with those given above, because of differences of temperature and feeding; but within the same set males and females may be compared. In Set D? (Table XXIII) at high temperature (30° C.), the individual records of two males and one female were taken. Each animal was fed 105 mg. of beef at each feeding period about thirty times during the ten weeks. The initial weights together with the percentage increments were as follows: MALES (TABLE XXIII) FEMALE (TABLE XXIII) In. wt. Loss In. wt. Increment grams per cent grams per cent INOW (Escape natnnee 1.037 ini INC ©) hedioe poragencpoct 1.389 2.4 18 Ada Springer Though the three individuals were fed the same amount of beef, both of the males lost in weight, while the female gained 2.4 per cent of the initial weight. In Set E? (Table XXV) at high temperature (30° C.), individual records of two males and one female were kept. Each animal was fed about forty times during the period. ‘The initial weights and percentages were as follows: MALES (TABLE XXV) FEMALE (TABLE XXV) In, wt. Increment In, wt. Increment grams per cent grams per cent IN Gs(2)) eee eieseeeee 1.539 22.8 INOS(3) cee cueciiane eee 1.172 57-4 ING) ects cera stocton 2.805 10.5 Here again the percentage in the female was considerably higher than in the males. From the tables, however, it will be noticed that the initial weights of some of the females are less than are those of the males. The question arises as to whether the difference in percentage increments might not be due to this fact rather than to sex. (b) Initial weight. By comparing the records of the five males in lable XVITI, it will be found that although there are exceptions, the animal whose initial weight is greatest shows the least per- “centage increment, as in the following: In. wt. Increment grams per cent INOS (3) Fas tars ceevscecs ms ecersichahe bee rate cals euenet eats Aaa,b ene sre ede arsia tora ole eranotereteteier Ne 1.577 38.9 INOS) As EAA A cleiciansn asthe oe lobgeiaieve'n a sieee Ao wie eae ens Costes tate raters 2:71.97, 20.1 These two weights represent the extremes and lying between them are gradations, showing that as the initial weight increases, the percentage increment decreases. By a comparison of the five females in (Table XIX), the results are in general the same, as follows: In, wt, Increment grams per cent INO C21) op. Es ALES oe eigatehapaie ate ue Si tage tleteyaeatetos ay cae em ener eelomt ernie steers Paes 1.237 64,8 NOH G0 eee Ge ci Or oon Om RE Se mo cmebRactacn tn Soo maraccuon anne 2,222 2247 The greater the initial weight the less the percentage increment A Study of Growth 19 Comparison of females and males of approximately the same initial weight, ought to show whether the percentage increment is greater in the females or in the males. The data on this point are not conclusive, as shown by a review of Tables XVIII and XIX. No. No. No. MALES (TABLE XVIII) In.wt. MALES (TABLE XXIII) In, wt. grams MALES (TABLE xxv) In, wt. grams whale) aie, 6s) vilelie jeje \e) 10 Increment per cent 38.9 28.8 26.8 28 .o1 20.1 Loss per cent 12 Increment per cent 22.8 10.5 FEMALE (TABLE XIX) In, wt. grams Nos 2) setae scene pene ae 1237, Non (4) seicreeaisie et ciereis 1.387 INOA((5)) Sterioc vercmerccere 1.447 INOW) ta Senin. ser erie cts 1.927 INO H(i) owes evetsnetateere eterets eoa2 FEMALES (TABLE xxXxIII) In, wt. grams NOG) e cmsn ne erates 1.389 FEMALES (TABLE XXv) In, wt. grams Nos ))tcnwcescnn ence 1.172 Increment per cent 64.8 54.2 35-1 35.6 2207) Increment per cent 2.4 Increment per cent 57-4 It is thus seen that the percentage increment for both sexes is very closely connected with the initial size. Whether there are also minor differences related to the sex of the individual can not be determined with certainty from the facts obtained. In order to find out more exactly the relation between food, initial weight and percentage increment the weights of each piece of beef taken by each of five males and five females (Tables XVIII and XIX), respectively, for a period of four weeks out of the ten were recorded. The following table shows the initial weights at the beginning of the period of four weeks, the amount of beef taken during the time and the percentage increments: 20 Ada Springer MALES (TABLE XVIII) FEMALES (TABLE XIX) In.wt. Amt. of beef Increment In, wt. Amt.ofbeef Increment grams grams per cent grams grams per cent Nios (@bocke 1.948 1.269 18.2 IN (Ges ase 1.68 1.254 20.5 Now (5) ier Depipes 1.246 1255 Now) Haece 1.92 1.272 2A INow(4) sacle 2.81 1.274 15.6 Wis (@))poane 1.987 1.290 19.5 INow(n)o=e- 3.04 1.291 Talo INoaiI(i)saeer 2.447 1.271 12.09 NOs) erie 1 3098 1.256 9.6 INDE EDle core Daisy W277 10.4 By comparing the five males it will be found that, while all have consumed approximately the same amount of beef, the percentage increments vary considerably. ‘Taking the extremes, Nos. 3 and 2, it appears that while both have eaten about the same amount of beef, yet the percentage increment of the first, the initial weight being 1.948 gram, was twice that of the second, whose initial weight was 3.098 grams. Between these extremes there are inconsistencies, yet in general the same relation holds. A similar comparison between the five females (Table XIX) shows the same relation. In Nos. 5 and 3, where the initial weights were 1.68 and 2.515 grams respectively, the amount of beef was approximately the same; yet the percentage increment in the first case was 20.5, and in the second 10.4. ‘These results are in accordance with those for the period of ten weeks, and show that with increase of initial weight there is a decrease in percentage Increment. In connection with these general results it should be noted that in Tables XVII and XVIII there are weekly variations for which the factors considered cannot account. Set C (Table XX) was composed of seven individuals consider- ably below the average in length and in weight. Because of their small size it may be assumed that they are younger indivi- duals. ‘They took pieces of beef thesame size (105 mg.) as did the average ones, with the exception of two, which after two weeks ate practically nothing; this makes the percentage increment somewhat lower than it would have been had all eaten the normal amount. ‘The average initial weight of the set was 0.91 gram, and after a period of six weeks the percentage increment was A Study of Growth DI 48.08, the largest percentage obtained for animals taking three pieces of beef per week, despite the fact that two individuals did not eat after the first two weeks, which must have lowered the total increment. These results are also in harmony with the individual records showing that the less the initial weight the greater is the percentage increment. (c) As to the other question — which arose as a possible factor in accounting for the variations in the results, viz: the effect of moulting; the data were not sufficient upon which to base a sug- gestion. II INFLUENCE OF TEMPERATURE The series of experiments of 1906-07, the object of which was to test the effects of temperature above and below the normal on the rate of growth, was repeated. In the former series it was impossible to show the rate of growth at different temperatures because the amount of beef taken by the animals kept at a low temperature, 10° C., was insufficient to preserve equilibrium of those at 20° C. and 30° C. In the present series the temperature conditions were practically the same as before, but the feeding was modified somewhat. Six sets, consisting of ten individuals each, were kept at three different temperatures, as follows: Average tem perature deg. C. estan ( Mecabaless NONGD) ror clnterae fetid. qcatat cya chet alee] ote ta abate BOS tekareie Lois as iets $e) Sfaval DES tes] ue: :4 0 Kae eRRRE Spe ec eGeEn ceoreoommacaracnrocce cto. 20 Sip BU (GE EO.) RR See RRR ae Pin Bye ee yen OR Ade 30 Sah THY (Citvisils ON). obonanooopamuminoopes cogoupe ondoooDdscaenananone ahose 20 Siar 12 (MIS BORD pasogacoodpoasonete0 coho solbonoumtogoumasonoodannoDd. 30 Sat TD) GRE DOA) hn om dadbdetetos capo ocHcooome comm aoc Undone no ronCCO lac 30 -The amount of beef taken by D was used as a basis for D* and D®. For the first three weeks the feeding was very regular, all eating three pieces per week, averaging 105 mg. each; after this time, because of the effect of low temperature, the animals became difficult to feed. 22 Ada Springer After a period of ten weeks the results were as follows: Increment per cent hal Da cep poocueeandeomccsenonce ids saa cbodesunoasa maga uobecoesAbabescuEsc 24.8 Gifu D Soe gansccondue sopHunoGop nAacoasoo bonddcocardoMduoen DNS SOeC Cocos odoUS 12.5 Seteb2utheremwas ailossuOie qe everelocretslcicvele ciel sicloreict rater iene ele he etter nya 14.4 The amount of food in case of the set at 10° C. was sufficient to give a percentage increase of 24.8; the same amount at 20° C. gave only 12.5 per cent; while at a temperature of 30° C. it did not prove sufficient to preserve equilibrium, a loss of 14.4 per cent was the result. The individual records of three animals from each set were kept; the initial weights with the percentage increments being as follows: SET D (TABLE XX!) SET D! (TABLE XXI1) SET D? (TABLE XXIII) In, wt. Increment In, wt. Increment In. wt. Loss grams per cent grams per cent grams per cent INOS ID) Heo a e374: 20.5 No. (1).... (lost) No. (1) .:. 4.037 —12.5 Now (@)eee- 1-947 21.4 No:-@)i.. 3.207) 1973 No. (2).... 1.389 (gain) 2.4 Now. (3))tee 2-232 8.4 Wop (eons Haeeyf Sides No @)) eae 5077, ae During the period of ten weeks Nos. 1 in D and D? were fed thirteen times (105 mg. average weight of piece), this amount being as much as No. 1 in Set D would take. No. 2 of D was fed 16 times; No. 2 of D‘ 18 times; No. 2 of D? 16 times; No. Bron). 1 0L)7. 5) 3yenmmes: The result of the individual records were the same in general as those for the entire sets. ‘The irregularities of the percentage increments in the two individuals of set D' (Table XXII) may be explained, or an explanation suggested. No. 3 was a large female, the initial weight being 2.007 grs. ‘The steady loss was not due to a diseased state, but to the fact that the amount of food was not sufficient to preserve the initial weight. ‘This was shown later when the experiment had closed; the animal was fed as much as it would take, and a steady gain followed. ‘The high percentage increment in No. 2, Set D', may be due to the small initial weight. Set E! (Table XXIV) at 20° C. was fed as much and as often as the individuals would take food, and Set E? (Table XXV) A Study of Growth 23 at 30° C. was fed a corresponding amount. ‘The initial average weight and the percentage increments were as follows: In.ut. Increment grams per cent BS SES Seteresetots er srer ore) ehetare eve’ «ial a.sictaleyareuavabeiarst deve corsia aareistelays viele, eisai elas 1.738 49-7 BS Gta DS eoreterelc eat c¥s 7a ssyomat ss s'rave istays' es sisls aacvey eta ics nate iaicgeteieie sraisrave eto crete = 1.69 29.07 This shows that the rate of growth of the animals at the two temperatures for the amount of beef was more than enough in both cases to preserve equilibrium, and even to add to the body weight. The individual records of three animals from each set were as follows: set E! (TABLE XXIV) SET E? (TABLE XXyY) In. wt. Increment In. wt. Increment grams per cent grams per cent IN (cb: (@)) mel Seeeaooaoede 1.507 47.1 INS (Quel ceacoceeoc Lsl72: 57-4 INoan(2)) male)... sty) =< W527 Banos) 8 Now(@)imales-eryeeeaaee 1.539 22.8 Noo ((@)smaale ery \ere rite -1- 2.307 ZIT Now) imalesaas eae 2.805 10.5 During the period of ten weeks Nos. 1 in Set E! and Set E? were fed 38 and 40 times respectively (105 mg. of beefeach time); No. 2 in Set E1 was fed 42 times; No. 2 in Set E? 30 times; No. 3 in Set E1 45 times; No. 3 Set E? 45 times. With the exception of No. 3 in Set E’, the results show the same relative percentages as did the averages taken from the entire sets, viz: the gain was greater at the room temperature. No. 3 of Set E? was a female of small initial weight; this fact as interpreted in the light of the data cited above will not be difficult to explain. Lastly, Set F (Table XXVI), at 30° C. was fed as much beef and as often as the individuals would take, regard- less of amount given to other sets. It was found, however, that this set ate about the same amount as Set E1 (Table XXIV), and Set E? (Table XXV); and this may be considered the maxi- mum amount of food that can be taken. ‘The initial weight and the percentage increment for the period of ten weeks was as follows: Initialeweieht (grams) ena ricer ccc yee stich crs eer cies meee 1.661 Increment, (per: Cent) ocmeqe stam eicss sorte c¥etos vies, sia ar Lela tote tein tei eee 25.06 24 Ada Springer The individual records of two individuals do not agree entirely with the average results of the sets, as shown below: In. wt. Increment grams per cent INGE C0) etree entra Naso OU re RAS a omega TR EIA AO binia o's oc 1.107 56.4 Boe Ga) ya9 aba 8 ots, arate ciate dior Pm nes eich oe Lecce ote ais te 1.632 43-3 No. 1 was fed 40 times (105 mg. average per feeding), and No. 2 was fed 39 times. The general results of the experiments carried out to test the effect of varying temperatures seem to show: 1. More food is required at a high temperature to preserve equilibrium than at a low temperature. 2. The maximum amount of beef that the animals will take at a low temperature is, on the average, one-third as much as for those at the room temperature (20° or at 30° C.). At a low temperature digestion probably takes place more slowly, or it may be less food is taken because waste is slower and less material for repair is needed by the body. 3. When a definite amount of beef was given to animals at three temperatures, viz: 10° C., 20° C. and 30° C.,; this being the maxi- mum amount that those at 10° C. would take, the rate of growth was greatest at 10° C., less rapid at 20° C., and at 30° C. the beef given was not sufficient to maintain equilibrium. When, how- ever, the animals at 20° C. were fed their maximum amount (the same as the maximum for those at 30° C.), which was three times as much as that given to the animals at 10°C., the percentage increment was almost twice as great as that of the animals at 10° C., and was greater than that at 30°C. ‘The percentage increment of animals at 30° C. was also greater on their maximum amount of beef than that of the animals at 10° C. on their maximum amount of beef. , By comparing the ratios between the amounts of beef taken and the percentage increments in the above cases, it will be found that the rate of growth in proportion to the amount of food taken is preatest at 10° C., less at 20° C., and still lessiat g03-€- A Study of Growth 2 III RESULTS OF THE SERIES OF 1906-07 (PART 1) INTERPRETED IN THE LIGHT OF FACTS OBTAINED IN THE PRESENT SERIES OF EXPERIMENTS The facts obtained in the present series of experiments in many cases show or indicate the causes of many of the irregularities of the Series of 1906-07. In Sets At and A? (Table I), composed of three males and three females respectively, where the food and temperature conditions were the same, the percentage increment in A‘ was 23.3 and in A? was 29.8. ‘The initial weight, however, of the first was 2.006 grams, while of the second it was 1.661 gram. ‘The difference in initial weight would seem to account for the difference in the rate of growth. In Sets Ct and C? (Table III) there was a similar difference, not so great, but which may be interpreted in the same way. Initial Increment weight per cent SE Crome as oe anon Ar Ae Ae ean Geet ace TocUaanS 1.853 18.4 Sete 2b so .mmcis a eyae elracis soos bho a einar cinta SCY Tee e eee 2.11 14.5 In Set D! (Table IV) where the animals were injured by cutting off the tail at the base, and in Sets A! and A? the normal control, the initial weights and the percentage increments were.as follows: Injured animals Normal animals After In. ut. Increment In. wt. operation Increment grams per cent grams grams Marat StI So ocsdacadessade 2.066 14.3 Sete es wicsie 3. 2.253 1.986 19.5 Set As ire sfeleverere cielo Mens eie 1.661 20.3 After nine weeks After nine weeks Seta c enigma ees 2.066 Dake Sete Diy cn cites UC) 1.986 Rn Wie Mes oeoocoouR pAb aC 1.661 29.8 The percentage in the injured set was greater than in one control and less than in the other, but when the initial weight is considered the difference between the two control sets may be explained. Based on this comparison, the evidence in favor of a greater percentage in the injured set is more striking. 26 Ada Springer A similar comparison to the one cited above may be made in the following sets (Table V): Injured animals After Normal animals In. wi. operation Increment In. wt. Increment grams grams per cent grams per cent Setibeancies ac 2.264 1.887 14.7 Seen Gar scsi tias ceieniontensts 1.944 II.4 Sety Eze iasnec =: 2.044 1-73 FOURS ISEE Gere +. cannes saan Aint (HOSS) Taz Set Hee anus cvs 1.878 1.614 Tare | WSet Girt eee cee eee 1.965 (lors) 6.5 In the injured Sets E* and E’, the initial weight in one case was ereater, while in the other it was less than the initial weight of the normal control sets, G' and G°; yet the percenatge increment in both cases was greater in the injured sets. In Sets E™ (Table VII), in which the animals were injured by cutting the regenerating stumps, in E"® (Table VII) in which the regenerating stumps were left intact, and in G' (Table VI), the normal control, the average initial weights and percentage incre- ments were as follows: After In.ut. operation Increment In.ut. Increment grams grams per cent grams per cent Seige recat 1.904 1.835 21 See B. ies eee 2.543 Bea Taking into consideration the difference in initial weight, the large percentage of the injured set does not appear to be so anoma- lous. ‘This was also the case in the following sets (Table VIII) where the experiment was the same. After In. wt. operation Increment In. wt. Increment grams grams per cent grams per cent See be 5 so baacar 1.954 1.878 Dit se Satta paddies n ae 2.089 18.2 See (Gt scoocccccsssesce 2.18 I1.4 In Sets H® (Table IX), where the animals were injured by cutting off the tails at the base, and H? (Table XI), where the tails were left intact, the average initial weight, and percentage increments were as follows: A Study of Growth 27 - After In. wt. operation Increment In. wt. Increment grams grams per cent grams per cent Sel CSC aeatee 1.446 127 31.6 SetcHe acne aes ac incres 1.65 20.3 The difference in the initial weights was not great, therefore the percentage may be taken as showing more nearly a correct relation between the injured and the normal sets. In Sets F'* (Vable X) in which animals were injured by cutting regenerating stumps, in F!? (Table X) in which the regenerating stumps were left intact, and in H? (Table IX), the normal control set, the initial weights and the percentage increments were as follows: After In. wt. operation Increment In. wt. Increment grams grams per cent grams per cent Sep IBS aomaor 1.345 1.268 DIB Sete Ba meysrerteecs hae 1.359 23.5 SEE REIAR cat arrest oreiereyehe 1.65 20.3 This is the only case throughout the experiments, however, where the percentage increment was not greater in the injured set. In Sets I', I?, I? (Table XII) the tails were cut six successive times, and the conditions of food and temperature were identical. The initial weights and the percentage increments of the injured sets together with those of the normal controls, At and A’, were as follows: After In.wt. operation Increment In. wt. Increment grams grams per cent grams per cent Sete yates 1738 1.425 42. Setg Almere cer rer 2.066 2Re8 See 12 os Beta 2.118 i O73) 24.5 SG neudaocccabonso os 1.661 29.8 Sie | Bipeceddme 2.489 1.977 2a Initial weights here also seem to account for the irregularities. That the percentage in Set [* is greater than either I? and I?, is probably due to the fact that the average initial weight is con- siderably less. ‘This is true also in the control Sets A! and A’. From the comparisons cited above it is found that in some of the cases the initial weight of the injured sets (before the tails were cut) was greater than that of the intact control sets, while 28 Ada Springer in an equal number of cases the initial weight before the injury was less than in the intact sets. ‘The percentage increment in the majority of cases, however, was greater in the injured sets. This would seem to indicate that the greater percentage increment in the injured sets was not due alone to a smaller initial weight before the tails were cut. It must be remembered, however, that after the tails had been cut off the weight in the injured sets in every case was less than that of the intact sets. It would seem to follow that the greater rate of growth in the injured sets was connected with the decrease in the volume, by removal of the tails. “There is another relation to be considered, viz: whether the condition of the tissues themselves play a réle or whether the result depends simply on relative volume, that is, whether there is any difference between two animals of the same weight, one having been a large animal when its weight was reduced by cutting © off the tail, the other an intact smaller animal. ‘The data are nsufhicient upon which to base a conclusion. The following table shows four sets that were starved for a period of five weeks, after which time they were fed 153 mg. of beef per week; also four sets as controls which had been fed for five weeks, after which they were also fed 153 mg. of beef per week. ‘The initial weights and the percentage increments were as follows: After five weeks? starving After five weeks of feeding In. wt. Increment In. wt. Incremen grams per cent grams per cent Sep Hom (lablevix®)ae-tee 1.65 Zong SeteGs ((CRablesVilo) ss. me2s 11.4 Set. B14 "(Qliable 2X8)... 1.345 21.8 Set E™"(Dable\Vill )5.7- 1-904 21.1 Set Fb (Table X)....... 1.359 23.4 Set E!> (Table VII ).... 2.543 nee Set F86 (Table XT)....... 1.056 27.8 Set E: (Table VIII).... 2.089 18.2 The four sets after starvation gained faster than did the sets after a period of feeding. Set B! (Table II) was starved for seventeen weeks, after which it was fed 153 mg. of beef a week for a period of nine weeks. When the percentage or rate of growth is compared with that of normal sets, A! and A? (Table I), for an equivalent period, the result is as follows: A Study of Growth 29 In.wt. Increment In.wt. Increment grams per cent grams per cent Sid? LMecossdaceosoueedod - 0.89 54.2 Sarva eh eounoc cocasbonee 2.066 23-3 DEGMAGr He iookeire ctor 1.661 29.8 After starvation the rate of growth was faster than in the nor- mal animals. When Set B? (Table II) is compared with Sets A‘ and A? the result is the same. ‘This was probably due to the fact that starvation had reduced the initial weight. IV. THEORETICAL DISCUSSION It has been shown that when two animals, one larger than the other, are given just enough food to preserve equilibrium in the large one, the smaller animal gains weight. Assuming that digestion and assimilation in both cases are the same, how may the facts be interpreted? It may be safely assumed that it takes a smaller amount of food to preserve equilibrium in case of the small animal, than in the large one, so that the material over and above that used in actual repair of the body waste goes to form new tissue, to increase the size. According to this view, as the small animal approaches the large one in size, which is the maxi- mum weight for the amount of food taken as a basis, the rate of growth should become less. It takes gradually more and more material to replace waste, because of the new tissue that has been continually added; hence there is less to be used in the formation of new tissue. Even if more food were digested by the larger individual the result would be the same. By feeding the large one more than enough to preserve equi- librium, and the small one the same amount, the large one will gain; but the rate will be slower than that of the small animal, for the same reason as given above; the small one uses less for repair and more goes to form new tissue. When animals are injured by cutting off the tails, the increase in rate of growth that follows may be due either to a stimulus produced by the cut, or to an increase owing to the reduction in weight; that is, to a change in the relation between the body material and the amount of food taken. By cutting off the tails 30 Ada Springer at the base about 15 per cent (average) of the body weight is removed, so that it may be supposed the food material that other- wise would have been used to repair the waste in the tail goes to increase the body weight. It has been shown that the smaller the animal the greater its rate of growth; it has also been found that after a period of starvation and consequent reduction in size, the rate of growth is faster than in the animals in a well-fed condition. It is prob- able that when the body material is reduced in animals by cutting off the tails, the increased rate of growth observed may be due to the reduction in size, that is, in proportion to the amount of food taken, rather than to the stimulus of the cut. Whether in addition to this the cut may also act as a stimulus, cannot be affirmed or denied from the experiments so far carried out. The reduction of the body weight by starvation and by cutting off the tails cannot, however, be considered equivalent factors. In the reduction of the initial weight by starvation the condition of the tissue is changed, while when the tails are cut off the remaining tissue still remains in the same condition as before the injury. SUMMARY 1 Increase in weight in adult Diemyctylus is due to an increase in the size of many of the organs of the body, and not to a stor- age of fat. The converse is also true, that decrease in weight is due to a decrease in the size of the organs. : 2 Percentage increment is directly proportional to the amount of food consumed by the individual; the more food consumed the faster is the rate. Rate of growth decreases as the maximum weight is approached. This maximum is determined within limits by the amount of food taken; for a certain amount there is a definite maximum or point where there is established a state of equilibrium between waste and repair. By increasing the amount of food the weight may be increased and a newcondition of equilibrium be reached. 3 By cutting off the tails the rate of growth is increased. This increase in rate is probably due to a reduction in the size A Study of Growth a1 of the animal, although from the data it is impossible to determine whether it might not be due to some extent also to the stimulus produced by the cut. Injury in this case reduces the weight of the body without affecting the amount of food digested, therefore it seems reasonable to suppose that less material is needed for repair (there being less material to be repaired), and more goes to increase the weight than is the case in normal control animals. © 4. The initial weight is the main factor in determining the percentage increment or rate of growth; the greater the initial weight the less the percentage increment; that is, the larger the animal the more food is used for the actual maintenance of the body material, and the less goes to an increase in weight. 5 Sex may influence the rate of growth, but so far as the data ‘go only indirectly through the initial weight. ‘The average initial weights of the females were less than those of the males. 6 After starvation and subsequent reduction in the initial weight, the rate of growth is higher than in the normal animals. This increase in rate is probably due to the reduction of the initial weight. 7 During starvation the rate of decrease in weight diminishes as the temperature is lowered. 8 If the maximum quantity of food the animals will receive at the lower temperature be taken as a feeding basis, the rate of growth diminishes with the increase of temperature; that 1s, the rate is highest at the lower temperature and becomes lower as the temperature increases. The quantity of food the animal will take, however, increases with the temperature. It was found that the quantity of food taken by the animals at the intermediate and higher temperatures was three times that taken by those at the lower. ‘Taking as a feeding basis the maximum quantity of food the animals will take at the intermediate temperature (which was also that for those at the higher) the rate of growth diminishes with the increase of temperature. The rates of growth at the intermediate and at the higher temperatures on their own feeding bases were both higher than that at the low temperature on its feeding basis, but the difference in the rates was not in proportion to the difference in quantity of food taken which was three to one. 32 Ada Springer Temperature, therefore, while influencing the amount of food the animals will take, also influences directly the rate of growth. I wish to acknowledge my indebtedness and to express my thanks to Professor Morgan, under whose direction the work was done. Zoélogical Laboratory Columbia University A Study of Grow:h 33 TABLE I Ser Al ; Ser A? Showing normal rate of growth. Fed 153 mg. Showing normal growth as in A}, also the (average) of beef a week rate of starvation 2 rs d 32 a ; & E Fe © r= q 8 E ce) =e Se le e | 3 8 Oley Se 8 pe | ah a) hee Be Wee ae as fo Oct. 16] 6 12.396] 2.066 6 9.968) 1.661 23, 6 11.876] 1.979] —0.087| —4.3 6 9.856] 1.643] —o.018| —1. 30| 6 12.54 | 2.09 O.111 5-4 6 10.326| 1.721; 0.078 4.6 Nov. 6) 6 14.208] 2.368] 0.278 {2.4 6 11.967| 1.994) 0.273 Wan nay 14. Dee] =O) Sirf 6 11.995, 1.999| 0.005 0.2 20| 6 14.32 | 2.386] 0.053 —2.2 6 12.222] 2.037| 0.038 1.8 270 10 14.298] 2.383] —0.003 Qi 6 11.917} 1.986] —0.051) —2.5 Dec. 4| 6 14-735| 2-456] 0.073 Ne 6 12.775 2 hKs)| GIy} 6.9 Te 15.167] 2.528] 0.072 2s 6 T3E2E |) 22202] OnO7a ee] 18} 6 15.667] 2.611] 0.083 Aiz 6 13.46 | 2.243] 0.041 1.8 24, 6 16.655) 2.776] 0.165 6.1 6 13.572| 2.262] 0.019 0.8 Began to starve Jans2 | 0 16.275| 2.712] —0.064| —2.3 6 12.625| 2.104] —o0.158 Gia? Sino 16.347| 2.724| 0.012 0.4 6 MeV UG|| TAC yal, Che isi 7-4 nig) © 16.86 | 2.81 0.086 Ao 6 11.165| 1.861] —o.092 4.8 22| 6 17-492] 2.91 . 100 Bic 6 11.135] 1.856) —o.005 On 29] 6 MN Dee 6 \ 3.8 Feb. 5) 6 18.294] 3.049] 0.139] f 28 6 LORG22 (e721 O11 © if 3.8 12-16 19.087| 3.181] 0.132! 4.2 6 10 1.667) —0.053 Zoi 19} 6 19.43 | 3-24 0.959 1.8 6 9.705] 1.618) —o.049 2.9 26| 6 19.135| 3-189] —o.o61| —1.8 6 9.425] 1.571| —0.047 2.9 Mar. 5| 6 19.21 | 3.201] 0.012 Ong 6 9.275| 1.546] —0.025 1.6 12 46 isin@Vsl| Zi satistel| —te)efeyriy =the 6 8.575) 1.429] —O.117 7.8 I9| 6 18 .595| 3-099 ~0.059) =i st 6 8.237] 1.373] —o.056 3-9 GeO 18.065] J.011] —0.088) —2.8 6 Ton \\ toy |, Celis! 6.2 Began to feed 459 mg. (average) of beef a week Apr. 2| 6 19.065] 3.177] 0.166) 5-3 6 7.289] 1.215] —0.075 5-9 9) 6 20.142| 3.357] 0.18 af 7.235| 1.206] —o.009| 0.7 16| 6 20.685] 3.447| 0.09 6 6.955| 1.159] —C.047 3.6 From From Oct. 16 to Dec. 18 ....| 2.338] 0.545) 23.3 Oct 16 to Dec. 18] 1.952] 0.582] 29.8 Oct. 16 to Deb. 19) =. -)) 2-653) 1.074) 442 Oct. 16 to Nov. 20} 1.849 0.376 20.3 Dec. 18 to Feb. 19 ....| 2.925] 0.629] 21.5 Dec.24 to Apr. 16) 1.710 1.103) 64.5 Oct. 16 to Nov. 20 ....| 2.226) 0.320] 14.3 H 34 Ada Springer TABLE II Ser B Set B? Showing the result of starvation and of subsequent Showing the result of starvation and of feeding subsequent feeding ee a 2 ed a 3) 2 Se z 5 Bet < en eas ae : | Pee Nae : Shes vain 2 , pe aes ae 9c g : ele ee | 2 | eae Oct 16/6 I1.561| 1.927 6 11.756 1.959 22 | eo 10.761) 1.794) —0.133 Tal 6 11.161] 1.86 | 0.099 5-1 30 6 10.278| 1.713) —o.081 4.6 6 10.376 1.729 0.131 Wes WOW) (6 9.76 | 1.627| —0.086 Rol 6 9.983 1.664, 0.065 a5 rig g.292| 1.548} —0.079 4.9 6 9-51 1.585) 0.079 4.8 20| 6 8.784 1.464) —0.084 Gals 6 9.068) 1.511) 0.074 Chey) 276 8.2 | 1.366) —o0.098 6.9 6 8.787 1.465 0.046) Bo Decaur4|n6 8.059] 1.343] —0.023 1.6 6 8.31 1.385, 0.08 5.6 od) 7.629) 1.271) —0.072 RAS 6 7.722| 1.287] 0.098) Wes 18] 6 iter || Wotitsy|| =Caesy 6.8 6 7.465 1.244, 0.043 Ba Began to feed 153 mg. of beef a week increase | % inc. al 16 6.842 1.14 | —0.047! 4. 6 8.926) 1.488) 0.244 yf jane 6 6.597 ei —0.04 3-5 6 8.435 1.406) —0.082| —5.6 S06 6.094, 1.016 —0.084) 7-9 6 9.002) 1.5 0.094 6.4 15} 6 5-925] 0.988) —0.028| 2.7 6 9.54 | 1.59 | 0.09 5.8 22| 6 5-905) 0.984] —0.004) 0-4 6 MAG || trey 0.12 72. 29 \ 3-45 6 | \ 4 Feb. 5| 6 5.505| 0.918 —0.066 J 3-45 6 ea) Ja) a ™45) [ 4: 2h ig 4-45 | 0.89 | —0.028 ae 6 11.34 | 1.89 | 0.035 1.8 Began to feed 153 mg. (average) of beef a week | increase % inc. 19} 4 | 5.007] 1.251 0.361 33-7 6 11.982] 1.997| 0.107 Se 26) 3 3-94 | 1-313, 0.062 4.8 6 11.95 | 1.991} —0.006| —0.3 Mar. 5] 3 4.085] 1.362, 0.049 3°5 6 12.125] 2.021] 0.030 1.4 12 4-005] 1.335} —0.027/ —2. 6 11.785] 1.964; —0.057| —2.8 KO | es 4.103] 1.368] 0.033 2.4 5 9.685] 1.937| —0.027| —1.4 AAS) 9) 4-105] 1.368 0.00 0.0 4 8.322] 2.08 0.143 Tol Apr. 2 3 4.085 1.362) —0.006| —o0.4 3 6.497} 2.166] 0.086 4. 9) 3 4.285] 1.428) 0.066 4:7 3 64.9) || 221163)" — 0.002) Onn 16] 3 4.655| 12552) Onra4 8.3 3 6.4351 2.145) —0.018) —o.8 FE|| el 1 ee oS; 0 « “ Geo creer oreo aoe From From Octar6itopfebsn2zear T4082 7] we 7300 Oct. 16 toDec. 18} 1.601] —0.715| —44.6 Octin6ito Dec 13 a eke G7 O-740)| e475 Dec. 18 toApr.16} 1.694) 0.901] 53.1 Heboyizitov Apr. 16%.) .-|/m.220) OnOGzN 54h Dec.18 to Feb.19} 1.620] 0.753) 46.4 Octar6ito! Dec! iter 1.599] 0.656) 41. | Dec. 11 to Feb. 12..... 1.08 Ongotle as 2 A Study of Growth 35 TABLE III Ser Cl Ser C? Showing normal rate of growth. Fed 102 mg. of beef Showing normal growth asin C1, Fed 102 (average) a week mg. of beef a week Z| 2 $ & = = = “a I 5 v Y = a ES o ¥ 2 oe rah eee y Soiae ie aes g 2 iS Es z 5 3 iS) S 3 H A A < 4 ow Z B < 4 ow Oct. 16} 6 11.118) 1.853 || worl win 226 II.131| 1.855} 0.002 o.1 6 | 12.805) 2.134] 0.024 holt 30| 6 II.277| 1.879] 0.024 1.2 6 | 12.991| 2.165) 0.031 1.4 Nov. 6) 6 11.966} 1.994] 0.115 5-9 6 | 13.821) 2.304] 0.139 6.2 13} 6 11.815] 1.969] —0.025| —1.2 6) |) Tg5O5|) 22257|) 0.053 e218 20| 6 11.595] 1.933} —0.036| —1.8 6 | 13.515] 2.252} 0.001] 0.04 27 6 II.79 | 1.965] 0.032 1.6 6 | 13.197| 2.2 =O) O5 22a Wecs 456 12.305] 2.051] 0.086 4.2 6 | 13/85 | 2-308] ‘07108 4-7 11; 6 nits || TeGEG] =Oouny| Sets 6 | 13.777| 2-296] —0.012| —90.5 18} 6 Tah a7 e223 0.295) 14.1 6 | 14.64 | 2.44 0.144 6 Began to feed 153 mg. of beef a week 24, 6 14.97 | 2.495] 0.265 Dez) 6 | 14.815] 2.469] 0.029 Hiei Jane. 26 14 T77\\ 2-403\) 0-032) — Taz 6 | 14.845) 2.474] 0.005 0.2 8} 6 14.977| 2.496] 0.033 ii5g) 6 | 14.897) 2.483] 0.009 0.3 m5 © 15.44 | 2.573] 0.077 3-03 6 | 14.775] 2.463) —o.02 | —o.8 22| 6 17.324] 2.887) 0.314] I1.5 6 | 15.305] 2.551] 0.088 BES 29| 6 Deo) On, } 3- Mey Gi © 18.105] 3.018 al 2.2 6 | 16.285] 2.714] 0.163 Be 12} 6 18.997| 3-166} 0.148 AT Cm 15.095] 2.516) —0.198] —7.5 19} 6 19.06 | 3.176] 0.01 0.3 6 | 16.037) 2.673) 0.157 6. 26, 6 18.555| 3-093] —0.083] —2.6 6 | 16.035 2.673 0.00 O. Mar. 5) 6 18.92 | 3.153] 0.06 1.9 6 | 15.965| 2.661; —0.012] —0.4 12} 6 18.365] 3.061] —0.092] —2.9 6 | 15.077| 2.513} —0.148) —5.7 19} 6 10}67/2) |) Bheu2 0.059 1.9 6 | 15.072] 2.512, —0.001 0.03 26| 6 18.425] 3.071] —0.049] —1.5 6 | 15.125] 2.521| 0.009 0.3 Began to feed 204 mg. of beef a week AGT a2 G 17-955| 2.992] —0.079] —2.6 6 | 14.865) 2.477 —0.044, —1.7 9) 6 18.295] 3-049] 0.057 1.8 6 15.607) 2.601] 0. 124| 16] 6 18.965) 3.161] 0.112 3.6 6 | 15.875, 2.646 0.045 eres ce #9 See wees Se eee oes s§/ § og aie |e 5 3 Pag oe Bs aes alt | on | Ay a wa — Ay & From From | | Oct. 16 to Dec. 18....| 2.046 0.377| 18.4 Oct. 16-Dec. 18] 2.27 0.33 14.5 Dec. 18 to Feb. 19.. a 2.703; 0.946) 34.9 Dec. 18-Feb. 19] 2.556) 0.233} g.1 Oct. 16-Feb. 19} 2.391) 0.563, 25 36 Set D! Ada Springer TABLE IV Rate of growth after cutting tails off at base. Fed 153 mg. of beef a week Set D? Tails cut at base as in D} Gy ie a = z <3 | ee ae : Bo] gee d ze z P - lee 5 & F ¥, + g fe P ¥, 2 8 Za se |) eal x 2 | | as x Oct. 23 + tails + tails 135 1O|e2 253 II.491| 1.915 Gen atarls 6 |— tails 11.916 1.986 9.998| 1.666 30) 6 13-478] 2.246] 0.26 122) 6 11.474] 1.912] 0.246 127 Nov. 6} 6 13-477| 2.246] 0.00 0.0 Began to starve 12 14.487| 2.414] 0.168) 72 20|) 16 13.745] 2.291 SOng|| ee? 2716 me 11.537, 2.307, 0.016 0.6 ° Dec. 4) 5 II.955| 2.391; 0.084 055 | a5 12.767] 2.553] 0.162 6.5 18} 4 10.71 | 2.678) 0.125) 4-7 6 8.892] 1.482 | Fed 153 mg. of beef a week DAN Xe) 8.096 2.698) 0.02 | ony 6m is21g 1.869) 0.387, 23.1 jane 2g 7-89 | 2.63 | —0.068) —2.5 6 II.412] 1.902| 0.033 iGny/ 81 3 8.085) 2.695, 0.065 hy 6 | 11.865! 1.978] 0.076] 3.9 DS 3 8.2 | 2.733] 0.038 1.4 6 13.105, 2.184 0.106] 5. 22) 3 7-545, 2.515} —0.218 838! 6 | 14.25 | 2.375 0.191) 8.3 29} 3 | | leas 6 | | 2.5 Bebra |3 8 765) 2.922 0.407] { 7.4 3 |i tie Aipls 0.125) Dak 12) ng 8.84 | 2.943) 0.021 0.7 | eGo) 2 756 0.256) 9-7 19} 3 8.92 | 2.973] 0.03 1.01 6 | 16.5 | 2.75 | —o.006] —o.2 26} 3 9.12 | 3.04 | 0.067 222 6 16.955; 2.826 0.076) 27 Mar. 5) 3 8.965, 2.988) =O. O52 leer? 6 | 17.565| 2.928] 102 Zio 12\) 53 8.78 | 2.926 —0.062| —2. 6 16.689) 2 782, SCs) =i IO a3 8.657) 2.886) —0.04 Tie Gill 1673 7) 62 789) 0.007) One 26)5 3 8.52 | 2.84 —0.046] —1.6 6 | 17.17 | 2.862] 0.073} 25 Began to feed 408 mg. of beef a week Began to feed 306 mg. (average) beef a week iNoye, Ff Sine | Boge} || Osu = s6G) 6 | 17.225] 2.871] 0.009 0.3 93 9.295] 3.098} 0.368 12.6 6 | 18.44 | 3.073] 0.202 6.7 16} 3 9.105} 3.035] —0.063] —2 6 | 19.205] 3.201] 0.128 4. From Oct. 23 to Dec. 24 | 2.342] 0.712 30.4 From Dec. 24 to Feb. 26 | 2.869] 0.342 | 11.9 Dec. 18-Jan. 15) 1.833] 0.702| 38.2 Oct. 23 to Nov. 27 |. 2.2 0.428 | 19.5 A Study of Growth TABLE V 37 Ser E? Set G? Rate of growth after cutting tails off at base. Fed 153 Normal intact control. Fed as in E* mg. of beef a week = . a 2 a 2/2 |? am ones : E o B nl AS BS E ae : P y ¢ F at lone x = Fe S a Ss © 5 a) —_ s g 5 3 sa ae 2 5 Seer ies £ s A a < 4 Py Bo) S < 4 Aa |+ tails 20.44 | 2.044 Noy. 13) 10 |-— tails ZO) | 2y.112| 2-001 ost || seelyet 17| 10 | 18.047] 1.805] 0.075 4.2 LOW) | 211-2072 121 Ono 0.4 20{ 10 18 .035| 1.804] —0.001| —0.05 Mey || Bee | eset) Siext |} 1136) 27 tO pases || wag! Cite) —iioh 10 | 20.067| 2.007| —0.074| —3.6 Dec. 4] 10 18.347] 1.835] 0.052 2.8 10 | 20.095] 2.01 0.003 o.1 ri] i) 19.067} 1.907} 0.072 3.8 Io | 20.445] 2.045} 0.035 17 18] 10 19.217| 1.922] 0.015 0.7 Io | 20.794] 2.079] 0.034 (6 By Soles cebe il dees a & z Ge) ¢ | 2 ff] , | 28 Peg ge ee oe From From Nov. 13 to Dec. 18....| 1.826| 0.192 10.5 Nov. 13-Dec. 18] 2.095] —0.032| —1.2 Ser Et Ser G! Same as Set E? Same as G? + tails | | | Nov. 13 22.635) 2.264 10 6 | — tails 10 | 19.442) 1.944! 18.865) 1.887) 17| 10 18.995| 1.899} 0.012 0.6 Io | 18.936 1.894, =O5 |] aE 20] 10 19.256] 1.926| 0.027 1.4 10 | 19.655] 1.966] 0.072 BT 27 10 20.295] 2.03 0.104 5.2 10 | 19.535| 1-954| —0.012| —0.6 Dec. 4] 10 20.765) 2.077 0.047 2.2 Io | 20.189 2.019) 0.065 Af II| 10 21.627) 2.163) 0.086 4. Io | 21.09 | 2.109} 0.09 4-3 18} 10 21.882. 2.188} 0.025 Dad Io | 21.795] 2.18 0.071 Be3 | > — “oO o = oF) eg | és gf) ¢ | 88 TO aaah 73/4 | 83 From From Nov. 13 to Dec. 18....| 2.037} 0.301 14-7 Nov. 13—Dec. 18] 2.062] 0.236] 11.4 Ada Springer TABLE V—Continued Set E% Set Gt Same as Set E? Same as G? oa —— —— & 30 icy) Sg & ak o g “eI as eS | — a I Ee = ss s o | wo =| Cs = vo vo r=) oy on Sc oO tn = a o iS) a fe © ° © i) be v o 3 Bale 3 5 S Seer Z B Z eae 4 om Z Ey alec 4 = + tails 18.778] 1.878 INGvamiia)|N LON stalls 10 | 19.65 | 1.965 16.14 | 1.614 itll 1K) 16.004, 1.6 | —o.014; —o.8 IO | 19.785, 1.979, ©.014, 0.7 20] 10 16.156) 1.616 0.015 0.9 IO | 20.19 | 2.019} 0.030) Die 27| 10 16.342| 1.634) 0.018 Wai 10 | 19.895] 1.99 | —0.029| —1.4 Dec. 4] 10 17.349| 1.735| ©.101 5-9 10 | 20.26 | 2.026) 0.036 ay) WE} ie) 17.469] 1.747, 0.012 0.6 IO | 20.595] 2.06 | 0.034 1.6 18) 10 18.62 | 1.862) 0.115 6.3 10 | 20.988 2.099 0.039 1.8 ae 8 gg! Pec ee oun |, 3 & 5 ae ee Soe Bre | Ue ooh eh Ee aie eee iar Ve =| a. 5 a. From | From Nov. 13 to Dec. 18....| 1.738] 0.248 14.2 Nov. 13—Dec. 18] 2.032} 0.134 6.5 A Study of Growth TABLE VI Ser EB? Set G! Rate after cutting tails off at base. Fed 153 mg. Normal control. Fed as in E% of beef a week dt of = : 2 2 =, S g > | 2 g g Vf 4 | a a te = os a S s B Bo 2 =| a 2 ¥, 9 g 6 = 5 3 5 ro) -_ s © 9 3 Suillus S E 6 = g 5 Z A < 4 oe Z a < 5 om + tails 18.778| 1.878 Nov. 13] 10 |-— tails 10 19.442) 1.944 16.14 | 1.614 20| 10 16.156] 1.616) 0.002 o.1 10 19.655] 1.966, 0.022 ea 27| 10 16.342| 1.634) 0.018 Tee 10 19.535| 1-954] —o.012|. —0.6 ECan LO 17-349] 1.735| 0.101 BaG) 10 20.189 2.019) 0.065 3.2 II| 10 17.469| 1.747| 0.012 0.6 10 21.09 | 2.109] 0.09 4-3 18] 10 18.62 | 1.862] 0.115 6.3 10 21.795) 2.18 0.071 an8 24| 10 19.7 | 1.97 0.108 5.6 10 23.535] 2.9541 oc174 7.6 Jan. 2] 10 19.547| 1.955] —O.015 | —0.7 10 23-367, 2.337| —0.017) —0.7 8} 10 20.222| 2.022| 0.067 3-4 10 Bel 2.485, 0.148 6.1 I5| 10 21.125] 2.113] 0.091 4-4 10 ipaea 2AAG| —On04i | ea 10 22] 10 22.485] 2.249] 0.136 622) 10 26.185) 2.619] 0.174 6.8 29} 10 4. 10 4-4 Rebs 95) 10 24.405| 2.441 0.192 Ate 10 28.615) 2.862) 0.243 4-4 12) 25.268] 2.527/ 0.086 3-4 10 29.105 2.911} 0.049 1.6 19] 10 26.255] 2.626} 0.099 3.8 10 30.597 3.06 0.149 4.9 26| 10 26.192] 2.619] —0.006 | —0.3 8 24. 105|| 3/5021), —-O-039| 1 1-4! Mar. 5] 10 26.645) 2.665) 0.046 1.7 8 23.515] 2.939] —0.082| —2.7 12| 10 26.725] 2.673} 0.008 0.2 8 23.385] 2.924) —0.015| —0.5 19} 10 26.875] 2.688) 0.015 0.5 26} 10 26.49 | 2.649] —0.039 | —1.4 Api. 2] 10 26.115] 2.612) —0.037 | —1.4 9) 10 26.425] 2.643] 0.031 gi 16] 10 26.475| 2.648} 0.005 O.1 | “oO re) + & ~~ Pr are j>5| 8 aon a S| i cE From Noy. 13 to Feb. 19 (ee 1.012 47-7 From | | Dec. 18 to Jan. 15 | 1.987) 0.251 12.6 Dec. 18-Jan. 15) 2.312} 0.265 11.4 Jan. 15 to Mar. 19 | 2.4 0.575 | 23.9 Nov. 13-Feb. 19] 2.502! 1.116) 44.6 Nov. 13 to Jan. 15| 1.863} 0.499 26.7 40 Ada Springer TABLE VII Set E!? Ser E1b Regenerating tails cut off at base. Fed 153 mg. of Regenerating tails intact. Fed as in E\* beef a week ea Ee lice a ee g ep o r g rs o FS a ‘3S 2 Es a oS Z a 4 g 2 | : B g 2 & Ss = < $ 8 3 2 Ss 3 = cS 5 x £ o 5 Lal ° 9 jo} As aie ieee 147 A oN Ces ON fears a a + stumps 9.52 1.904 stumps 5 12.705! 2.541 IDEs US) 0.205 | 0.041 — stumps 9-177 | 1.835 PA TH | ithe! 2.00) 0.2250) ey 5 TQe0E | 2.622) 8 (OLS | (emia Jan: 2) ~5 | 10.311 | 2-062)| 0.002 0.09 5 12.927) 2.505) —O.037\) 1.4 8| 5 |. 10.837 | 2.167| 0-105 4.9 5 13.14 | 2.628] 0.043 1.6 15a | Llesa7e 2-207|,0.10 os 5 13.414| 2.683] 0.055] 2. coed es ag es ene a8 I Lal From From Dec. 18 to Jan.15..... 2 eORT 102492) ||) 2. Dec. 18 toJan. 15] 2.613] 0.142} 5.4 A Study of Growth | 41 TABLE VIII Ser E%4 Set E2> Regenerating tails cut off at base. Fed 153 mg. of Control. Regenerating tails intact. Fed beef a week asin B® & a=} & me =} e| 4 o g B | @ | oo 2 3 5 2 5 5 o B 5 os o o = os = o o — oe B oo 77) =] as A) 72) =| ro) =z S $ rs ) s s A} 8 3 3 $ 3 5 3 6 $ 3 5 Z A < 4 ow Z A < 4 = + stumps JEU Ls stumps Dec. 18) 5 0.21 5 10.447| 2.089 — stumps 9-39 1.878 . 24 5 10.436 | 2.087] 0.209 | 10.5 5 11.165] 2.233] 0.144 6.6 janie ines 10.535 | 2.107] 0.02 0.9 5 II.167| 2.233] 0.000 8| 5 11.417 | 2.283] 0.176 8. 5 12.342] 2.468] 0.235 | 10.0 1515 11.63 2.326] 0.043 1.8 5 12.545] 2.509] 0.041 1.6 ~o © o 2 x S & a a 3 & a Ss) pe| g | bs eee | ee = Lal From From Decyx18) to fan 15)... ||) 2.102] 0.448 |) 21.3 Dec. 18 toJan. 15] 2.299] 0.42 18.2 42 Ada Springer TABLE Ix Ser H3 Ser H? Tails cut off at base after 5 weeks starvation. Fed 153 Normal control after 5 weeks’ starvation. mg. of beef a week Fed as in H? seid las poy tas Nees 3 eal es ful) pee lore Z & a Wane £ Ea, Ware lice | = | & | 3 3 a gy % rs) S ca yy FA a OH = 3 o = 2 Ss oe is) ad < ® | cs) ° =] ny @ ° é = g 3) | a S iS) > 2 o Z BR < 4 by Bi Z A < 4 = +tails | 14-455 | 1-446 | | tails | Dec. 18} 10 1.515 | 0.152 | 10 | 16.495 | 1.65 — tails | 12737025) e277 24) 10 15.242) 1.524] ©.254 | 18.1 LO!) |) 19222) er 09221) (On2 72) aide janes 2) 6 16.357 | 1.636} 0.112 he 10 | 18.857 | 1.886} —0.036| —1.8 8] 10 16.562 | 1.656) 0.020 1-2 10 | 19.632 | 1.963] 0.077 4. EG oO 17-477 | 1.748] 0.092 5-4 ro | 20.229 2.023 0.06 3 | a =e ot 23> eee gs & g a 3 5 & g =r Seals me ae ace 2 eae From From Dec. 18 to Jan. 15 ....| 1.509] 0.478] 31.6 Dec. 18toJan. 15] 1.836) 0.373] 20.3 A Study of Growth 43 TABLE X Ser Fi Ser Fib Regenerating tails cut off at base after 5 weeks’ starva- Regenerating tails intact after 5 weeks’ tion. Fed 153 mg. of beef a week starvation. Fed as in F'® 4 ~ A - a ~ = | . a fee We : Slee eo : 3 5 EB | Fe 3 S | = s 2 So 2 | y 2 Sp 2 q 3 2 s S g 2 a 2 3 5 : pe) o (3 a . vw bs re Pree esi meee ol fi Ze were. Wee eee | ae | = ——_|——_|_ a + stumps 6.715 | 1.345 | stumps G 6.795 | 1-359) | Dec. 18) 5 0.185 —stumps 6.34 | 1.268 | 24) 5 7.855 | 1.571 0.303 Zia 5 8.4 1.68 Oeeyil| — Paitel Jan. 2) 5 F735 ele S4 70.024 Tas 5 8.025 1.605, = GVO 5) aud Sins Go || UolYu| —Oxcorl] Csi 5 8.309 | 1.662) 0.057 Bel lif) 7-897 | 1-579] 0.035 2.2 5 8.6 1.72 | 058 a4 22s atk) || Mogiyfs| Teles s}) Tata 5 9-575 | 1-915) -195| 10.7 | 29 4.5 5 I 2.9 Reba) 15) 5 9-745 | 1.949] 0.171 4-5 5 10.15 | 2.03 | 0.115 J 2.9 12/5; 10.175 | 2.035] 0.086 43 5 | 10.825 | 2 165, 0.133| 6.3 1O| 5 10.35 | 2.07 0.035 ey) 5 | 11.192 | 2.238} 0.073 Roa 26) 5 10.755 | 2.151] 0.081 3.8 5 | 11-225 2.245, ol 0.3 Miaens|) 5) 10.805 | 2.161] 0.01 0.4 5 | 11.335 | 2-.267| 0.012 0.5 T2}) 35 HOS |) Borwill| Certo) 511 5 | 10.033) || 2.207) —0.06 | —2-6 19 5 HORII |} Yo rtiGil| —Lo)(o%ey | git 5 | 10.915 2.183 —0.024| —I. 26) 5 ital |) Pei Cari 5.1 5 | 11.31 | 2.262} 0.079) Bas Began to feed 204 mg. of beef a week Began to feed 408 mg. of beef a week As Al 5 Tt 0 || 12.2221 0.000 0.0 5 I1.395| 2.279] 0.019 0.8 a 5 Hit ask O.1II 4.8 & 12.605] 2.521]/ 0.242) 10. 16) 5 12.405} 2.481) 0.15 6.2 5 13.205 2.641] 0.12 Tals beets as = BC eee ees Piece. oe OR pia tess: $3) € | 28 Fela ene | See From From Dec. 18 to Jan. 15 -.... 1424 Osh | eee Dec. 18 to Jan.15| 1.539] 0.361) 23.4 Dec. 18 to Mar. 26....| 1.744] 0.954] 54.6 Dec. 18 toMar.26| 1.81 | 0.903| 49.8 44 Ada Springer TABLE XI Ser F# Ser F3b Regenerating tails cut off in middle after 5 weeks’ starva- Regenerating tails intact after 5 weeks’ tion. Fed 153 mg. of beef a week starvation. Fed as inF* | = a a a é = | 3 g a | & | 8 g a é % | & z aa ee 2 ay a (5) =) os {8 ° = s v = S 3 S o e é 2 Oo oO M fo) 5S 2 5 Us| 2 ae) At ee es é Z| ee ae a +tails 6.36 1.272) Dec. 18) 5 | —tails 5 5.28 | 1.056 6.286 | 1.257 24| 5 7-592 | 1.518] 0.261 | 18.8 5 62567) ||) 3131) o.257 eae |B 2) 5 7-595 | 1.519] 0.001 0.06 5 6.43 | 1.286] —0.027 8| 5 7-869 | 1.574| 0.055 3-5 5 6.72 | 1.344| 0.058) 4. TS) 5 8.375 | 1.675] 0.101 6.2 5 6.987 | 1.397| 0.053) 3.8 aS o = | Gi 2 . gat | a Wi gue ge) 3 Soe SS es oe = .3)| 8 sue From From Dec. 18 to Jan 15. ....| 1.466] 0.418 | 28.5 Dec. 18 toJan.15| 1.226) 0.341| 27.8 6°z £Lo-o grsic. |Zor sn gl gto'o orcee, Snes or aon fo'o Sbr-z | Lo +1 9? rae) Egi°z 11 L-1— | zvo-o— | Sib-z |Lev-b1 *¥— | ggo-o tgo'z | Lob z1 38 6°8 I1Z‘0 “Svc | vl o1 tor || trz-0 6g1°z |S10° £1 gbz-z |SLb kr gS6°1 |S€L-11 $9 Lt1"0 sie] — Sr 6g0°0 sprey + LeE-z | go-br 110°z |Sgo'zI MAY ae SIME) oz'z |voz fr zeb"r |ztS*11 z'9 giro sje} — SoS) toro sie} — 61f-z |L16°f1 6g6°1 |gf6-11 Seat SUBS Se Igi'z |160°f1 Seg: |Lot-11 “€ ggo"o sey — f-9 Z1'0 spre} — = glz-z | Lo fr Sg6:1 |€SZ-r1 3 SM ar SEES Sr O ieee oz br gtg'1 |Si1o°11 IL'o g10"o spre} — I'O— | Z00'o— sprey — Ss gbz-z |Leb: fr 6gg'1 |See-x11 =a STEED oats SIME Se < ziz-z |S6E-fr 16g'1 {Sherr a $5 — gor°o— spre} — ++ ggo'o— sey — ate en zb6'1 |bSg°It ‘ wane 4 — ggt-z |lzt-r1 fo:z |gli°zr SAL = SEY 6gh-z |St6-+1 glitz |SoL 71 SUB ar SIME. ir eae ee deus 8 “ s | = 8 : eae 1] sv auvg 1] sv auvy sf 14S eL TAS IX ATaV.L gt | Soro Lie goo gi 6to'o 76 gI'o 9°6 LLi‘o “L Szi‘o 8 gtt-o EC) ZI0'O Soe I0°O— g 2 “st 9% S =) rg = ig 661° Sto: 616°1 Zbl t 17g’ \o lon eo} — TS ~ isa} + wn No) No) 4 i] Lea) ~~ o) _ qysiam advI2AV 761 tt Sg6:z1 SoS: z1 [aren Lori spiey — tiS-i1 sprey + |bSb-o1 syrey — Sz6'o1 syrea + LLY‘ o1 sprey — SoS :o1 spre} + 69°6 SpioL — 616°6 Sp}08 Sg°6 S[k — 6f0°o1 S]iD1 I'Ol S/O vor spiey + WYZIOM [eIOT, woeeowe So S[PUITUR JO “ONT Si 8 6 oix{l $z gl II y “Deg Lz oz €1 “aon yaam v faaq fo ‘du ES pag “Saudi IQISSIIINS 9 ind SPOL i 24g Alla Springer a - fai; a6) e TABLE XII—Continued I) ...|/From Noy. 13 to Jan. 15 IP so6 Noy. 13 to Jan. 15 JE Gos Noy. 13 to Jan. 15 (average) Mean 1.976 7 Mf ay — * ww 2 eae oS os me [8 0.774 | 42.7 0.486 | 24.5 0.531 23-5 ., Mi i Z fe Ls is Feb. A Study of Growth 47 TABLE XIII Set G2 Ser G3 Tails cut 6 successive times. Fed 153 mg. of Tails cut Q successive times. Fed as in G*b beef a week 4 a a s “3 oa Ps =| “BI 3 E ot cra aia ® | 8 & ci i Reel Men Mies g Seca). eee |) Seg cle SLiliree |i (a cane ge es x 2 Nese ites tee ates + tails} + tails 9-437) 1-887) 9-75 | 1-95 tails — tails 18) 5 0.22 | 0.034| Dec. 18) 5§ 9-545| 1.909 — tals | + tails 9-235, 1.847 9-97 | 1-994 + tails all i = ae 0.085] 4.3 10.209] 2.042) 0.195) 10. 9.867) 1.974 24| 5 tails + tails ©.197| 0.029 9-76 | 1.952 — tails | Janeee |) Ig = tals —0.022| —1 1 10.064) 2.013 9-57 | 1.914 + tails | + tails 10.03 | 2.006, —0.007| —0.3 10.047| 2.009 BN gi. > paeks | 8| 5 | —zails —0.095, 4.8 0.17) 0.034, 9.857) 1.971 — tails | + tails 9-914, 1.983) 10.429] 2.086 + tails uG|| Is — tails Qoine) Hale 10.295| 2.059} 0.076) 3.7 10.07 | 2.014 | 8| 5 tails | | + tails 0.252| 0.054) 10.805} 2.161 — tails, 22) ene — tails 0.147 7 9.96 | 1.992) 10.622] 2.124| + tails + tails 1O.9) | 2.16 0.168) 8.2 II.055| 2.211 15| 5 | tails 29/05: a taals 0.087} 4. 0.335| 0.067] 10.76 | 2.152 — tails| + tails 10.366) 2.073 Te Slee + tails | 0.152] 7. Feb. 5] 5 | —tails 0.148, 6.6 22| 5 I1.125| 2.225 | TL.15 | 2.28 — tails) | + tails 10.961) 2.192 TiN A) 285 29| 4 9.055| 2.264] 0.072/ 3.2 12) Staal Cpivye|| G7 5 3 7.465) 2.488] 0.224) 9.4 11.425) 2.285 12s Talat! Palsehss) Ciely |) tat) 19} 4 9-637] 2.409] 0.124) 5.2 19} 3 7 .895| 2.631] 0.093 21655 26| 4 9-155) 2.288, =O212il| = Gott 26) 3 8.or | 2.670] 0.039 1.4 Mar. 5| 4 9-967 2.492 0.204, 8.5 48 Ada Springer “TABLE XIII—Continued me ; g “Bo ‘Ss & § 2 a e iy nD Cc ‘S| es S s 8 A s 5 K 2 ee Ala gees gy Mees deck é Mar. 5] 3 8.077] 2.692 0.022 0.8 12| 93 7-905 22835) —0.047| —1.7 19] 3 7-975| 2.658] 0.023 0.8 26} 3 7.965] 2.655) —0.003 O.1 ApIem 23 7-965) 2.655 0.000 S o 5 aa ie 2 8. ee From Decrx8 to Peborg. 3 ja2/104| ate TOl ish? .8 Big ya hh Aad = ae Mar. 12] 4 9.825) 2.456 19] 4 9.845) 2.461 26] 4 9-79 | 2.447 Apr. 2] 4 9-555| 2.388 9} 4 | 9-795) 2.448 16] 4 | 9.935] 2.484 ee s 5 eas From Dec. 18 to Mar. 19 | 1.997 3) aes g r= ao o 5 ez 5 a 0.036) 1.4 0.005} 0.2 —0. 014 Ong —0.059| —2.4 0.06 2.4 0.035, 1.4 2. aaa o ous Pa sy es 3 | ae 0.974, 48.7 A Study of Growth 49 TABLE XIV Ser F! Ser H! Rate after cutting tails off at base. Starved. Normal intact. Starved. ma =| 4 7) ra a=} | 4, a g "Bo hae iS | oo | @ 8 q Ss | - = i ee eet 2 Es ea lee g a each a g ° =| ~ é =) ° “4 Re a 3) S} ° > ) Bi) } ro Ba g B Za A < 4 oS A a < 4 ow +tails | | 10 20.679) 2.068 10 20.287, 2.029 Nov. 13 | —tails | | 17.552) 1.755 20| 10 17.095| 1.71 ©.045| 2.5 10 19.429) 1.943) 0.086 Area 27| 10 16.077| 1.608] 0.102] 6.1 10 17.605, 1.761; 0.182 9.8 Dec. 4] 10 TS ee elms 545 | OnOOR mean 10 16.615) 1.662) 0.099 Ba in|] TASB 27 |ei483\ 9 wOnKit2 nes 10 15.935) 1-594| 0.068 4-1 1g| 10 12-50 | L-350|| ©-.082) 558 10 15.542| 1.554| 0.04 Diaby | | } Gay 2 Bheteas ee ta) =| 210) a & 5 7) 8 B 5 5 | 7) 8 2 SU TS Sl ersia Seo ea Ee From From Nov. 13 to Dec. 18.... | 1.553] 0.404] 26. Nov. 13-Dec. 18] 1.791] 0.475 | 26.5 Ser F? Ser H? Same as Set F} Same as H} | Nov. 13] 10 +tails 19.886 1.989 10 22.445) 2.245 —tails | 16.872| 1.687 | 20] 10 16.664 1.666) 0.021 ge) 10 Piimctoxll eaic) ||| | (o)ativtls Gar? 27| tO 15-47 | 1.547] 0.019 ial 10 | 19.76 | 1.976) 0.154 | 7-5 Wee: 4\) 410 15.092) 1.509] 0.038 2.4 8 Co) 19.075| 1.908) 0.068 | 3.5 II| 10 14.177| 1.418] 0.091 6.2 10 U7 575 |) Lagi5Si 0 cOnns 8.1 18) 10 13.162] 1.316] 0.102 7-4 10 16.495, 1.65 | 0.108 | 6.3 “eo = “o = Pepe en ie ve eS iee keg | oR = ae =I From | From Nov. 13 to Dec. mer 1.501] 0.371 24.6 Nov. 13-Dec. 18 1.947 ©2595) || 3025 50 Ser F3 Same as F1 Eee ie a Os] e eo 6 3 - } J vo Ze eed Nov. 13} 10 | + tails 1762s 7152 — tails |S) 20| 10 14.815] 1.482 27| 10 13.825] 1.382) Dec. °41) 910 13.287] 1.329] II] 10 12.397| 1.24 18] 10 11.64 | 1.164 ~ vo oS From Nov. 13 to Dec. 18....| 1.357 Ada Springer TABLE XIV—Continued Loss 0.068 o.1 0.053 0.089 0.076 Loss 0.386 Per cent loss No. of animals 10 From Ser H3 Same as H} Total weight 19.607 18.771 17-307 16.379 15-332 14.455 Nov. 13—Dec. 18 = ue] a = 8 © a o Eee = 1.961 1.877! 0.084 4.3 1. 734|) O-0AGn| more 1.638] 0.093 5-5 1.533] 0.105 6.6 1.446] 0.087 6. a Ss Sass 2 =I 26) 3 | eae 1.703} ©.515 | 30.2 Ser F? Regenerating tails. Rate of growth after starvation. Fed. 153 mg. of beef a week A Study of Growth TABLE XV Rebs 5 Mar. 5 ie 2 From Eee ; 3 2 & PA = 2 eres iris A fe) 13.162] 1.316 10 16.437| 1.644] 0.328 22 10 16.04 | 1.604] —o.o40] —2.4 10 16.507] 1.651 .O47 2.8 10 17.269] 1.727] 0.076 4-4 10 tect) || Moki) — Cait 6.1 6.5 10 20.935| 2.094 0.257 6.5 10 21.465] 2.147/ 0.053 2.5 10 23.035] 2.304] 0.157 Fc Io 23.095] 2.31 0.006 0.2 fe) 23.294] 2.329] 0.019 0.8 fe) 22.925] 2.293] —0.036] —1.5 Co) 23.095] 2.31 0.017 0.7 10 22.812] 2.281) —o.029] —1.2 10 Pere Wiils| ered] CC) 8h “oo o b=) Dec. 18 to Mar. 5 ....| 1.822] 1.013 55-5 5! Ser H? Normal intact control after starvation. Fed as in F? knee ae : SUE ge J = 3 E aS s FE oR 2 q ra (een Wie acres fe) 16.495] 1.65 10 19.222] 1.922| 0.272] 15.2 10 18.857| 1.886] —o.036| —1.8 fe) 19.632] 1.963] 0.077 4. io) 20.229] 2.023] 0.06 ao ie) 21.69 | 2.169} 0.146 6.9 10 \ 5-3 10 24.13 | 2.413 0.244] f ine fe) 24.89 | 2.489] 0.076 Qn2 fe) 25.885] 2.589] 0.1 3°9 10 Be eAN | |2 54-0. O40 |i Tie9 Ff 19.035} 2.719] 0.179 6.8 5 127-537) 22507 0-200) — Oak 5 12150227510 0,007) OL2 5 Tigeys || Moekts}| eis) = Fo/1 5 I1.595| 2.319] —0.049; —2. 5 11.925] 2.385} 0.066 2.8 5 12.44 | 2.48 0.095 3-9 From Dec. 18 toMar. 5) 2.184] 1.069 48.9 52 Ada Springer TABLE XVI Ser A Ser D Rate of growth at high temperature 28.2° (average). Fed Same as A. Starved ~ ) =| A a PI fet i) 2 “Eb ss 2 “Eo EO fe —. 7) a tae 7) ee "a ans E ‘Ss cn E o » - Siesta 2 & & A at 2 & a Ste ery ond Or ae 8 S| ray al ene & . (3) fo) ~ an a Be SN el MeKe A eee fle eS x Bil (Sex es Seal fi Feb. 21 6 |11.8 | 1.967 Mar. 5] 27.5 | 6 |10.057| 1.676] —o.291| —15.9 6 | 10.225] 1.704) Al poy || A |) ioe | teks Susi) SG) act 6 8.95 | 1.492) —o.212 14-2 19] 29. On e8eh 1-478) —OsOrg ata 6 7.825] 1.304, —0.188) 13.4 26| 27. 5 | 7-425] 1.485] 0.012 0.8 3 3-49 | 1.16 | —o.144| 117.6 iMees PA exo | Ly |) G/sleys || teztets| —ekeyydl —tyde 3 3-105) 1.035) —0.125| 11.3 9] 29. 5 | 7-095] 1.419] 0.011 0.7 3 2.905} 0.968 —o.067 6.6 rel] Ao || || Spout || ior) 0.011 0.7 3 2.797] 0.932|. —0.036 37 a8 a Be ae Sill ve o 8 D a v a v A = opiates eet From From Mar. 5 Feb. 21 to Apr. 16] 1.698|— .537| —31 COVAIprai Ovo) 906) On7772) SC DG Ser B Ser E At room temperature 22° (average). Fed Same as B. Starved Feb. 21 6 |13.08 | 2.18 Maar. 5] 21.5 | 6 |12.355] 2.059] —o.121 S07) 6 |10.33 1.721 12] 22.1 6 |11.327| 1.888) —o.171 —8.6 i eh? 1.64 0.081; —4.8 19] 22 6 |10.745| 1.791] —0.097| —5.2 5 7205 ||) .449)) —Onto7|erany 26| 22 6 |10.465| 1.744] —0.047| —2.6 5) Os64e0 e325) Orns ae ANDY eee) 2AE Sais al Oc 2oal WO S 7K Os Oo) eek. 5 | 6.04 | 1.208] —0.120| —9.4 9| 22 5 | 8.475| 1.695} 0.038 2.2 FN) Gere |) oie) =Oeege||) =~ 16] 22 5 | 8.647| 1.729] 0.034 1.9 i We Relio. "| sie) |) SCvenG)) ite “oO re “oO +“ Set acm tetae afl sar ee 2 ee ino ES =e 8 ep ae) =] A = S) = From From Mar. 5 Feb. 21 to Apr. 16 I.954|— .451|—23.08 toApr.16..| 1.425] —o.591| 41.4 A Study of Growth Ser C TABLE XVI—Continued At low temperature I1.2° C (average). Fed Av. temperature degrees C. No. of animals ~ ie) oo n Apr: 2] 12. ~ \o a] i] n AMmManninnnan From Feb. 21 to Apr. 16 Pale: a a = Ps 2 So Be | 4 10.712) 1.785 (72) | ensgs3 11.285) 1.881 11.585] 1.931 9-87 | 1.974 9.625] 1.925 10.435| 2.087 9.907) 1.981 vo al 1.883] 0.196 Increase increase 1.04 Per cent | Ww vb AN wo Set F Same as C. Starved 53 ole ean ae g "a oO ity ean ne - = E &% I ) = s s 8 3 fo) 5 3 3 wl Ey ie es sy Gea e1Ong25(eteom 3 Souls |) ei 0.001 0.05 2} 3.222] 1.611| —o.109 6.5 2 3.005] 1.502] —0.109 oe 2 2.94 | 1.47 | —0.032 2.1 2 2.93 | 1.46 | —o.0or 0.06 2 2.865] 1.432] —0.028 1.9 Ca =) eralkees Uex2 = 5 é eee} ~ onl A From Mar. 5 Apr t6>. |) 1576)! 0.289) —18ieg 54 Ada Springer TABLE XVII SerA SET Males. Showing normal rate of growth. Fed 315 mg. Females. Normal rate of growth. Fea of beef per week as in Set A a AL os. : = F = 3 ah o g g aah 5) Je 5 5 = i f= 3 = = Es 2 2 = E 2 g 5 6 a = S & x) Ta Ny ° = S E 5 g 5 6 E > g Ea Z & < Si a Z a < a Oct. 22] 10 21.235| 2.124 10 17.214| 1.721 28| 10 21.978] 2.198} 0.074 3-4 10 19.052| 1.905 | 0.184 | 10.1 Nov. 4]/ 10 22.874| 2.287} 0.089 3-9 10 19.148) 1.915 | 0.01 0.5 I1| 10 22.934] 2.293] 0.006 Oe) ie) 19.402) 1.94 0.025 ee 18] 10 23.721) 2.372| 0.079 aia 10 20.343| 2.034 | 0.094 4-7 25| 10 24.749] 2.475] 0.103 4.2 10 21.806] 2.18 0.146 6.9 Decy 32/10 25.085) 2.509) 0.034 Tie 10 22.708] 2.271 | 0.091 4. 9] 10 26.366) 2.637} 0.128 4.9 10 23-575| 2.358 | 0.087 Bei 16} 10 27.397) 2.740| 0.103 3.8 Io 24.394) 2.44 0.082 3-4 23] 10 28 .378| 2.838! 0.098 I85 10 25.815) 2.582 | 0.142 5.6 31; 10 28.645| 2.865} 0.027 0.9 10 26.387| 2.639 | 0.057 Ql | “> ete mos 3 2 From | | From Oct. 22 to Dec. 31 vee 2.494 0.741 | 29.7 Oct. 22 toDec. 31} 2.18 0.918 | 42.1 A Study of Growth 55 TABLE XVIII Males. Individuals showing normal rate of growth. No. 1 No. 2 an Aorireeee oI o 35 3 o £ Fy |e ees Pabee 2 Z © < Sh 3 8 8 6 oe 3 8 ; 30 °c 3 b ; ob ‘o 3 M Be locas (ren |e fe ey Soy are fi Oct 22 I 2027 I 2.787 28 I 315] 2.672} 0.045 1.6 1) ays 2755 |i OL O82 eta Nov. 4 I |.315| | 2-687] 0.015 0.5 I | 315 |q] 2-855} 0.1 Babs II Te |g es @ 2.720| 0.033 1.2 HR etal Ey 2-795) 0.00) | 2). o o 1S ee Tae eg 2 | 2.910 0.19 6.7 iy (2 2.92 0.125 4-3 25) ip oeinG || a) Se eheyVis| “tiie 1 Sire |g egs|| = ‘osrze! 5-4 DEC 2 ed 315 hy || Clr hitts|| Cots Tsu) 3-098] 0.003 0.09 Oy pate Aollg| — alts 3-9 I | 298 3.227| 0.129 4. 16) 1 | 316 3-446] 0.283 8.5 if || gyeke 3.468) 0.241 Tee Pell Tl eles 3-439| —0.007| —0.2 I | 329 AeKOG|| OTA) =) Bul) ae || eer 3-483) 0.044 fine Ton 320 3-412| 0.047 Teg ra Ps > Fe | oa aegis g 2) eg | dee irae: Solas eb) 6 | ae ee okey SS ela) ene From From Oct. 22 to Dec. 31 ....] 3.055] 0.856 | 28.01 Oct.22 to Dec.31| 3.099] 0.625) 20.1 ID)385 2 fo) IDA Zit gaeal) youl] lez ee) etal Dec. 2 to Dec:31)) 3.255) 0.314 9.6 No. 3 ‘No. 4 Oct. 22) 1 Sf I 2.509 25 ahs, 1.587| 0.01 0.6 Tenn es 2.472| —0.037| —I.4 Nov. 4] I | 315) @| 1-659] 0-072) 4-4 1 | 315| | 2.539] 0.067 2.6 ary] 9G) eens a 1.744| 0.085 4-9 Ty Shs = 2.537; —0.002) —0.07 o rice] = ath nits z 1.799| 0.055 aya ESE Si 2.592/ 0.055 2p acl a 3x5) | ¥.997| 0-498] ‘To.4 Eur sl -2.7ae le ora Siam ee Decwe2 I 315 1.948] —0.049| —2.4 i |Lesiiuts 2.81 08 2.8 9 297 2.09 0.142 ‘lo 1 | 303 3-044 234 7-9 16 I 325 2.197 0.107 4.9 I 324 2.998! —0.046) —1.5 Z| ak |] abt 2.23n\| 102138 6. 1 ee NN sleexa 128] get 31 I | 329 2.339} 0.004 O.1 ton |eQ2z | 3-287, 0.161 Ge From From Oct, 22) towDWeengr-. |) 1-958), 04702) 38!-9 Oct.22 toDec. 31} 2.898) 0.778) 26.8 Dec. 2 to Dec. 31....| 2.143] 0.391 18.2 Dec. 2 to Dec.31 | 3.048} 0.477) 15.6 Ada Springer TABLE XVIII—Continued No. 5 “a ay ee g ce ae Achaea ML Sauiienc soul hae 8 roy a0 ‘Oo 3} be Zo Ne B | 4 a Oct. 22 I 1.885 28 I 315 E.927| 0.042 Z.2 Nov 4! 1 | 315] g] 1-959] 0.032 1.6 adh ae |] Gaia a 2.035] 0.076 3.8 LS | nest - 2.052] 0.017 °.8 Ae |) Sei|| | anil Gaeta 2.8 Dee: 2 I 315 2.222) 0.11 5. 9} 1 | 286 2.301} 0.079 3-4 ne} te || Get 2.438] 0.137 Say Bali si | gpl Doll || (oe 183} 31 ir |l) sya Za52 ||) 0.08 2 a &h 2 q 2 From Oct. 22 to Dec. 31 | 2.202] 0.635 | 28.8 Dec. 2 to Dec. 31 | 2.371| 0.298 12.5 A Study of Growth ah TABLE XIx Females Individuals showing normal rate of growth. No. 1 No. 2 =| 2 S = o ae. Om site oo He Z ie sen eels 2 © 2 ml! e S % ° a |) 8 8 Do ‘o 3) i ; 8p ‘o 3 = ie tee aang f Peake eal ea huge f Oct 22) 1 | 1.927| I 1237 Deh] ae || EROS 2.142] 0.215] 10.5 ih hae ies icity! Ooze] Belek Nov. 4 1 | 315] o| 2.122) —0.020) —0.9 Dy | Sega 1-487 0.04) | ee rag II I 315 a oii i7h| A} (9}y| Cle I 315 1.497, 0.01 0.6 rit ae | 305 | 2.194, 0.077] 3-5 E (oars - 1.637| 0.14 8.9 25) 1% | 315] | 2-452) 0.258) 411.1 P| 3x5) 57) 1.844! 0.207) “a21-8 Dee, 2) ae | ai 2.447, —0.005| —0.2 iv | i 1.92 0.076 4. 9 || key 2.649] 0.202/ 7.9 I 318 2e72\NOZER 123 16 |i ete [ea eine) SOCylh| Hee) I | 300 2.09 | —0.083 3.8 23 I 328 | 2.691 OstrS\) -4ua I 330 2.288] 0.198 9. 31 Te eaes 2-762| 0.071) 2.6 Teeegaa: 2.424| 0.136 ng) 28, g% ae a & a ee SC os os S a s oOo sx ele le 6 Se ae: See ee SOP ae pects From From Oct. 22 to) Deeygn .. =.) 2.344) (0-835) 315.6 Oct. 22 toDec.31| 1.83 1.187| 64.8 Dec. 2 to Dec. 31 ....| 2.604 0.315] 12.09 Dec. 2 toDec. 31] 2.172| 0.504) 23.2 No.3 No. 4 Oct 22 I D222) I 1.387 | Ae}| an) epils DEAR? |e Or20 ils ae Or. I 315 1.462) 0.075 Ree Nov. 4| I | 315] o| 2-357 —0.075| 5) oil I 315| g| 1-689) 0 227 eae! mil | ae leit I 25204. —0.073) —2) 5 I 315 a 1.807) 0.118 6.7 iiss) ae Sl] adil = 2.367) 0.083] 3.5 I 315 2 1.797| —0.O1 =)a'5 7s) ik |) HI | Daeiiall” = eroyds|| a(t I BUG || eG Al ORNS 5 8.2 Dec ez eet) e305 AciGl) aie) Geli Hy egag 1.987] 0.035] 1169) 9) ete a7 2.698 eiia| Fle I 314 2.186, 0.199 9-5 TO] # rgir 2.637 —0.061 — 22 I 321 | 2.217, 0.031) 1.4 2A etn 3277 2.696 059] 2.2 I 332 2.339, 0.122 Gaal atl) VT 322 soreyy| to)tel il | Gols I 323 2.418] 0.079] Big “ae | a5 “> = s 3 5 } 5 s 5 5 e age i Pcie =|) Semiber e From | From Octa22zstoDeciain ae 2.507) Om5 iL 02227) Oct. 22 toDec. 31) 1.902} 1.031) 54.2 DecyzstowDecagte.)-.) ZOGA NO). 27.5|TOnd. Dec. 2 toDec. 31] 2.202 19.5 0.431) Ada Springer TABLE XIX—Continued No. 5 ogee g eo = S ze “ g I Toho ars =» | § & : ao ‘oO 5 Me Zea ee ean cee 3 aes Oct. 22 I I 447, 28 |" ap | ang 1542), o.egs) 16,3 Nov. 4; 1 | 315] g] 1-552} 0.01 0.6 nid) Sh || Sys a I.507| —0.045| —2.9 re}] ak |) ahs z 1.567, 0.06 3-9 2u “Fo NQtsl || T7249 Owe7i= Jong IDs, All af eels 1 .63))|) 0.044 2G GIy ah |leayee 1.943} 0.263] 14.5 mS ae |ghies 1.999} 0.056) 2.8 | ZA at eS 2.051| 0.052) 2.5 il) 36 |) eye) 2.065] 0.014; 6.8 ae 2 ay o f o Guns il From Oct. 22 to Dec. 31 | 1.756] 0.618) 35.1 Dec.2 to! Dec. 31 |) 1-872) Vo.g8s5| 520.75 TABLE XX Set C Normal growth. Fed 315 mg. of beef per week a ~ =, a E 4s | ‘3s g ‘3 “3 5 = < 2 g 2 | ey Wee BL ea g fe} i) ala = 5 Z ial =< | os Ay Nov. 11 7 6.371] 0.910 18 7 7.021] 1.003} 0.093 9-7 215 |e 7-929) 1.133| 0.13 120 Deca 2 v7, 7.945| 1.135] 0.002 o.1 Hy 8.534| 1.219] 0.084 Wiad 16] 7 8.821] 1.26 | 0.041 23 23 16 8.92 1.486 0.226 | 16.4 ye hea ae oe vo a 2 oC © From Noy. 11 to Dec. 23) 1.198] 0.576 | 48.08 A Study of Growth TABLE XXI Ser D Growth at low temperature (10° C. average). Fed as much as animals would eat i) oe Ss) ere ‘s 8 Bui eee eee E i S ee o 2 F So % Fi to} om & = CG © oO : £2 £ o rs 2) oe iS) > a 5 Z a a < a iw Oct. 22] 10 L722 T7722 28] 10 10 20.731| 2.073 0.351; 18.5 Nov. 4] 10 12 21.676} 2.168 0.095, 4.4 II| 10 13 21.766) 2.177 0.009] 0.4 18] 10 II 2174.0!) 20117.50 ||) —O)O2|) O09 25} 10 10 MebeGeiil Pole || ——Io)\seysh| — Ski Dees 2\) 10 10 Minos Acid || —CiCyiy/| shor 9| 10 22.45 | 2.245 0.069} 3.1 16) 10 12 22.041| 2.204 | —o.041| —1.8 23] 10 II 23.061; 2.306 ©.102] 4.5 31; 10 12 22 LOK 22 —O.096| 42 “o oe + 8 © 5 g eel g | gs From Oct. 22 to Dec. 31 1.966] 0.488 | 24.8 4 No. 1 (Male) No. 2 (Male) is) oes g | oe g aol ince Bs 5 Beccles a te uw fs r=) & q BS rr RS r g a ) o 85 S a = ) reel erse Py o :: . - a pa . s - — MH 3 LORS, o g 5 2 LORS g S a) A = = 4 py Z = = 4 Py Oct. 22 I 1.374 I 1.947 28) 1 B15 em et734ie cOngOn is 23h I 315 | 2.202) O.255) ize Nove 4) 2 Bu aint) anit The, I gui || up|) Teh — Od II I HLO) |) 1-527) —O. 047) —2).15, I 315 | 2.302] 0.11 4.8 18 I LOS | olin 747/\sO- 00) | 4d I 110 | 2.457 155 6.5 25 3 On t-7.97/ OOS 2.8 I 105 | 2.532; 0.075 a IDisten pl) TOS ay eek Of |i O 27) Ni —7759 I nels |) Aertel) Chet) Goi 9} 1 © | 1.697| 0.027) 1.6 105 105 | 2.436] 0.03 1.2 16} 1 © | 1.605} —0.092] 5.5 I © | 2.284| —o.152| —6.4 20 110 | 1.8 195} 11.4 I 105 | 2.422} 0.138 5.8 sy) a 105 | 1.688) —o.112} 6.4 I 105 | 2.415] —0.007 aa 2 | ae beg ee 2 ge) 2 | 83 ie: : 5 I From | From Oct. 22 to Dec. 31 | 1.531] 06.314] 20.5 Oct. 22 toDec. 31) 2.181' 0.468 60 Ada Springer TABLE XXI—Continued No. 3 (Male) ig o> = pis ee eu coy aes We ro) So 5 “Bo © 8 S oo % a) 5 u Ziv geen. Wee 5 a ‘Octa 22 I 2.232 28 I 315 | 2.607 0.375 Se Nov. 4) 1 315 | 2.89 0.183) 6.6 16) ad B05) 2637) —0.2.53)5— Our 18 I TOS 2 0G] e—- On OF eae 25nd Oo | 2.567| —o-040) —1.5 Dec 2 ex Ge | 2505) 0.042) 6 Gi) 105 | 2.538} 0.023] 0.9 10) /n © | 2.493) —0.045| —1.7 272) like) || Bplsicel| @aiies Ang ii @ ana, | —o.183| Tee o c et) 2a) be ene From Oct. 22 to Dec. 31 | 2.331] 0.198} 8.4 A Study of Growth TABLE XXII Ser D! Growth at room temperature (20° C. average). Fed same as Set D Hn (peer) cae 8 Be tee Weenie 5 ae ees S g S) a) ° > a o Z a a g E Bie | g eile test ll tee | ace . ieee an alee - ee cial a | By | Seo Hite : é oe o o FF 6 oe wi) is) ACM WM a bap 4 ow Bal eray eral es Miei ies ow Oct. 22 I 1.297 I 2.007| 28 I ang || E.592|| O-295] 20.4 I 315 | 2.134) 0.127 Nov. 4] 1 Bice tt. 070 Nem -OS7 mas 23 I 215 || 2.162)) 0.028 riail| ai Ais, |) Taliry/| “Oacits| = sec I 305) | 2107 —O-onk 18 I 315 1.777; 0.08 4.6 I 105 | 2.092) —0.015 25 I 105 | 1.864; 0©0.087/ 4.7 I 105 | 1.982) —o.110 Dec: 2 | HOS || Seo || ==@stekv ll, 2.533 I 1.916 —0.066 Oe | TOS ele O47 ee Ont27 in 6.7 I 1.822} —0.094 nO] a | OW ete 522i On 24 merit I © | 1.737, —0.085 23 I 110 | 1.944] 0.121 6.4 I IIO | 1.777| 0.04 20) et TOS L892 | OLO52 27 I On| 1-7) J —ofo77) 22) eo Lae eee From From | Och27to Dec: 40...) 1594/8 C5595] 3723 Oct. 22 toDec. 31) 1.853] 0.307 62 Ada Springer TABLE XXIII Ser D? Growth at high temperature (30° C. average). Fed same as Set D. isha iets Roe Hock See Scr ca 5 2 be B < = = Oct. 22] 10 15.849) 1.585 28] 10 35 15.714| 1.571] —o.014| —o.8 Nov. 4] 10 29 15.477| 1.548] —0.023] —1.4 itil ni) 28 15.537| 1-554] 0.006] 0.3 18] 10 28 15-2 87/\00529|) 0-025 |e ato 25| 10 29 15.245| 1.525| —0.004] —0.2 Decse2| 9 26 12.982] 1.442] —0.083] —5.5 9) 9 28 13.427| 1.492] 0.05 3-4 16} 9 29 13.078] 1.453] —0.039|.—2.6 23) a9 31 12.78 | 1.420] —0.033] —2.2 31, 9 2p Teg RIS) To ayall) =Ceys]| = 55 AM Phe From Oct. 22 to Dec.31 1.478| 0.214] 14.4 No. 1 (Male) No. 2 (Female) ees s Eig ei Sa 2 = Stas ae lie 2 Smee ounceyg| Peat all me | eee geal ae 3 oo 5 oO 3) 4 6 bo | 7d 3 Zoe, peal a 5 oe ra Ware = 4 Oct: 22) = 1.037 I 1.389 28 is) 2047) ORor 0.9 I 315 | 1.484) 0.095 Nov. 4] 1 ails | tisha) Ciel 2} 7 I 315 | 1.432] —0.052 ri eee IIO | 1.044) 0.03 2.9 I 315 | 1.462 03 18} I TOS eUsOle) |e OnO22 art I 105 | 1.547; 0.085 25) © | 0.967) —0.045| —4.5 I 105 | 1.637 09 Dec 2) or 105 | 0.947) —0.02 | —2. I 105 | 1.548) —0.089 OW os © Or962)i) 5O-OUG | alls I Ios | 1.589) 0.041 rc: o | 0.88 | —0.082) —8.9 I ° 2a) ar IIo | 0.918] 0.038) 4.2 I 105 | 1.479) —0.110 Bul) 105 |*O.QI5| —O.003) —0-3 I TOG) lh. 423'| OOK Fee eae Pela From | From Oct; 22\to| Dec. ater nc 0.976, 0.122] 12.5 Oct. 22 toDec. 13} 1.406] 0.034 Per cent inc. Per cent Ww orn increase A Study of Growth TABLE XXIII—Continued No. 3 (Male) F I 3) = | ON r=] oO) leee || fee Ziyi,| eae ane eel es Octar22 1 1.677 28 relly Lepes nel Cy! || 73571 Nov. 4) 1 315 | 1.517| —o.12 | —7.6 pi] B05 1-537 O02 Ting 1S) 110 | 1.524) —o.013} —o.8 25 I O | 1.442] —0.082) —5.5 Decan2 et OF | 1-987) —O.O5h |e —on8 Gy © | 1.408) 0.021 1.5 LOl I © | 1.425/ 0.007] 0.4 23} ee IIO | 1.419] —0.006] —0.4 atl r I Hegyey) eyrmnipl) Ber) co 2 a x a ee Du ee 2! A = From Oct. 22 to Dec. 31 | 1.49 Wee) Hise Ada Springer TABLE XXIV Set E} Growth at 20° C. (average). Fed as much as animals would eat 2 a Boe Oulece Clee 3 yl Sees dee: 2 = Stl Ege Mee & g 2 ees A veces : 2 |e Be poe Gee re Oct. 22] 10 17 -377| 1.738 28| 10 20 22.126] 2.213] 0.475] 24. Nov. 4] 10 20 21.756) 2.176] —0.037| —1.6 LTO 20 22.356, 2.236] 0.06 2.7 18} 10 19 23.899] 2.39 0.154] 6.6 25} 10 20 25.364) 2.536] 0.146) 5.9 Dec ez 10 20 25.89 | 2.589] 0.053] 2. g| 10 21 26.637 2.664] 0.075) 2.8 16] 10 20 27OT TN 2e7qol) AO:O74l) e277, 23] 10 20 28.04 | 2.804) 0.066) 2.3 hii] 0) 20 26.05 | 2.89 0.086] 3. From Oct. 23 to Dec. 31 | DeAlA) Ti UG? AOE, No. 1 (Male) No. 2 (Male) Bias| Suites g Bl Seal 2 ‘g |S & e E |2 & ve Be LG ra a S oe Neca 2 g = Fal exe lites fh] Legioe | ac | reine aes ne ae ° cons g 2 o S Sa a, 3 S Fee a es 4 oe Z |e 2 4 ow Octe22 ar 2.307 I 1.527 25a a 630 | 2.407 | 0.1 4.2 I 630 | 1.854 0.327| 19.3 Nov. 4] 1 630 | 2.442 | 0.035 1.4 I 6308 2017, 0.163; 8.4 | 525 | 2.494 | 0.052 21 I 525 || 2.047 0.03 1.4 18} Reem 02/7 mnOraag 5.1 I RR aay 0.21 9-7 25) GIS ze Ohe sl O.228 8.3 I 420 | 2.424 ©. 167| e7ien Dee, Ba 420 | 2.911 | 0.056 1.9 I 420 | 2.476 ©.052) 2.0 9) x 525) ge Cog) || 1OLOg2 3-1 I 420 | 2.53 0.054| 2.1 1G |e ot AE engi O28 || O-03 0.9 I 110 | 2.649 ©-119]) = 455 2a 1 AUS |) 3kO8) || C.075, 2.4 I Bis | AoE) || = @aelis)| 2.6 31 I ality | shout) || Coley) Hels I ails |) Delt 0.048 1.8 From From Oct) 22 towDecign 27820 O15 50n | ated Oct. 22—Dec. 31] 2.077 I.101] 53.05 A Study of Growth TABLE XXIV—Continued No. 3 (Male) Cae g 3 = 2 Sie alk gi 2 ra 5 ons a os o sé [esl s 5 5 Zee = 4 4 Och 22 I 1.507 28; 1 (Yo || Baar 0.735} 39-2 Nov. 4|) 630 | 2.307 0.065} 2.8 II I Rok || aang. || =@piiG) Bou S| ee 525 | 2.297 0.105, 4.6 25 eel 525) 2687, 0.39 | 15.6 Dec) 2) 1 420) ||" 21510) | -—- O07) — 0.8 Ol) a Bulls || Balsre) 0.012} 0.4 16) 5x 105 | 2.635 0.113] 4.3 23a 315 || 2.437 | Slay) Fok) ail] (died) “o o = 3) uw From Oct. 22—Dec. 31 197/25 tORQ2On nazar 66 Ada Springer TABLE XXV Set E? Growth at 30°C. (average). Fed as in E} | Pei Se ek : a a wa oa = siae|2 | 8) € 13 : = 2 = i 5 s 2 | 6 So) a) ne f Oct. 22] 10 16.899] 1.69 28} 10 35 iipfouly || woWG| @oerG|P ie! Nov. 4] 10 29 17.661] 1.766} 0.051 : I1| 10 28 18.227] 1.823] ©.057| -3,1 18} 9 28 18.669] 2.074) 0.251) 12.8 25) 9 29 19.967| 2.218] 0.144, 6.7 Dec 2 | eG 26 20.002] 2.222| 0.004) 0.1 g| . 8 28 18.864] 2.358] 0.136) 5.9 16] 8 29 19.406] 2.426] 0.068) 2.8 23| 8 31 19.49 | 2.436] 0.01 0.4 31]; 8 32 18.120] 2.265] —0.171| 7.2 OM a = ix} GC} is] o a s o 5 tae pee ame From Oct. 22 to Dec. 21 1.977] 0.575] 29.07 No. 1 (Male) No. 2 (Male) g si = i) Shee =) 5 3 2 aS a 3 al es uw 2 g | 3S eis 2 g r=] Pace Geel eel S cceal Seo 5 a Ss a is) b S) oo a) i) ba Pee Ae le x alee =| 8 a Oct. 22) 1 2.805 | 1.539| 28 I 420 | 2.627| —o.178; —6.5 I 420 I 582 0.043} 2.7 Nov. 4 I 840) 1927.57) 0205) 049 I 840 1.582 0.000 Fully ca 525 | 2.687] o.110) 4.1 I 525 | 1.677| 0.095) 5.8 18] 630 | 2.93 0.243} 8.6 I 630 | 1.862) 0.185) 10.4 215 Lay || eyCiGl| GeuGSl) Gol I ATG) L887) 9 O.O2 5) masked Deca I 420 | 3.003] —0.082| —2.6 I 420 | 2.05 On163) esi) at Ceiie h OnG)| Cerna “opsy I 420) || 2507 So oe 16|) eet S15) ese 32 7\ee O.002)) a 2ar I uy || way Oe) 74055 23 I QS |eg e200) O-O20 eee I 315 Lay || =e 007, 0.3 31 I anid | ever) =). y) Ree I 215) ||) 9371) 0023) Male = © Z H & < 4 m4 Oct 220 16.606] 1.661 28) 10 35 | 17-449] 1-745] 0.084] 4.9 Novy. .4| 10 29 18.324] 1.832] 0.087) 4.8 Gi 8G 28 17.679] 1.964] 0.132] 6.9 18} 8 28 16.564] 2.071 ©. 107) 523 25, 8 29 17.092] 2.137| 0.066) 3.1 Deco 29 18 26 17.919] 2.239] 0.102) 4.6 Ones 28 18.751] 2.344] 0.105) 4.5 TOl ey 29 16.265] 2.323] —0.021| —o.9 23) 0 7) 31 T5706 2.2.52 On Ot earl 20 7 32 14-962! 2.137)| —Onhis| — 5.2 ies al) reali aae aes | From Oct. 22 to Dec. 31 | 1.899 | © 4761 25 .06| No.1 No. 2 Ely > | a —E |y > E: & z- Sal me a) | >=} 5 rs) 80) = » 2 EN PRS mss at ae ah sa cm ee Niue: é 3 co Ss 3) (3) | 5 3} co os Oo 3) by} esi face a a 2 \S |e eee Oct.» 22 I 1.107 I 1.632 28) 1 420 | 1.417 0.310] 24.5 I 420 | 1.627 | —0.005) —0.3 Nov. 4] I 52s, || )L.489))|| 0-020) 129 I 525 kia, OL12Z0/ yor II I 525 Les 12 Cyimgl| Biov I 420 | 2.612 0.865) 39.6 18} 1 Qs) |) 1677 0.165| 10.3 I 420 | 2.735 GO) eenAcO Zire me fly || toy hy) Os E1063 I 525 | 2.855 0.120] 4.2 Dec a2 I 420 1.912 0.125 6.7 I 420 | 2.872 0.017; 0.5 O|auear py || Piteer) 0.090] 4.5 I 420 | 2.992 Onl2Oli— Ae TO eek 315 | 2.069 0.067| 3.2 I 205 |) 2.945 0.047 ek Dall i AiG || BCR || SOzCKKO) S18 7/ I ats || BEeG-s |), —Oer nl) gio 31 I ALi || Ayes! |] —=C)Clss|) = eha'y I 315 | 2.674) —o.180| 625 “o ) = a = r= From From Oct) 22 toy Deci ier eet a542 0.871} 56.4 Oct22—Decyaine2eni52 1.042] 48.3 STUDIES ON CHROMOSOMES IV THE “ACCESSORY” CHROMOSOME IN SYROMASTES AND PYRROCHORIS WITH A COMPARATIVE REVIEW OF THE TYPES OF SEXUAL DIFFERENCES OF THE CHROMOSOME GROUPS! BY EDMUND B. WILSON Wiru Two Pirates anp Two Ficures In THE TEXT Since the unpaired idiochromosome (‘‘accessory chromosome’’) was first discovered by Henking (’g1) in Pyrrochoris apterus L. this species has been reéxamined by only one observer, Dr. J. Gross (07), with results that are in substantial agreement with those that pe had reached in an earlier investigation (’04) on the coreid species Syromastes marginatus L. In both cases his conclusions hre in conflict with the view advanced by McClung (’o02), and first 1 Terminology. With the advance of our knowledge of the chromosomes that form the distinctive differential between the chromosome groups of the two sexes, and between the male producing and the female producing spermatozoa, it becomes increasingly difficult to find a common name that will apply equally to their various modifications. ‘Terms such as the “‘accessory,” ‘“‘odd” or ‘‘heterotropic” chro- mosome, or ‘‘monosome,”’ that are based on the condition of these chromosomes in the male only, are. misleading or inappropriate; and some of them are in certain cases inapplicable, even in the male— e. g., in Syromastes, where the ‘‘accessory” chromosome is not univalent but bivalent. Such terms as “‘heterochromosome” or “‘allosome’”’ (Montgomery) seem to me unsatisfactory, since they designate the m-chromosomes as well as the differential chromosomes, though these are obviously of quite different nature. Since it has now become evident that a univalent ‘‘accessory’’ chromosome in the male is exactly equivalent to what I have called the “‘large idiochromosome” in other forms, I think these chromosomes should be designated by the same name, and one that will apply equally to both sexes. While there are some objections to the word ‘‘idiochromosome” as a general term for this purpose I am not able to suggest a better one; and since it has already been thus employed by some writers, I shall use it hereafter in a broader sense than that in which I first proposed it, to designate the differential chromosomes in general, whether they are paired or unpaired in the male, and whether one or more pairs are present. A univalent or odd idiochromosome in the male will be called the unpaired idiochro- mosome (or simply the idiochromosome), while the word ‘“‘heterotropic” may perhaps conveniently be used as descriptive of its passage without division to one pole in one of the maturation divisions. In Syromastes, as will appear, the ‘‘accessory” or heterotropic chromosome represents a pair of idiochro- mosomes; while in Galgulus there are several pairs of these chromosomes. THE JouRNAL or ExprerRIMENTAL ZOOLOGY, VOL. VI, NO. I. We) Edmund B. Wilson shown to be correct in principle by the work of Stevens and my- self, that half the spermatozoa are male producing and half female producing. ‘This view rests on the following facts. When the male somatic chromosome groups contain an odd number, includ- ing an odd or unpaired idiochromosome (as in Anasa, Alydus, or Protenor) the female groups have one more chromosome, being duplicates of the male groups with the addition of another chro- mosome like the unpaired one of the male. When the male groups contain an even number, including a large and a small idiochro- mosome (as in Lygzeus, Coenus or Tenebrio) the female groups contain the same number, but include two large idiochromosomes in place of a large and a small one. In the first type half the spermatozoa receive the odd idiochromosome while half do not, the former accordingly containing one chromosome more than the latter. In the second type all the spermatozoa receive the same number of chromosomes, but half receive the large idiochromosome and half the small. It follows from these rela- tions that eggs fertilized by spermatozoa containing the odd chro- mosome, or its homologue the large idiochromosome, must pro- duce females, those fertilized by the other spermatozoa males. These cytological results, first reached by Stevens (’05) in Tene- brio (which has a pair of unequal idiochromosomes in the male) and myself (o5b, ’o5c, ’06) in Anasa, Protenor, Alydus and Harmostes (which have an unpaired idiochromosome in the male) and in Lygzus, Coenus, Podisus and Euschistus (which agree essentially with Tenebrio), have since been confirmed in.a con- siderable number of species and extended to several other orders of insects.? They have recently received indirectly a_ striking experimental confirmation in the important work of Correns (’07), which proves that in the dicecious flowering plant, Bryonia dioica, the pollen grains are likewise male determining and female deter- mining in equal numbers. Gross’s conclusion in the case of Syromastes and Pyrrochoris is opposed to all these results in that only one of the two forms of spermatozoa is supposed to be functional (those containing the ?See the tabular review in the sequel. Studies on Chromosomes 71 “accessory” chromosome) the others being regarded as in a certain sense comparable to polar bodies (as was also supposed by Wallace (05). This result was based mainly on the numerical relations, and especially on the belief that in both these forms the number of chromosomes is an even one and the same in both sexes—twenty- two in Syromastes, twenty-four in Pyrrochoris. Since the com- plete reduced number (eleven and twelve in the two respective cases) 1s present only in those spermatozoa that contain the “accessory”’ chromosome, Gross argues that this class alone can be concerned in fertilization, as follows: Syromastes....Egg 11 + spermatozoén 11 = 22 (d'or 92) Pyrrochoris ... Egg 12 + spermatozoén 12 = 24 (cor 2) whereas in Anasa or Protenor the relations are: Anasaeen see Egg 11 + spermatozoén 10 = 21 (6) Egg 11 + spermatozo6n 11 = 22(@) Protenor....Egg 7 + spermatozoén 6 = 13 (<) Egg 7 + spermatozoobn 7 = 14(?) In the hope of clearing up this perplexing contradiction I endeavored to procure material for a reéxamination of the two forms in question, and through the great kindness of Professor Boveri, to whom my best thanks are due, was fortunate enough to obtain an abundant supply of both, though unluckily it includes no female material.‘ As faras the relations can be worked out on the male alone they give, I believe, the solution of the puzzle and bring the two species in question into line with the general princi- ple that has been established for other forms. This is evidently true of Pyrrochoris. Syromastes, however, constitutes a new 3 At first thought this seems to be in harmony with the remarkable discovery of Meves (’03, ’07) that in the male honey bee actual polar bodies are formed which produce abortive spermatids. Butobviously the two cases are not parallel, for in the bee the fertilized eggs produce only females; and this finds a natural explanation, in accordance with the general conclusions of McClung, Stevens and myself, in the assumption that it is the male producing class that degenerate as polar bodies. ‘The material, fixed in Flemming’s fluid and in Bouin’s picro-acetic-formol mixture, is of excellent quality and gave preparations of perfect clearness. The Flemming material is on the whole the best. For single stains Zwaardemaker’s safranin and iron hematoxylin were employed (the latter especially for photographs). Various double stains were also used. One of the best, which I can strongly recom- mend to other workers in this field, is the combination of safranin and lichtgriin, which gives prepara- tions of admirable clearness and is also easy to use and certain in its results. 72 Edmund B. Wilson type that is not yet known to be exactly paralleled in other forms; though, as will appear, the genus Galgulus presents a somewhat analogous case. It does not seem to have occurred to Dr. Gross (as it did not to me until I had carefully studied both forms) that Syromastes and Pyrrochoris might be of different type, but such is evidently the case. I shall endeavor to show that Pyrrochoris is of quite orthodox type, having an odd somatic number in the male and a typical unpaired idiochromosome. Since I am compelled to differ with Dr. Gross in regard to this species, I am glad to admit that the doubts I formerly expressed as to his account of the spermatogonial number in Syromastes, were unfounded. In regard to the female number, on the other hand, I believe he was misled by a wrong theoretic expectation (as he evidently was in case of the male Pyrrochoris), though it is possible that his determination of the apparent number was also correct, as indicated beyond. SYROMASTES MARGINATUS L. Gross’s account of this form was as follows: The somatic groups in both sexes are stated to show twenty-two chromosomes. The “accessory” chromosome arises by the synapsis of two spermatogonial chromosomes, and is therefore a bivalent. It divides equationally in the first spermatocyte division but fails to divide in the second, passing bodily to one pole in advance of the other chromosomes without even entering the equatorial plate. All of the spermatid-nuclei thus receive ten chromosomes and half of them in addition the “accessory.” ‘These are the essen- tial conclusions; but they are complicated by the following singular view of the relations between the “accessory” and the micro- chromosomes or “mm-chromosomes.” ‘The chromosome nucleolus of the growth period is supposed not to give rise (as it does in Pyrrochoris and other forms) to the heterotropic or “accessory” chromosome of the spermatocyte divisions, but to the m-chromo- some bivalent—the same view as the earlier one of Paulmier (99) which has since been shown to be erroneous (Wilson ’o5c). But, on the other hand it is believed to arise, not from the Studies on Chromosomes 73 m-chromosomes of the spermatogonia, but from two larger chro- mosomes, while the spermatogonial m-chromosomes are supposed to be converted into the “‘accessory” (!). I will not enter upon the very ingenious, if somewhat fantastic, conclusions that are based on these results, for, as I shall attempt to show, the results them- selves cannot be sustained in some important particulars. But apart from this [ am glad to be able to give the most positive con- firmation of Gross’s interesting discovery in regard to the numer- ical relations in the male. Syromastes 1s indeed a case in which the spermatogonial number 1s an even one (twenty-two), while there is a heterotropic chromosome in the second division. Half the sperma- tozoa seem to receive ten chromosomes and half eleven, as in so many otherspecies of Coreidae. But as Gross also correctly de- scribed, the heterotropic chromosome is here a bivalent which represents two chromosomes united together. The true numbers characteristic of the two classes of spermatozoa are therefore ten and twelve, respectively. For the sake of clearness [ will here point out that this becomes at once intelligible under the assump- tion that the female number is not twenty-two, as Gross believed, but twenty-four; and such I believe will be found to be the fact. That Gross was mistaken—doubtless misled by the earlier conclusion of Paulmier (’9g), nm which he was at first followed by Montgomery (’o1)—in supposing that the chromosome nucleolus of the growth period divides to form the m-chromosomes, is | think thoroughly demonstrated by my preparations. In the case of Anasa and Alydus I showed (’05c) that the m-chromosomes are not formed in the way Paulmier believed, but arise from two small separate rod-like chromosomes that are in a diffused condition during the growth period and only condense to form compact bodies at the same time that the condensation of the larger chro- mosomes takes place. I have since found this to be true of many other species. It is confirmed in the case of Anasa by the smear preparations of Foot and Strobell (’07), and I have also since fully established the same conclusion by this method, by means of which every chromosome in the nucleus may be demonstrated.* 5 This is opposed to the conclusion of Montgomery (’06). 74 | Edmund B. Wilson Although I have no smear preparations of Syromastes it is perfectly clear from the sections that the facts are the same here as in Anasa Alydus, and other forms. In the early prophases of the first divi- sion (at a period corresponding to Gross’s Figs. 31 to 37) when the plasmosome has disappeared or is greatly reduced in size, the nuclei contain both the chromosome-nucleolus and the m-chro- mosomes. ‘This is shown in great numbers of cells with unmis- takable clearness and after various methods of staining, particu- larly after safranin alone or combined with lichtgriin. In the early part of this period the chromosome nucleolus is at once recognizable by its intense color and sharp contour and 1s not for a moment to be confused with a plasmosome. ‘The ordinary bivalents are still in the form of ragged pale bodies, having the form of longitudinally split rods or double crosses. ‘The m-chro- mosomes have the same texture and staining reaction, but are much smaller and never show the cross form. While it is difh- cult to show the facts to demonstration in photographs of sections they may be fairly well seen in the following. Photo 18 shows the chromosome nucleolus (not quite in focus,) one of the large biva- lents (two others barely appear) and both m-chromosomes. Photo 1g is a similar view (the m-chromosomes more condensed), while Photo 20 shows the m-chromosomes and three of the ordi- nary bivalents. ‘The succeeding changes must be rapidly passed through, since the successive steps are often seen in the same cyst, passing from one side to the other. In these stages the large bivalents rapidly condense and regain their staining capacity, finally assuming a bipartite or quadripartite form. The m-chro- mosomes undergo a similar condensation, being finally reduced to ovoidal or spheroidal bodies. The chromosome nucleolus, on the other hand, becomes somewhat looser in texture and assumes an asymmetrical quadripartite shape, in which form it enters the equatorial plate to form the eccentric “accessory” chromosome. The period at which the m-chromosomes condense varies consider- ably, and the same is true of their relative position; sometimes they are in contact, sometimes more or less widely separated, even lying on opposite sides of the nucleus. Photo 21 shows two nuclei, one above the other, in each of which appear both m-chromosomes, Studies on Chromosomes 75 the chromosome nucleolus and a number of the other bivalents. Photo 22 shows the same condition. Photo 23, from the same cyst, is slightly later, showing the two spheroidal m-chromosomes wide apart, the chromosome nucleolus, and several of the other chro- mosomes. (Ihe chromosomes nucleolus, perfectly recognizable in the preparation, is in the photograph hardly distinguishable from the other bivalents seen endwise.) Up to this point, which shortly precedes the dissolution of the nuclear membrane, the chromosome nucleolus is still immediately recognizable by its deeper color (especially after safranin). There follows a brief period in which this distinction disappears, but the chromosome nucleolus is still recognizable by its asymmetrical form. That it gives rise to the eccentric “accessory” 1s, I think, beyond doubt. The evidence is demonstrative that it does not divide to form the m-chromosomes, and that the latter arise from separate rods as described. Gross appears to have seen these rods at a much earlier period (cf. his Fig. 10) and correctly identifies them with the spermatogonial m-chromosomes; but he believed them to give rise to the “‘accessory.” The relation of the chromosome nucleolus to the spermato- gonial chromosomes cannot be determined in Syromastes with the same degree of certainty as in Pyrrochoris (as described beyond), but the size relations leave hardly a doubt that Gross was right in asserting its origin from two of the larger of these chromosomes. The study of these relations 1s of importance because I believe they justify the conclusion that the chromosome nucleolus, and hence the “accessory,” is nothing other than a pair of slightly unequal idiochromosomes, which can readily be recognized in the spermatogonial groups. Study of the spermatogonial groups in detail shows that twenty of the chromosomes may be equally paired, while the remaining two are slightly but distinctly unequal in size. “These can always be recognized as the smallest of the chromosomes except the m-chromosome. Photos 1 and 2 show two groups in which this clearly appears. These photographs are reproduced in Text Figs. ta, 1b, with two others, c and d, the chromosomes in question being designated as I and. 76 Edmund B. Wilson It is evidently this pair that give rise to the bivalent “accessory” (eccentric) chromosome of the first division and hence to the chromosome nucleolus of the growth period. Gross correctly describes this bivalent as a quadripartite body or tetrad, but overlooked the fact that it is composed of two slightly unequal halves, and these correspond 1 in relative size to the unequal pair in the spermatogonia. This appears unmistakably in a great number of polar views of the first division metaphase (though it is not always apparent) and is clearly shown in Photos 3, 4 and 5. It is evident that the bivalent is so placed in the equatorial Cc Fic. 1. Four spermatogonial chromosome-groups of Syromastes marginatus; a and b are reproduc- tions of Photos 1 and 2.* * The drawings are not made from the microscope with the camera lucida but directly upon enlarged photographs of the objects. Since I believe this method to be superior in accuracy for the representa- tion of such small objects I will briefly describe it inthe hope that others may find it useful. The original negatives are taken directly from the sections at an enlargement of 1500 diameters (2 mm. oil immersion, compensation ocular 6). From these negatives enlarged bromide prints are made (with a photographic camera) three times the size of the original negatives (1. e., 4500 diameters) upon double weight paper, which givesa good surface for pen drawings. The drawing is then made directly on the print with waterproof ink, and when thoroughly dry the remains of the photograph are bleached out in a mixture of sodium hyposulphite and potassium ferricyanide. The enlarged prints of course show the chromosomes with more or less blurred outlines (though they are clearer than might be supposed); but by working with an ordinary print and the object before one for comparison the drawings may nevertheless be made with great accuracy. They may be tested and if necessary corrected, by the use of a reducing glass. Studies on Chromosomes TT plate as to undergo an equation division, like the idiochromosomes of other Hemiptera heteroptera. In uniting to form a bivalent before the first division these chromosomes differ from those of most other Hemiptera, but in all other respects up to the end of the first division they correspond exactly with them. But even’ this difference is bridged by a condition occasionally seen in other forms, for instance in Lygzeus and Metapodius.* In the last named form the typical and usual condition is that the idiochro- mosomes are in the first division quite separate, lying eccentrically outside the principal ring of chromosomes like the unpaired idio- chromosomes of other coreids (Photos 6 and 7), and in this posi- tion they separately divide. Exceptionally, however, they lie in close contact (Photos 8 and g), forming an asymmetrical bivalent precisely like that of Syromastes. In both cases this bivalent divides equationally, giving two asymmetrical daughter-dyads, thus a i The exactness of the correspondence up to this point seems to leave no doubt of the homology of this pair of chromosomes in the two forms. In the second division, however, the two species show a remarkable contrast. In Metapodius, as in Lygzeus or Euschis- tus, the two idiochromosomes are always united to form an unsym- metrical bivalent which enters the equatorial plate and is separated into its two components, half the spermatids receiving the large one and half the small. In Syromastes, on the other hand, the idiochromosomes remain united and do not enter the equatorial plate at all, but pass directly to one pole where they are included in the daughter-nucleus, as Gross has described (Photos 11 to 17). Owing to this behavior of the idiochromosome bivalent, polar views of the second division always show but ten chromo- somes instead of eleven (Photo 10). In this case therefore half the spermatid nuclei receive two more chromosomes than the others, the two classes having respectively ten and twelve chromo- somes. As the idiochromosome bivalent passes to the pole its two components are usually closely united, and often cannot be 6 The latter remarkable genus, which presents the phenomenon of the ‘‘supernumerary chromosomes” (Wilson ’o7c), will form the subject of a forthcoming fifth ‘‘Study.” 78 Edmund B. Wilson distinguished; but in some cases they may still be seen, as in Photo 12. As the nuclear vacuole forms the ordinary chromosomes rapidly diminish in staining capacity, while the idiochromosome bivalent retains its compact form and dark color, like a nucleolus, ‘and thus comes conspicuously into view, particularly after safranin. Its double nature is at this time often more clearly apparent than in the preceding stages. It disappears from view some time after the reconstruction, at a much earlier period than in Pyrrochoris. Only exceptionally in my preparations do the chromosomes of the second division show a quadripartite form as Gross figures them. Their usual form is dumb-bell shaped or dyad-like; though as the two halves separate they are often connected by double fibers, as is the case with many other species of Hemiptera. PYRROCHORIS APTERUS L. As already stated, Pyrrochoris is of different type from Syro- mastes and agrees precisely with other forms having an unpaired idiochromosome, such as Anasa or Protenor. Aside from the interest that this species possess as the one in which Henking first discovered the idiochromosome, it is in other respects a peculiarly interesting form for the study of the general spermatogenesis, particularly in respect to the presynaptic and synaptic periods. I shall here, however, confine myself mainly to the numerical relations and the history of the idiochromosome. Henking him- self somewhat doubtfully concluded that the spermatogonial number was twenty-four: “Ich habe in drei Fallen die Zahl 24 erhalten, in einem Falle die Zahl 23. Da die Bilder tberall die gleichen sind, so habe ich das Zahlgeschaft nicht an einer grosseren Zahl vorgenommen und glaube die theoretisch zu erwartende Zahl 24 als das Normal ansehen diirfen”’ (op. cit., p. 688, italics mine). It is clear enough from this that Henking, too, was misled by a false theoretic expectation; and a study of his figures (op. cit., Figs. 6, a, b. c, 7) will show that they are very far from decisive. In the case of the female, Henking speaks much more positively (‘92) and there is hardly a doubt that his count of twenty-four chromosomes was correct, since he found this number “ unverkenn- Studies on Chromosomes 79 bar” in the dividing odgonia, and in the connective tissue cells of the ovary, and also figured (Fig. 39) a double group (exactly such as I described in Anasa, Wilson ’o6, Fig. 2,k), showing forty-eight chromosomes. Gross accepts Henking’s account without question, treating the numerical relations in rather summary fashion as follows: “Die Aequatorialplatte der sich teilenden Spermatogonien enthalt 24 Chromosomen. Daieselbe Zahl hat Henking ausser in den Sperma- togonien auch in den Oogonien gefunden. Ebenso konnte ich in den Follikelzellen der Eirdhren, also in somatischen Zellen, kon- statieren. 24 ist also die Normalzahl der Species” (07, p. 277). In support of this are given two polar views of spermatogonial metaphases (the female groups are not figured) each showing eight small and sixteen large chromosomes (Figs. 9 and 10). His ac- count continues as follows: The idiochromosome appears aready in the syanaptic period (synizesis) as a double nucleolus-like body, assumed to be a bivalent body that arises by the synapsis of two of the spermatogonial chromosomes, though none of the earlier stages were followed out. At a later period it appears as a single spheroidal body owing to the close apposition of its two halves. This chromosome divides in the first spermatocyte division, but in the second lags behind the others and passes undivided to one pole, as Henking described. All of the spermatid-nuclei thus receive eleven chromosomes, while half of them receive in addi- tion the idiochromosome. Since both sexes were supposed to con- tain twenty-four chromosomes, Gross drew the same conclusion as the one previously reached in the case of Syromastes, namely, that only the twelve-chromosome spermatozoa are functional. In regard to the spermatocyte divisions my own results are perfectly in accord with Henking’s and Gross’s. As to the spermatogonial number, I must say that after having immediately confirmed Gross’s account of Syromastes (which [ examined first) I was fully prepared to find a similar relation in Pyrrochoris. It was therefore with astonishment that I found everywhere twenty- three instead of twenty-four spermatogonial chromosomes. ‘This number appears with diagrammatic clearness in a great number of spermatogonia from different individuals (testes from 35 different 80 Edmund B. Wilson individuals have been sectioned) and is shown both in camera drawings and in photographs. Eight of the latter are shown (Photos 24 to 31), and these same groups are also represented in the drawings, Text Figs. 2, a, b, c, d, e, 7, k, 1, together with four others (7, g, h, 1), also from photographs. Inspection of these photographs and drawings will show that the unpaired idiochro- mosome is at once recognizable by its large size, which renders Fic. 1. Spermatogonial groups of Pyrrochoris apterus (drawn on photographic enlargements, as explained under Fig. 1); a, b, c, d, e, j, k and / are reproductions of Photos 24, 25, 26, 27, 28, 29, 30 and 31, respectively. it almost as conspicuous as in Protenor (heretofore described by Montgomery and myself). I find the size relations not quite the same as Gross describes them. ‘There are, as he states, eight chromosomes that are considerably smaller than the others; but two of the others are but slightly larger. The remaining twelve paired chromosomes are much larger, though the contrast is in my material not so great as Gross figures it. The idiochromo- Studies on Chromosomes SI some is nearly twice as large as any of the others, and is obviously unpaired. I have examined a large number of spermatogonial groups with great care with a view to the possibility that this chro- mosome might in reality be double, but am thoroughly convinced that such is not the case. This is unmistakably evident when this chromosome has the form of a straight or only slightly curved rod (Photos 24 to 28, Text Fig. 2, a to 7), and these constitute the great majority of observed cases. I have, however, found a few cases where it has a very marked sigmoid curvature; two or three of these give at first sight the appearance of two chromosomes in contact (Photos 29, 30, 31; Text Fig. 2, 7, k, 7). Even here close study shows that it is a single body; but such forms might readily mislead an observer having a preconceived idea of the number to be expected. That this is a single chromosome that is identical with the idio- chromosome of the growth period and the maturation divisions is placed beyond doubt by a study of the presynaptic stages, which were not examined by either Henking or Gross. This period is of such interest in Pyrrochoris as to merit a special study. With only a single exception I know of no other form in which the history of the idiochromosome and the succession of the stages can be so completely and readily followed at this time. Throughout this whole period, beginning with the telophases of the last sperma- togonial division, the idiochromosome can be traced step by step as a single body, and it is evidently identical with the large un- paired spermatogonial chromosome. In the stages that immediately follow the last spermatogonial telophase (Photos 32 and 33) the chromosomes still retain their boundaries, though they show a looser texture, vaguer outlines and diminished staining capacity (by which characters the post- phases are readily distinguishable from the prophases). ‘The large chromosome (idiochromosome) is clearly distinguishable at this time, both by its size and by its deeper color. In the stages that immediately follow a remarkable contrast appears between this chromosome and the others. ‘The latter rapidly lose their visible boundaries and their staining capacity, breaking up into a fine net- like structure in which traces of a spireme-like arrangement may 82 Edmund B. Wilson sometimes be seen. The idiochromosome, on the other hand, retains its identity and deep color and now appears as a conspicu- ous elongated body (“caterpillar stage’). “Though its outlines are still somewhat ragged and its color less intense than in the succeeding stages, it already appears in sharp contrast to the pale reticulum (Photos 34 and 35). It sometimes extends straight across the whole diameter of the nucleus; but beside such forms, in the same cysts, are often curved and shorter forms. At this time it is usually surrounded by a distinct clear space or vacuole, as I hope the photographs may show; and there are also in the nucleus from one to three much smaller nucleolus-like bodies which (on account of the staining reactions) I believe to be plas- mosomes, but these soon disappear. Splendid pictures of these and the following stages are given by the safranin-lichtgriin combination, which shows the idiochromosome at every stage bright red, while in properly differentiated preparations the reticu- lum is pure green.’ The idiochromosome now takes up a pe- ripheral position and the clear space surrounding it disappears. It acquires a more definite contour, stains still more intensely, and rapidly shortens until it is converted into a condensed ovoidal or spheroidal chromosome nucleolus that may be traced without a break through every stage up to the prophases of the first sperma- tocyte division. As it shortens it may undergo a variety of form changes. In what I regard as the typical process it shows no indication of duality at any period up to the full contraction phase (synizesis) being progressively reduced to a short rod and finally to an ovoidal or spheroidal body (Photos 36 to 42). In the mean- time the nuclear reticulum contracts more and more, usually towards one side of the nucleus, becomes coarser in texture, and increases in staining capacity, until at the climax of the process 7 The effect of this stain depends in some measure, of course, on the relative degree of extraction of the two dyes. My method is to stain in safranin for two to four hours and then to place the slide at once in strong alcoholic solution of lichtgriin for ten to twenty seconds. This is at once followed by_ rapid washing in 95 per cent and absolute alcohols. The alcohol is then replaced by clove oil and the latter by xylol. In all cases the chromosomes of dividing cells and the chromosome nucleolus of all stages appear brilliant red, the achromatic fibers and general cytoplasm pure green. ‘The relative intensity of red and green depend on the length of immersion in the green solution. The description here given applies to sections rather strongly stained in the green. Studies on Chromosomes 83 it forms a close knot, or rounded mass, staining almost black in hematoxylin, at one side of which is the idiochromosome (now a condensed chromosome nucleolus). These structures lie in a large clear nuclear vacuole, as shown in Photos 39 to 42. The stage thus attained is the characteristic contraction phase or synizesis, which in this species is extremely marked.°® In the safranin-lichtgriin preparation at this period the chro- mosome nucleolus is, as always, intensely red. The synaptic knot varies with the relative intensities of the red and green, being in some preparations distinctly red, in others pure green, in still others of mixedappearance. In the succeedingstage the chromatin emerges from the synaptic knot in the form of separate spireme threads which lose their staining capacity for hamatoxylin and in the double stain are again pure green (Photos 43 and 44). In the middle and late growth period they are still more or less green but contain red granules. In the prophases of the first divi- sion they at last lose Hei affinity for the green and finally appear pure red; but this does not occur until just before the dissolution of the nuclear membrane. Since the idiochromosome always retains its intense red color it may thus be followed from stage to stage with great certainty. The study of the whole cycle of changes from the last sper- matogonial division onward gives certain very definite results in regard to synapsis in general, and especially in regard to the idio- 8 Many recent writers have expressed the opinion that the synizesis stage is an artifact produced as a shrinkage product, though Miss Sargant (’96) stated very explicitly that she had seen it in the living cells, and this has recently been confirmed by Overton (’o5). I can fully substantiate this in the case of Anasa tristis. The perfectly fresh testis, gently teased apart in a Ringer’s fluid in which the sperma- tozoa continue their active movements, very clearly shows nearly all the features of the spermatogenesis, including the number, shape and size relations of the chromosomes, their characteristic grouping and behavior in the spermatocyte divisions, the double rods, crosses and other prophase figures, the spindle fibers and asters, and even, I believe, the centrosomes. In this fresh material the synizesis stage appears in essentially the same form as in the sections, the nuclear knot lying in a large clear vacuole. These nuclei only appear in the same region of the testis as in sections, and they show a conspicuous contrast to those of earlier and later stages that lie near bythem. In the post synaptic stages the chromo- somes, in the form of spireme threads can be seen again spreading through the nuclear cavity. These observations leave no doubt in my mind that the synizesis is a normal phase of the spermatogenesis in these animals, though it is not improbable that the contraction may be somewhat exaggerated by the reagents. It is evident, however, from such studies as those of the Schreiners (’06) and others that the synizesis does not occur in some forms. 84 Edmund B. Wilson chromosome. Concerning the first point I will here only indicate one principal conclusion. It is quite clear that in Pyrrochoris (and I think the same holds true in other Hemiptera) synapsis, or the conjugation of chromosomes two by two, does not occur in the closing anaphases of the last spermatogonial division as was described by Montgomery (’oo) in Peripatus and Euschistus (‘Pentatoma’’), by Sutton (02) in Brachystola, by Stevens (’03) in Sagitta, and by Dublin (05) in Pedicellina. Although the number of chromosomes in the postphases immediately following this division (Photos 32 and 33) cannot be exactly made out, it is perfectly evident that it is not the reduced number but approxi- mates to the somatic number (twenty-three). “The chromosomes, therefore, have not paired two by two in the spermatogonial ana- phases. It is equally certain that this stage does not pass directly into the synizesis but is separated from it by a long “resting period” (Photos 34 to 38)—as is demonstated by the topographical relations as wellas by the progressive stages of the idiochromo- some—in which the ordinary chromosomes lose their sharp boundaries and their affinity for nuclear stains. In this respect Pyrrochoris shows a close similarity to ‘Tomopteris, as described by the Schreiners (’06), whose original preparations, by the kindness of Dr. Schreiner, I have Bad opportunity to examine. This comparison has convinced me that synapsis occurs at the same period in both—whether by parasynapsis (side to side union) or telosynapsis (end to end union®) or in some other way I am not prepared to say. There can be no manner of doubt that the first division of the bivalents is a transverse one, as described by Paulmier and Montgomery; but it has been rendered evident enough by recent studies on reduction that this in itself gives no trustworthy evidence regarding the mode of synapsis. The direct investigation of the process in the Hemiptera presents great difficulties. The foregoing general conclusion regarding the time of synapsis is of importance for the more specific one in regard to the idio- chromosome. During the entire earlier presynaptic period the 9T have for some years made use of these terms in my lectures on cytology. Studies on Chromosomes 85 elongated idiochromosome is manifestly a single body. As it shortens and condenses to form the chromosome nucleolus, it shows a considerable variety of forms; and the rate of condensation also varies, cells that are already entering the synizesis stage being sometimes seen in which the idiochromosome is still distinctly a rod (Photos 35 and 36). In most cases it is at this time a single ovoidal or spheroidal body; but not infrequently it appears more or less distinctly double (Photos 37 to 39). This condition is however not produced by a previous synapsis of two chromosomes, as Gross believed, but arises, [ think, from a tendency of the chromatin to accumulate towards the ends of the rod; and when this is very marked it may assume an appearance of duality, even in the ear- lier stages (Photo 37, below), though this 1s relatively rare. In the later stages a double appearance is not infrequent, dumb-bell forms being thus produced, which sometimes give in the synizesis stage apparently double bodies. ‘The earlier stages conclusively show that this is a secondary appearance. In the later (postsyn- aptic) stages (Photos 34 and 35), and throughout the growth period, it always appears as a single spheroidal body. In view of these facts I think the conclusion inevitable that the chromosome nucleolus is a univalent chromosome that arises by the condensa- tion of the unpaired large chromosome of the spermatogonia. I have little to add to Henking’s and Gross’s acounts of the maturation divisions. As will be seen from Photos 45, 46, 50, 51, the size relations are correlated with those of the spermatogonial chromosomes. In polar views of the first division appear with great constancy four smallest chromosomes, one slightly larger one, and seven still larger ones, or twelve in all. The idiochromo- some is one of the largest, but cannot be distinguished from the others (as is also the case in Protenor). ‘This is obviously due to the fact that the idiochromosome is still a univalent or single chromosome, while each of the others represents two of the sper- matogonial chromosomes united. Since all have nearly the same dumb-bell shape as seen in side view, the idiochromosome appears from the pole approximately but half as large, relative to the others, as in the spermatogonia. ‘The same size-relations appear in the second division, but all the chromosomes are much smaller, as the photographs clearly show. 86 Edmund B. Wilson I have not succeeded with Pyrrochoris (as I have with several other genera) in obtaining photographs of both anaphase daughter groups showing all the chromosomes; but it is perfectly evident that all divide equally in the first division, and all but the idio- chromosome in the second. This chromosome lags behind the others and then passes undivided to one pole where it is included in the daughter nucleus (Photos 47 to 49) as Henking described. This pole thus receives twelve chromosomes, the other but eleven. As in a number of other species the idiochromosome retains its compact form and deep-staining capacity long after the reconstruc- tion of the nuclei and the breaking up of the other chromosomes. It may thus be distinguished (especially well in safranin prepara- tions) up to a rather late period stage of the spermatids, even after the tails have grown out. It finally disappears from view, and the mature spermatozoa show no visible indication of their dimorphism. GENERAL If my conclusions are correct, Pyrrochoris agrees exactly with other forms in which an unpaired idiochromosome is present. Syromastes however presents a new type in which the “accessory” chromosome is not univalent but bivalent, and in which accord- ingly half the spermatozoa receive two more chromosomes than the other half. If we may apply the same rule to Syromastes as that which holds for other Hemiptera we may expect the sperma- tozoa that receive the “accessory” to be female-producing, the others male-producing. ‘The fertilization formulas for the two species considered in this paper should therefore be as follows: PYRROCHORIS Egg 12 + spermatozoén 11 = zyote 23 (<') Egg 12 + spermatozoién 12 = zygote24(Q) SYROMASTES Egg 12 (including J and i)* + spermatozoén 10 = zygote 22 (including I and 7)(@) Egg 12 (including Jandi) + spermatozoén 12 (including J and) = zygote 24 (including Iie US oy) (C2), * The formation of a reduced female group of this composition may readily be explained if it be sup- posed that in synapsis the two small idiochromosomes couple with each other to form the bivalent ii, the two large ones to form the bivalent J/. Studies on Chromosomes 87 The correctness of my deduction may readily be tested by a reéxamination of the female groups. Gross, it is true, states that he has found but twenty-two chromosomes in the female (follicle cells); but I think no one is likely to consider as in any way conclusive the single figure that he gives in support of this (op. cit., Fig. 111). Not less than five of the twenty-two chromo- somes figured are deeply constricted; and any one of these might in reality be two chromosomes in contact) I hope that Dr. Gross himself may be willing to reéxamine this point, in view of the possibility here suggested. It is however also possible that the two members of each of the idiochromosome pairs in the female may be united to form a bivalent, in which case the female would apparently show but twenty-two chromosomes; but even if this be so the two members must separate again when transferred to the male. In regard to Pyrrochoris, there is little doubt that the determina- tion of the female number by Henking and Gross as twenty-four was correct; and since the idiochromosome in the male is the largest of the chromosomes we may expect the female groups to show two such chromosomes. !° I should state the expectation less confidently in the case of Syro- mastes if it stood entirely alone; but another case has now been made known in which the male and female groups differ by more than one chromosome. ‘This occurs in the genus Galgulus, which has been worked out in my laboratory by Mr. F. Payne (whose results are now in press)" on material collected by myself. ‘The following facts are very clearly shown in this form. ‘The sperma- togonial number is thirty-five, the female number thirty-eight. In the second division five of the chromosomes are always asso- 1° Henking’s figures (’92) give considerable evidence that such is really the case. His Fig. 83 of the first polar metaphase shows one of the twelve bivalents fully twice the size of the others; and the same is true of Fig. 68, which shows a side view of the second polar spindle, though not all the chromo- somes are shown. With this accords his Fig. 39 of a double group from a connective tissue cell of the female showing forty-eight chromosomes, of which four, of nearly equal size, are nearly twice the size of the others. This agrees precisely with the relation shown in a double group of Anasa figured by mein a former paper (’06, Fig. 2,k) which shows twice the normal number of both the largest and the smallest chromosomes. 1 Since published in Biol. Bull., xiv, 5. 88 Edmund B. Wilson ciated to form a definite pentad element of which four pass to one pole, one to the other, while the remaining fifteen chromosomes divide equally. Half the spermatozoa thus receive sixteen chro- mosomes and half nineteen. From these facts it is clear that the sixteen-chromosome class must be male producing, the nineteen- chromosome class female producing, according to the formula: GALGULUS Egg 7 + spermatozoin % — 3 = zygote n — 3 (d) Egg g =F spermatozoén 7 = zygoten (2) This case, together with that of Syromastes (if my inference regard- ing this form be correct) shows that we must considerably enlarge our previous conceptions as to the relations between sex produc- tion and the chromosomes; for we can no longer hold that only a single pair are involved. In Syromastes there are two such pairs, in Galgulus several pairs. A COMPARATIVE REVIEW OF THE TYPES OF SEXUAL DIFFERENCES OF THE CHROMOSOMES It is evident that a greater variety of types exists in regard to the sex differences than was indicated in the brief general review given in the third of my “Studies” (Wilson ’o6.) In that paper I distinguished three types, examples of which are given by Protenor, Lygzeus and Nezara; but the number must now be increased to at least five, and possibly to seven, of which I will now give a brief synopsis. With the exception of Syromastes and Diabrotica this synopsis includes only species of which both sexes have been accurately determined. Forms like the aphids, in which idiochromosomes have not yet been positively identified, have been omitted. Seventeen of the species are here reported for the first time (one or both sexes) from my own results hitherto unpublished. I am indebted to Dr. Stevens for permission to include her results on the Diptera and on Diabrotica, which are now in press (’o8a, ’o8b). Studies on Chromosomes 89 I Both sexes with the same number of chromosomes; a pair of equal idiochro- somes present in both. No visible difference between the two classes of sperma- tozoa or between the male and female somatic groups. FERTILIZATION FORMULA Egg 7 + spermatozoon ™ = zygoten(c‘or 2) Described Case 3 Ss Z a | 2 3 Specie Order | Family < 5 Authority ae | = Nezara hilaris Say Hemiptera heteroptera | Pentatomide | 14 | 14 | Wilson (’06) | | To this type belongs also Oncopeltus fasciatus Dall, one of the Lygaidx. It is further probable that here belong many forms in which no visible sexual differ- ences are to be seen, and in which idiochromosomes have not been identified. If a particular pair of chromosomes, corresponding to idiochromosomes, are of general occurrence, though not visibly distinguishable from the others, it is probable that this type represents the most frequent condition in animals generally. Mt Both sexes, and both classes of spermatozoa, with the same number of chromo- somes. The male with a pair of unequal idiochromosomes, half the spermatozoa receiving the large one and half the small. In the female a pair of equal idiochromo- somes like the large one of the male. FERTILIZATION FORMULA Egg ” (including J) + spermatozoon ® (including 7) = zygote n (including I7) & Egg ™ (including J) + spermatozoin" (including J) = zygote n (including IT) 9 go Edmund B. Wilson Described Cases a liee a ee Species Order Family E E Authority A | ® ive) Or Oebalus pugnax Fab. | Hemiptera heteroptera | Pentatomide 1o | 10 | Wilson Euschistus fissilis Uhl. Hemiptera heteroptera | Pentatomide 14 | 14 | Wilson’osb, osc, ’06 ictericus L. Hemiptera heteroptera | Pentatomide 14 | 14 | Wilson’o6 servus Say Hemiptera heteroptera | Pentatomide 14 | 14 | Wilson by tristigmus Say Hemiptera heteroptera | Pentatomide 14 | 14 { Ree | ‘i variolarius P. B. Hemiptera heteroptera | Pentatomide 14| 14 | Wilson’o6 Coenus delius Say Hemiptera heteroptera | Pentatomide 14| 14 | Wilson’osb,’o5c, ’06 Stiretrus anchorago Fab| Hemiptera heteroptera | Pentatomide 14 | 14 | Wilson Podisus SN eee oar \ Hemiptera heteroptera | Pentatomide 16 | 16 [3 ser eC oe (spinosus) f \ 2 Wilson ’osb, osc, 06 Banasa dimidiata Say Hemiptera heteroptera | Pentatomide 16 | 16 | Wilson’o7b calva Say Hemiptera heteroptera | Pentatomide 26*| 26 | Wilson’o7b Lygeus turcicus Fab. Hemiptera heteroptera | Pentatomide 14 | 14 | Wilson’osb, ’o5c, ’06 bicrucis Say Hemiptera heteroptera | Pentatomide 14 | 14 | Wilson Tenebrio molitor | Coleoptera Tenebrionide | 20 | 20 | Stevens’o5 Chelymorpha argus Coleoptera Chrysomelidz | 22 | 22 | Stevens’o6 Trirhabda virgata Coleoptera Chrysomelide | 28 | 28 | Stevens ’o6 canadense Coleoptera Chrysomelide | 30] 30 | Stevens ’06 Drosophila ampelophila} Diptera 8 8 | Stevens ’o8a Musca domestica Diptera 12 | 12 | Stevens’o8a Calliphora vomitoria Diptera 12 | 12 | Stevens’o8a Sarcophaga sarracinie | Diptera 12 | 12 | Stevens ’o8a Scatophaga pallida Diptera 12 | 12 | Stevens ’o8a Tetanocera sparsa Diptera 12 | 12 | Stevens’o8a Eristalis tenax Diptera 12 | 12 | Stevens ’o8a *See Type Ila. III . The female chromosome.groups with one more chromosome than the male. Male with an unpaired idiochromosome and an odd spermatogonial number, half the spermatozoa receiving the idiochromosome and half being without it. Female with an equal pair of idiochromosomes like the unpaired one of the male. Egg Egg NSws FERTILIZATION FORMULA (including J) + spermatozoén 7" — 1 = zygote n — 1 (including) & (including J) + spermatozoén " (including J) = zygote n (including IT) ? Studies on Chromosomes Described Cases gi Species Largus cinctus H. S. succinctus L. Pyrrochoris apterus L. Alydus pilosulus H. S. Harmostes reflexulus Std. f Protenor belfragei Hag. Leptocoris trivittatus Say Archimerus calcarator Fab. Pachylis gigas Burm. Anasa tristis DeG. armigera Say sp. Euthoctha galeator Fab. Leptoglossus phyllopus | L. Margus inconspicuus H. S. Chariesterus antennator Fab. Corynocoris distinctus Dall. Aprophora quadrang- ularis Peeciloptera septentrionalis pruinosa Elater, sp. Blatta germanica Anax junius Order Hemiptera heteroptera Hemiptera heteroptera Hemiptera heteroptera Hemiptera heteroptera Hemiptera heteroptera Hemiptera heteroptera Hemiptera heteroptera Hemiptera heteroptera Hemiptera heteroptera Hemiptera heteroptera Hemiptera heteroptera Hemiptera heteroptera Hemiptera heteroptera Hemiptera heteroptera | Hemiptera heteroptera Hemiptera heteroptera Hemiptera heteroptera Hemiptera homoptera Hemiptera homoptera Hemiptera homoptera Coleoptera Orthoptera Odonata Family Pyrrochoride Pyrrochoride Pyrrochoride Coreide Coreide Coreidz Coreide Coreide Coreide Coreide Coreide | Coreidz | Coreide Coreide Coreide Coreide | Coreide Jasside | Fulgoride | Fulgoride Elateride | | Blattide | Aeschinde o Somatic No. 21 S Z | g Authority fo} A OF 12 | Wilson 14 | Wilson Fs yi 2 Henking ’91 2 i 3 Wilson 14 | Wilson ’osb, ’osc, ’06 | f & Montgomery ’o1 4 )\ 9 Wilson ’06 7 3 Montgomery ’or 4 \ Q Wilson ’o5b, ’o5c, 706 14 | Wilson 16 | Wilson 16 | Wilson 22 | Wilson ’osb, ’o5c, ’06,’07a if o Montgomery ’06 *? )\ 9 Wilson 22 | Montgomery ’o06 22 | Wilson 22 | Wilson 24 | Wilson 26 | Wilson 26 | Wilson 24 | Stevens ’06 28 | Boring ’o7 28 | Boring ’o7 20 | Stevens ’06 | if oc Stevens ’06 74 i 2 Wassilieff ’07 28 | LeFevre and McGill ’08 *This number, disputed by Foot and Strobell (’o7a, b), has since been confirmed by my own reéxami- nation (’o7a, ’08) and by that of Lefevre and McGill (?o8) and others. Q2 Edmund B. Wilson IV Female groups (by inference only) with two more chromosomes than the male. In the male a pair of unequal idiochromosomes, half the spermatozoa receiving both these chromosomes, and hence two more than the other half. In the female (by inference only) two such pairs. FERTILIZATION FORMULA Egg ® (including I,7) + spermatozoén® — 2 = zygote n — 2 (including I,7) 3 Egg ™ (including I,7) + spermatozoén ™ including/,;) = zygote n (including I, I, 7, i,) Q (by inference only) Described Case | | gis Specie Order | Family a | 3 Authority | | BS "| OF | | , = OHesves | Hemiptera heteroptera | Coreide 22 2a Gos oe marginatus L | | | 2 Wilson (inferred) V Female groups with three more chromosomes than the male. Half the sperma- tozoa receiving three more chromosomes than the other half. Egg “ Gis spermatozo6n — 3 = zygote n — 3 (SA) Egg % + spermatozoén ” = zygote n (2) Described Case yl GS so | a | | 12/8 Specie | Order Family gl Authority | | 3a | Fo | oO | Galgulus oculatus Fab.) Hemiptera heteroptera | Galgulide | 35 | 38 | Payne ’o8 | | At least two of the foregoing types may be complicated by the presence of certain additional chromosomes, present in some individuals but not in others of the same species, to which I have applied the name of ‘“‘supernumerary chromosomes.’”” The number of these varies from one to six in different individuals but is constant in the same individual. In some forms (Metapodius, Banasa) these supernumer- 1 Wilson ’o7b, ’07c. A detailed description is now in preparation. Studies on Chromosomes 93 aries accompany a typical pair of unequal idiochromosomes (as in Type II). In other forms (Diabrotica), the supernumeraries accompany an unpaired idiochromo- some (as in Type III). In these cases definite numerical formulas cannot be given, since the distribution of the supernumeraries is variable and both sexes show a variable number of chromosomes in consequence (directly known only in Meta- podius.) Forthe present these cases may most conveniently be treated as sub-types as follows: Ila Forms that agree with Type II except that certain individuals may possess, in addition to a pair of idiochromosomes, one or several supernumerary chromosomes. The cases described, with the numbers of chromosomes observed, are as follows: ow) <2 Species Order Family Somatic | Somatic | Authority No. | No Banasa calva Hemiptera Pentatomide 26 | 26 Wilson ’07b heteroptera [26 + 1] Metapodius terminalis Hemiptera Coreide 21* 22 _ Wilson ’o7b, 08 heteroptera 22 22+1 22+1 Za 22-2 22+3 | 22+3 | 22+4 | femoratus Hemiptera Coreide 22 22+1 | Wilson’o7b, ’o8 heteroptera 22+2 | 22+2 22 Ae 22+4 | 22+6 granulosus Hemiptera Coreidz 22 | 22443 heteroptera Ae || Aap Wilson ’o07b, ’08 22+2 2243 | 2244 | | e225 * This number occurs only in Montgomery’s (’06) material of this species, identification of which though probably correct, is not absolutely certain. This case will be considered in a later publication. Illa Forms that agree with Type III except that certain individuals may possess, in addition to an unpaired idiochromosome, one or several supernumerary chromo- somes. Described cases as follows: 94 Edmund B. Wilson ey 2 Specie Order Family Somatic | Somatic Authority No. No. Diabrotica 12-punctata | Coleoptera Chrysomelidz 19 Stevens ’07, 708 soror I9+I 19+2 19+3 19+4 Despite the apparent diversity of the types that have been enu- merated all conform to the common principle that the spermatozoa are of two classes, equal in number, that are respectively male producing and female producing. In the case of Type I this is no more than an inference, since the two classes cannot be distin- guished by the eye; but its great probability will be admitted in the fact that the forms with equal idiochromosomes are connected by forms (such as Mineus) in which only a slight inequality exists, with those in which the inequality is very marked (Wilson ’o5a). The facts now show that the difference between the two classes of spermatozoa is not always confined to a single pair of chromosomes, but may affect two pairs (Syromastes) or even alarger number (Galgulus). It is noteworthy that in every case where a quantitative difference of chromatin exists between the sexes It is always in favor of the female, whether it appear in a larger number of chromosomes or in the greater size of one of them. But I must again emphasize the fact that this quantitative differ- ence cannot be considered as the primary factor that differenti- ates the two classes, for inthe first class such a difference does not exist, while in Metapodius, even in the same species, it is some- 13 I based this type on the facts observed in Nezara, where the idiochromosomes are equal in size in both sexes. This is not in accordance with the later observations of Montgomery (’06) who believes that in the Hemiptera generally the two components (paternal and maternal) of every chromosome pair are at least slightly unequal—though he finds the idiochromosomes of Oncopeltus equal as I have also since observed. A reéxamination of Nezara confirms my original account of this form, though in some individuals the idiochromosomes often appear very slightly unequal. A careful examination of the other chromosomes, particularly the small m-chromosomes (which are most favorable for the purpose) in Alydus, Anasa, Archimerus, Pachylis, and other genera, leads me to a very skeptical view of Montgomery’s general conclusion on this point. It is true that the two members of each pair vary slightly in relative size, and are not always exactly equal; but, in my material at least, it is clear that Studies on Chromosomes 95 times the female, sometimes the male, that has the larger number and quantity. I therefore adhere to the view that if the primary and essential difference between the two classes of spermatozoa inhere in the chromosomes (there is of course room for difference of opinion on this point) it must be, or originally have been, qualitative in nature. Since the appearance of my third “Study,” in which some general discussion of the sex chromosomes was offered, there has appeared an important paper by Correns (’07) on the higher plants, the results of which, as he points out, harmonize remarkably with those based on the cytological evidence. ‘The most important of his results is the experimental proof obtained by hybridizing experiments on Bryonia, that in the dicecious species the pollen grains are male producing and female producing in equal num- bers, quite in accordance with the view put forward by McClung (02) in regard to the spermatozoa of insects and proved to be correct in principle by the work of Stevens and myself. ‘That the same result should appear from investigations carried out on such different material and by such different methods certainly gives good ground for the belief that as far as the male is concerned the phenomenon is at least a very general one. Professor Correns points out in some detail the extraordinarily close parallel between his experimental results and the cytological ones of Stevens and myself; but the interpretation that he offers differs materially from both those that I suggested in an analysis of my observa- tions (Wilson ’06). According to my first interpretation (Castle’s) both sexes are assumed to be sex hybrids or heterozygotes. ‘The conclusion of Correns is that, in respect to the active sexual ten- dencies of the gametes that produce them, only the male is a sex hybrid or heterozygote (* (2)), while the female is a homozy- gote (@ 2). This interpretation explains the numerical equality this is merely a casual fluctuation, the general rule being equality. This variation appears in dif- ferent cells of the same cyst (as may be seen with especial clearness in the m-chromosomes in side views of the second division where errors due to foreshortening may be eliminated). It would be indeed strange if these relations were subject to no variation whatever. Tt is necessary to an understanding of Correns’s view to bear in mind that the gametes are not considered to be ‘‘pure” in the original Mendelian sense, but to bear both sexual possibilities, one of which is ‘‘active,” the other ‘‘latent.”’ 96 Edmund B. Wilson of the sexes in accordance with the Mendelian principle without the necessity for assuming selective fertilization. It is so simple, and seems to be so clearly demonstrated in the case of Bryonia, that its application to the interpretation of sex production in general is very tempting. Correns himself believes it “very prob- able”’ that his conclusion will apply to all the dicecious flowering plants, and possible that it may also hold true of animals (op. cit., pp- 65, 66). It is evident that in their superficial aspects the cyto- logical results seem to bear this out. Wherever the sexes show visible differences in the somatic chromosome groups the female groups consist of two series in duplicate, while the male groups show two series that are not duplicates, only one of them being identical with one of the female series. As far as the chromosomes are concerned, and from a purely morphological point of view, the female is therefore in fact a homozygote, the male a hetero- zygote, in these animals. But when more closely scrutinized from this standpoint the interpretation seems by no means so clear. As I showed in my third “Study” the odd chromosome of the male must be derived from the egg; and if this chromosome bears the sexual tendency, it must under Correns’s hypothesis carry the female tendency—which is a reductio ad absurdum, since It is not accompanied by a male-bearing mate or partner in the male. I think this brings clearly into view the following alter- native. Either the females of these insects must be physiolog- ically heterozygotes (as I assumed), or the so-called “sex chromo- somes” (idiochromosomes) do not bear the sexual tendencies but only accompany them in a definite way. Which of these possi- bilities is the true one may be left to further research to decide. I will only point out that Professor Correns carefully considers the difficulties that his interpretation encounters in some other direc- tions, and admits that it must be modified in certain cases—for example in the honey bee and in Dinophilus, in which latter case he too is compelled to admit the possibility of selective fertiliza- tion. The parthenogenetic females of such forms as the aphids and phylloxerans, which produce both males and females without fertilization, are still considered by Correns as homozygotes, the production of males being assumed to be determined, if I under- Studies on Chromosomes 97 stand his conception, by the activation of the “latent’’ (not to be confused with the “recessive”) male possibility in the male pro- ducing eggs. ‘This is doubtless an admissible assumption, though it seems to me to put a considerable strain upon the general hypothesis. ‘The more natural view would seem to be the one directly suggested by the facts, 1.e., that the parthenogenetic stem- mother aphid is a heterozygote, the male tendency being in the condition of a Mendelian recessive. But I will not enter upon a discussion of this question, which is now in a condition where a little observation and experiment will outweigh a large amount of hypothesis. I think, however, that the first of the interpretations that I suggested (following Castle) should not be rejected without further data, and especially not until the question of selective fertilization has been put to the test of direct experiment. ZoOlogical Laboratory Columbia University February 13, 1908 WORKS REFERRED TO Borinc, Atice M. ’07—A Study of the Spermatogenesis of twenty-two Species of the Membracidz, Jassidz, Cercopide and Fulgoride. Journ. Exp. Zool., Iv, 4: Correns, C.’07—Die Bestimmung und Vererbung des Geschlechtes. Berlin, 1907. Also (in abbreviated form) in Arch. f. Rassen- u. Ges.- Biologie, iv, 6. Dusuin, L. I. ’05—The History of the Germ-cells in Pedicellina Americana. Ann. Ne YeAcad:-Scr. xvisl: Foor, K., and Stropett, E. C. ’07a—The “‘ Accessory Chromosome” of Anasa tristis. Biol. Bull., xii. ’o7b—A Study of Chromosomes in the Spermatogenesis of Anasa tristis. Am. Journ. Anat., vii, 2. Gross, J. ’04—Die Spermatogenese von Syromastes marginatus. Zool. Jahrb., Anat. u. Ontog., xx. ; ’o7—Die Spermatogenese von Pyrrochoris apterus. Jbid., xxiii. Henxinc, H. ’91—Ueber Spermatogenese und deren Beziehung zur Eientwick- lung bei Pyrrochoris apterus. Zeitschr. f. Wiss. Zool., li. *92— Untersuchungen iiber die ersten Entwicklungs-vorgange in den Eiern der Insekten, Ill. bid, liv, 1. 98 Edmund B. Wilson LeFevre, G., and McGu11, C. ’08—The Chromosomes of Anasa tristis and Anax. junius. Am. Journ. Anat., viii, 4. Meves, F. ’03—Ueber “ Richtungskérperbildung” im Hoden von Hymenopteren. Anat. Anz., xxiv. °o7—Die Spermatocytenteilungen bei der Honigbiene, etc. Arch. mik. Anat.; xx, 3. Montcom_ry, T. H. ’00o—The Spermatogenesis of Peripatus, etc. Zool. Jahrb., Anat. u. Ontog., xiv. ’o1—A Study of the Chromosomes of the Germ-cells of Metazoa. Trans. Am. Phil. Soc., xx. ’04—Some Observations and Considerations upon the Maturation Phe- nomena of the Germ-cells. Biol. Bull., vi, 3. ’06—Chromosomes in the Spermatogenesis of the Hemiptera Heteroptera. , Trans\Am. Ebi Soci; xx. McCune, C. E. ’o2—The Acessory Chromosome—Sex Determinant? Biol. Bull. ii. Overton, J. B. ’05—Ueber Reduktionsteilung in den Pollenmutterzellen einiger Dikotylen. Jahrb. wiss. Bot., xlii, 1. Pautmigr, F. C. ’99—The Spermatogenesis of Anasa tristis. Journ. Morph., Supplement. Payne, F. ’08—On the Sexual Differences of the Chromosome groups in Galgulus oculatus. Biol. Bull., xiv, 5. SarGANT, ETHEL ’96—The Formation of the Sexual Nuclei in Lilium martagon. I, Odgenesis. Ann. Bot., x. SCHREINER, K. E. and A. ’06—Neue Studien iiber die Chromatinreifung der Geschlechtszellen (Tomopteris). Arch. Biol., xx. STEveENS, N. M. ’03—On the Ovogenesis and Spermatogenesis of Sagitta bipunctata. Zool. Jahrb., Anat. u. Ontog., xviii. ’o5—Studies in Spermatogenesis with Especial Reference to the “ Acces- sory Chromosome.’ Carnegie Institution, Washington, Pub. no. 36. ’06—Studies in Spermatogenesis. II. A Comparative Study of the Hetero- chromosomes in certain Species of Coleoptera, Hemiptera and Lepidoptera, with Especial Reference to Sex Determination. Toed.,, Pab. 36, EL: ‘o8a—A Study of the Germ-cells of Certain Diptera, etc., Journ. Exp. Loose v5.2. °o8b—The Chromosomes in Diabrotica, etc. Ibid., v, 4. Sutton, W. S. ’°02—On the Morphology of the Chromosome group in Brachystola magna. Biol. Bull., iv, 1. Studies on Chromosomes 99 Wattace, L. B. ’05—The Spermatogenesis of the Spider. Biol. Bull., viii. WassiuierFF, A. ’07—Die Spermatogenese von Blatta germanica. Arch. mik. Anat., Ixx. Witson, E. B. ’o5a—Studies on Chromosomes. I. The Behavior of the Idiochromosomes in Hemiptera. Journ. Exp. Zodl., ii. °o5b—The Chromosomes in Relation to the Determination of Sex in Insects. Science, xx, p. 500. ’o5c—Studies on Chromosomes. II. The paired Microchromosomes, Idiochromosomes and Heterotropic Chromosomes in the Hemip- teran, Jbids ii. ’06—Studies on Chromosomes. III. The Sexual Differences of the Chromosome Groups in Hemiptera, with some Considerations on the Determination and Inheritance of Sex. Jbid., iii. o7a—The Case of Anasa tristis. Science, xxv, p. IgI. ’o7b—Note on the Chromosome groups of Metapodius and Banasa. Biol. Bull. o7c—The Supernumerary Chromosomes of Metapodius. Read before the May Meeting of the N. Y. Acad. of Sci. Science, xxvi, 677. ’08—The Accessory Chromosome of Anasa tristis. Read before the Am. Soc. of Zodlogists, December, ’07. Science, xxvii, 690. EXPLANATION OF PLATES All of the figures are reproduced directly from photographs by the author, without retouching. The originals were taken with a Spencer 3; oil-immersion, Zeiss ocular 6, which gives an enlargement of 1500 diameters. The admirable method of focusing devised by Foot and Strobell was employed. They are reproduced at the same magnification. Pirate I (Photos 1 to §, 10 to 23, Syromastes marginatus; 6 to 10, Metapodius terminalis; 24 and 25, Pyrro- choris apterus). 1and2. Spermatogonial groups of Syromastes; copied in Text-fig. 1, a, b. 3 to 5. Polar views, first maturation metaphase; m-chromosome at the center, idiochromosome- bivalent (‘‘accessory’” chromosome) outside the ring at the left. 6 and 7. Corresponding views of Metapodius, typical condition with the two separate idiochromo- somes outside the ring at the left. 8 and 9. The same; exceptional condition, with the idiochromosomes (at the left) in contact. 10. Polar metaphase, second division, Syromastes. 11 to 17. Side views of the same division. The duality of the idiochromosome appears in 12, 16 and 17. 18 to 23. Early prophases of first maturation division, Syromastes. Each of these shows the separate m-chromosomes, and in all but No. 20 the chromosome nucleolus (idiochromosome bivalent) also appears. 24 and 25. Spermatogonial metaphases of Pyrrochoris (copied in Text-figs. 2, a, b). STUDIES ON CHROMOSOMES PLATE 1 Edmond B. Wilson. The Journal of Experimental Zoology, Vol. VI, No 1. WILSON, PHOTO. Prate II Pyrrochoris apterus 26 to 31. Spermatogonial groups, each showing twenty-three chromosomes, including the large unpaired idiochromosome; 30, 31 illustrate the rare case in which the latter appears double, owing to marked sigmoid curvature. These photos are copied in Text-figs. 2, c, d, e, j, k and /, respectively. 32 and 33. Post-phases shortly following last spermatogonial division; the chromosomes still distinct, idiochromosome recognizable by its large size and deeper color. 34. and 35. Presynaptic stages following the last, showing “‘caterpillar” stage of idiochromosome and small nucleoli. In the last two the shortening has begun. 36 to 38. Further condensation of the idiochromosome; initial stages of synizesis; apparent duality of the idiochromosome in two of the cells. 39 to 42. Synizesis, showing various forms of the chromosome nucleolus. 43 and 44. Early post-synaptic stages. 45 and 46. Polar metaphases, first spermatocyte division. 47 to 49. Side views of second division. 50 and 51. Polar metaphases, second division. STUDIES ON CHROMOSOMES PLATE 11 Edmond RB. Wilson. The Journal of Experimental Zoology, Vol. VI, No 1. WILSON, PHOTO. FURTHER STUDIES ON THE CHROMOSOMES OF THE COLEOPTERA BY N. M. STEVENS With Four Pirates In three previous papers (’05, ’06, ’08) the chromosomes of several species of Coleoptera have been described and figured, and the role of the heterochromosomes in sex determination dis- cussed. The following pages are a further contribution to our knowledge of the character and behavior of the heterochromo- somes, and of the methods of synapsis in this order of insects. The methods used in handling the material have been the same as in previous work: fixation with Gilson’s mercuro-nitric, Flemming’s chromo-aceto-osmic, and Hermann’s platino-aceto- osmic fluids, and staining with iron hematoxylin or thionin. The aceto-carmine method has been used in testing fresh material, and as a check on the section method. PHOTINUS PENNSYLVANICUS (FAM. LAMPYRID#) PHOTINUS CONSANGUINEUS (FAM. LAMPYRID&) In my 1906 paper on the spermatogenesis of Coleoptera and other insects, the spermatogonial plate of one of the fireflies, Ellychnia corrusca, was shown on PI. XIII, Fig. 236. The mate- rial was obtained from adults in September at Woods Hole. Only spermatogonia and growth stages of the spermatocytes were present in the testes. “The spermatogonial plate contains nineteen chromosomes (Fig. 1), two V’s, two long rods and fifteen shorter rods. Constancy in form made this material seem very favorable for study of the individuality of the chromosomes, but I have not been able to get the maturation stages. ‘Two other fireflies Tuer JouRNAL or ExPERIMENTAL ZOOLOGY, VOL. VI, NO. I. 102 N. M. Stevens however have been studied, Photinus pennsylvanicus and Pho- tinus consanguineus. In Photinus consanguineus, all of the spermatocyte stages were found in the testis of the adult in summer at Cold Spring Harbor, and dividing spermatogonia were obtained from the larva in October. In the adults of Photinus pennsylvanicus only ripe spermatozoa were present; in the larve collected in October and November, only spermatogonia and growth stages of the sperma- tocytes. A large number of larve were collected and kept during the winter in battery jars with a piece of turf in the bottom. Some of the jars were placed in the greenhouse, others in a cool basement room. ‘The material from both sets of jars was tested from time to time with aceto-carmine, but not until May when the larve were beginning to pupate, were any maturation mitoses found. From May 8 to May 13 both larve and pupz furnished good material. Asin Tenebrio molitor, the pupz contained favor- able divisions in somatic cells. Fig. 2 is the equatorial plate of an odgonium of Photinus pennsylvanicus, from an ovary sectioned in October. There are twenty chromosomes, two longer and two shorter than the others. ‘The two smallest correspond to the small odd chromo- some (x) of the male, seen in Fig. 3, the spermatogonial. plate of nineteen chromosomes. Fig. 4 is an equatorial plate from a male somatic cell, found in mitosis in the digestive tract of a male pupa. In Ellychnia corrusca the synizesis and synapsis stages are similar to those previously described for several of the Coleoptera (Stevens ’06, Pl. IX, Figs. 37 to 42, 61 to 62, and PI. XH, 153 to 154; Nowlin’o6, Pl. I, Figs. 2to5, and Pl. HH, Figs. 52 to 54)—a dense group of short loops at one end of the nucleus in synizesis, followed by a stage in which the loops straighten and unite in pairs. In Photinus pennsylvanicus these stages are quite different. After the last spermatogonial division, the chromo- somes evidently remain condensed for some time, for we find many cysts on the border line between spermatogonia and spermato- cytes in which the nuclei have the appearance of Fig. 5. A slightly later stage shows the chromosomes more crowded together and at one side of the nuclear space. The next stage, which might Further Studies on the Chromosomes 103 be called the synizesis stage, shows the odd chromosome (x) still condensed, and the others forming a rather fine and closely wound spireme (Fig. 7). The fine spireme of Fig. 7 gradually thickens and spreads out (Fig. 8), and later loses much of its staining quality (Figs. 9 and 10). In these later growth stages the spireme winds in such a way as to appear in tangential sec- tion (Fig. 10) to radiate from the heterochromosome (x). There seems to be no question in this case but that synapsis must occur in the stage shown in Figs. 5 and 6, since the spireme, once formed (Fig. 7), remains unbroken until the prophase of the first maturation division. When the chromosomes come into the first spermatocyte spindle, nine of them are often typical tetrads, and one a dyad (Fig. 11). A comparison of x which is univalent with the bivalents con- vinces one that this is a reducing division for the bivalents and quantitative for the odd chromosome. In early metaphase the chromosomes viewed from one pole of the spindle are circular or oval in outline (Fig. 12), but in metakinesis and anaphase (Fig. 13) nine are dumb-bell shaped and one, the heterochromo- some, is circular. Figs. 14 and 15 show the typical metakinesis and early anaphase, while Fig. 16 is a late anaphase. The second spermatocytes all contain ten chromosomes, of which one, the daughter heterochromosome (x) usually stands out to one side of the equatorial plate (Fig. 17), and nearer one pole of the spindle in metaphase and anaphase (Figs. 18 and 19). In most of the Coleoptera where an odd chromosome has been found, it passes undivided to one pole of the first spindle, and ~ divides in the second division, as in the Orthoptera and many Hem- iptera homoptera; but in Photinus we have a case like that of Anasa and several other Hemiptera heteroptera where the unpaired chromosome undergoes its quantitative division in the first spermatocyte while its bivalent companions are being separated into their univalent elements. In Photinus consanguineus the number of chromosomes is the same as in Photinus pennsylvanicus, twenty in the female and nineteen in the male. The chromosomes of the daughter plates are easily counted after the cell has divided (Fig. 20), 104 N. M. Stevens proving that the heterochromosome is not divided in this mitosis and that the spermatozoa are dimorphic, half of them containing ten, the other half nine chromosomes. The chromosomes of an egg follicle cell are shown in Fig. 21, and those of a spermatogonium in Fig. 22. The synizesis and synapsis stages are quite different from those of Photinus penn- sylvanicus, and similar to those of Ellychnia. The synizesis stage has the short crowded loops (Fig. 23). The synapsis stage is less distinct than in some of the cases previously described. One occa- sionally finds a nucleus with the longer synaptic loops (Fig. 24), but more often synapsis and union of chromosomes occur at the same time, giving a mixture of loops, and spireme with sharp angles of which Fig. 25 is perhaps a fair specimen. ‘The hetero- chromosome may be seen in this stage but is more conspicuous in the later pale spireme stage (Fig. 26). Fig. 27 is the first spermatocyte equatorial plate, Fig. 28 and Fig. 29 the metaphase and anaphase, showing the unpaired chro- mosome dividing late. In fact, it frequently divides so late that the two daughter-heterochromosomes are still quite close together and connected by linin fibers in the metaphase of the pairs of second spermatocytes, as shown in Figs. 30 and 31. A pair of daughter plates are given in Fig. 32, showing that as in P. pennsyl- vanicus, the heterochromosome does not divide in the second spermatocyte. Usually it lags behind the daughter plate to which it belongs, so that the two anaphases (Figs. 29 and 33) are charac- terized by a pair of daughter heterochromosomes, and by a single heterochromosome, respectively. These two species of Lampyridz are the only cases which have been found, where the unpaired heterochromosome divides in the first spermatocyte instead of the second. In one, Photinus con- sanguineus, it divides very late, in a stage which is a late anaphase or telophase for the other chromosomes, while in the other species, P. pennsylvanicus, it divides at the same time with the other chromosomes, or only slightly later. It will be interesting to study the spermatogenesis of other Lampyridz for comparison on this point. An abundance of adult material of several other species has been secured and examined, but only spermatozoa Further Studies on the Chromosomes 105 were found. It will therefore be necessary to obtain the larve or pupz before the maturation stages can be studied. LIMONEUS GRISEUS (FAM. ELATERID) In previous work an odd chromosome was found in two species of Elateridz (’06). In both, the male number of chromosomes was nineteen, and in one the female number (twenty) was determined (o6, Pl. XIII, Fig. 229). In both species the unpaired chromo- some was the smallest one. In Limoneus griseus there are seventeen chromosomes in the sper- matogonia (Fig. 34), and the heterochromosome (x) is the largest. The synapsis and synizesis stages are similar to those of Photinus pennsylvanicus. ‘The most conspicuous stage in the transition from spermatogonia to spermatocytes is one in which the condensed chromosomes appear as approximately spherical bodies which nearly fill the small nucleus (Fig. 35). In Fig. 36 the chromosomes are united and somewhat elongated. Elongation continues until all traces of the individual chromosomes, with the exception of the odd chromosome (x), are lost in the fine, closely wound spireme with which the heterochromosome remains connected by linin threads (Fig. 37). As the nucleus enlarges, and the spireme becomes thicker and less stainable, the heterochromosome shows a central vacuole (Fig. 38), and a little later it appears like a spireme wound about in plasmosome material and still connected with the much paler general spireme (Fig. 39). At this point it resembles in its behavior the “‘accessory” of Orchesticus and Xaphidiam (vicClung ’o2,.Pl. Vil; Figs-.45.5; 12). dm the later pale spireme stage (Fig. 40) the heterochromosome is again con- densed. In the first spermatocyte spindle the odd chromosome appears in the equatorial plate in metaphase (Figs. 41 and 42), does not divide, but lags behind the daughter plates (Fig. 43). The chro- mosomes of a pair of daughter plates are shown in Fig. 44. In the telophase (Fig. 45) the heterochromosome holds the hzma- toxylin after the other chromatin has been almost entirely de- stained. In the second spermatocytes (Fig. 46) the unpaired 106 N. M. Stevens chromosome (x) frequently divides somewhat later than the others. Polar plates of the two classes of second spermatocyte mitoses are shown in Figs. 47 and 48. So far as investigated the Elateride have an unpaired heterochromosome which differs from that of the Lampyridz in dividing in the second spermatocytes. NECROPHORUS SAYI (FAM. SILPHID#) Necrophorus sayi has an unpaired heterochromosome, while Silpha americana (’06, Pl. XI, Figs. 141-150) has an unequal pair of heterochromosomes. Necrophorus also differs from Sil- pha in having a much smaller number of chromosomes, thirteen in the spermatogonia (Fig. 49), while Silpha has forty. The synizesis stage is of the finely wound spireme type with the odd chromosome usually visible. There is no preliminary stage that can be pointed out as a synapsis stage, but as the chromosomes remain united in a spireme up to the prophase of the first maturation division, when they appear in the reduced number, it seems certain that synapsis must occur at the close of the last spermatogonial division before the synizesis stage. ‘The bivalents of the prophase of the first spermatocyte mitosis (Fig. 50) have the appearance of two spermatogonial chromosomes united end to end, and in the spindle they are merely somewhat shortened (Fig. 52). The first spermatocyte has seven chromosomes with the unival- ent one oftenest at the center of the group (Fig. 51). Fig. 52 shows the seven chromosomes of one spindle drawnat three different levels of the same section. Here the odd chromosome 1s at the periphery of the plate. ‘The centrosomes in this form are very large. Polar plates of one spindle are shown in Fig. 53, and in Fig. 54 equa- torial plates of the second divisions which proceed as in other sim- ilar cases giving the usual dimorphic spermatids, containing in Necrophorus six and seven chromosomes. CHRYSOMELA SIMILIS (FAM. CHRYSOMELIDZ) Most of the Chrysomelide have an unequal pair of hetero- chromosomes, but Chrysomela similis, like the Diabroticas, has an odd chromosome. At the close of the synizesis stage this Further Studies on the Chromosomes 107 form often shows synapsis with unusual clearness (Fig. 55): Fig. 56 shows a late growth stage with the spireme still staining more deeply than in most cases, and Fig. 57 the equatorial plate of the first spermatocyte with twelve chromosomes. Metakinesis of several of the bivalents and division of the centrosome are shown in Fig. 58, and an early anaphase in Fig. 59. The second sperma- tocytes contain eleven and twelve chromosomes (Fig. 60), as do also the spermatids and spermatozoa. The sperm heads (Fig. 61) have a large middle piece which stains in iron haematoxylin, but not in thionin. LISTOTROPHUS CINGULATUS (FAM. STAPHYLINIDZ) STAPHYLINUS VIOLACEUS (FAM. STAPHYLINIDZ) Three rove-beetles have been examined with a rather small amount of material in each case. All have an unequal pair of heterochromosomes. Listotrophus cingulatus has twenty-six chromosomes in the spermatogonia (Fig. 62), one being very small. The heterochromosome pair is distinguishable in the synizesis stage, which is of the spireme type, and in the later growth stages both members of the pair are clearly separated and associated with a large plasmosome. ‘The chromosomes of the first sperma- tocyte are shown in Figs. 63 and 64, and those of the second divi- sion in Figs. 65 and 66. In the blue rove-beetle, Staphylinus violaceus, the heterochro- mosome pair associated with a plasmosome is shown in Fig. 67. The first spermatocyte contains twenty-two chromosomes (Fig. 68), and the unequal pair shows clearly in a section of a spindle (Fig. 69). The two second spermatocyte equatorial plates appear in Figs. 70 and 71. Another brown rove-beetle, not identified, has twenty-eight chromosomes in the spermatogonia and fourteen in the spermatocytes. TETRAOPES TETRAOPHTHALMUS (FAM. CERAMBYCID&) CYLENE ROBINIA (FAM. CERAMBYCIDZ#) Tetraopes, the common red milkweed beetle, has twenty chro- mosomes. ‘l'wo spermatogonial plates (Figs. 72 and 73) show the different appearance of the chromosomes in different cysts. 108 N. M. Stevens The two smallest are the heterochromosomes. The synizesis and synapsis stages are of the loop type, though not especially clear. Fig. 74 is the equatorial plate of the first maturation divi- sion; and Figs. 75 to 77, metaphase and anaphase, show the une- equal pair of heterochromosomes dividing either earlier or later than the other chromosomes. Only one specimen of Cylene gave any maturation divisions. The remainder of the testes examined contained only spermatids and spermatozoa. The number of chromosomes is the same as in Tetraopes, twenty. The larger heterochromosome shows the peculiarity of holding the stain longer than the other chromosomes. Figs. 78 and 79 are metaphases of the first division from an iron hematoxylin preparation much destained. EPICAUTA CINEREA (FAM. MELOIDZ) EPICAUTA PENNSYLVANICA (FAM. MELOIDZ) Two varieties of Epicauta cinerea, one with all gray elytra and the other with a lighter gray border around the elytra, were studied, and the chromosomes in both, as well as in Epicauta pennsylvanica, found to be of the same number and character— nineteen large and one small chromosome in the spermatogonia. The synizesis and synapsis stages are of the loop type and the maturation divisions result in spermatids one half of which con- tain the small heterochromosome, and one half the large one. Figs. 80 to 83 show the chromosomes of the spermatogonia, first and second spermatocytes. The larger heterochromosome holds the stain as in Cylene. PENTHE OBLIQUATA (FAM. MELANDRYIDZ) Only one pair of Penthe obliquata has been captured. The ovaries and one testis were fixed in Gilson’s mercuro-nitric fluid, and the other testis in Flemming. The Flemming material alone gave any satisfactory results. The presence of an unequal pair of heterochromosomes is shown in Figs. 84 to 86—a _ sperma- togonial plate, a section of a first spermatocyte spindle, and two second spermatocyte equatorial plates. Further Studies on the Chromosomes 10g CICINDELA VULGARIS (FAM. CICINDELID&) In Cicindela primeriana (’06, Pl. XIII, Figs. 198 to 206) the number of chromosomes was twenty, and the heterochromosome pair a large trilobed bivalent. In Cicindela vulgaris the number is twenty-two, three larger than the others (Fig. 87). In the first spermatocyte spindle the conspicuous elements are the trilobed heterochromosome group and a four-lobed or cross-shaped macro- chromosome (Fig. 88). The divisions are like those of Cicindela primeriana. OTHER CHRYSOMELIDZ Among the Chrysomelidz, several other cases of an unequal pair of heterochromosomes will be briefly referred to. Lema trilineata (Figs. 89 to 92) has thirty-two chromosomes, one very small. ‘The synizesis stage is of the loop type followed by synapsis. Doryphora clivicolis is quite similar to Doryphoria to-lineata (o6, Pl. XII, Figs. 151 to 186), and the character of the hetero- chromosome group is much more easily determined. The reduced number of chromosomes is seventeen, instead of eighteen as in 10-lineata. [The chromosomes of the first and second maturation divisions are shown in Figs. 93 to 96. Chrysochus auratus has the loop type of synizesis and synapsis and a typical pair of quite unequal heterochromosomes (Figs. g7 and 98). The reduced number is thirteen. Haltica chalybea, the steel-blue flea-beetle, has twenty-two chromosomes in the spermatogonium (Fig. 100). Only occasionally a specimen of this species has been found in the net, and these have been studied with the aid of aceto-carmine. No drawings have been made of synizesis, synapsis or growth stages. Fig. 101 is a prophase showing the heterochromosomes (/; and f,) and another con- densed pair of chromatin elements which may be m-chromosomes. In the later prophase when the chromosomes are coming into the spindle, the heterochromosomes are often widely separated (Fig. 102), and the same is true of the metaphase (Figs. 103 to 105), so that twelve chromosomes show in the equatorial plate of the first spermatocyte (Fig. 106), but in the late anaphase (Fig. 107), the IIO N. M. Stevens two heterochromosomes are always found between the two polar masses of chromatin separating like any other unequal pair. In the telophase and youngest spermatids they are usually still sepa- rate from the general mass of chromatin (Fig. 108). An abund- ance of material for further study of this form is much to be desired. The chromosomes of Coptocycla clavata are very similar to those of Coptocycla guttata (Nowlin ’06). There are eighteen in the spermatogonium, the two smallest being the unequal pair of hetero- chromosomes. ‘The testes of Lina laponica were examined with the hope that there might be some perceptible difference in chro- mosomes corresponding to the dimorphism described by Miss McCracken (’06) but none was found. The first spermatocytes contain seventeen rather small bivalents, one of which is quite unequal, and the second equatorial plates show clearly the BEE dimorphism. MISCELLANEOUS A number of the Coccinellidz have been found to have nineteen large and one small chromosome in the spermatogonia and ten in the spermatocytes as in Adalia bipunctata (’06, Pl. XIII, Figs. 193 to 197). An unequal pair has also been found in one of the Rhynchophora, Phytonomius punctata, and in Obera tripunctata, one of the Lamiinz. DISCUSSION The character of the heterochromosomes has now been deter- mined for more than fifty species of Coleoptera, belonging to sixteen families. In twelve species an unpaired heterochromo- some has been found, in all of the others an unequal pair, and in Diabrotica soror and Diabrotica 12-punctata (’08) from one to four small supernumerary heterochromosomes may be present in addition to a large unpaired one. In connection with the work on the odd chromosome in the Coleoptera, new material of Stenopelmatus (’05) has been studied, and the spermatogonial number determined as forty-seven instead of forty-six. It was also possible to count one cyst of second Further Studies on the Chromosomes Ill spermatocytes; the numbers are twenty-three and twenty-four, and the odd chromosome can be identified among the twenty-four. Stenopelmatus material, at best, is unfavorable for accurate count- ing on account of the large number of chromosomes and the fact that they rarely form flat plates. There seems at present to be no doubt that whenever an unpaired heterochromosome is present in the first spermatocyte, the sperma- togonial number is odd in the Coleoptera, Orthoptera and Hemip- tera (cases with supernumeraries are an exception to the rule which however, applies if the supernumaries are counted out). It is equally certain that the unequal pair of the spermatocyte is also found in the spermatogonia. In Tenebrio molitor and Photinus pennsylvanicus the chromosomes of the somatic cells of the male have been shown to be of the same number and character as those of the spermatogonia. Not quite so certain is it that an equal pair of heterochromo- somes in the female always corresponds to the unpaired one or the unequal pair. This paper adds two more, Photinus penn- sylvanicus and Photinus consanguineus, to the four species of Coleoptera previously recorded (’05 and ’06) as having such an equal pair of female heterochromosomes. No exceptions have been found, but it proves to be difficult to get suitable material for determining the number of chromosomes in somatic cells. The pupz would probably give good somatic mitoses in nearly every case, but they are rarely to be had unless one can breed the insects. In the flies there was no difficulty in finding dividing oogonia and egg-follicle cells, and in every case an equal pair of large heterochromosomes corresponded to the large and the small one of the male (’08). In the Hemiptera an equal pair of hetero- chromosomes has been determined in the female of a compara- tively large number of species (Wilson ’o5, ’06, ’07, Stevens ’of, Boring ’07). The consensus of evidence would therefore indi- cate that this is the rule for these orders of insects, and that the determination of sex is closely connected with fertilization, since it is evident that only those eggs that are fertilized by spermatozoa containing the odd chromosome or the larger of an unequal pair of heterochromosomes can develop into females, and the males 112 N. M. Stevens must be the result of fertilization by spermatozoa which contain either no heterochromosome or the smaller of an unequal pair. The only other alternative for these insects seems to be that sex is already determined in the egg before fertilization either as a matter of dominance or as a result of maturation, and that fertiliza- tion is selective; i.e., the eggs that are already predetermined to produce females can be fertilized by those spermatozoa only which contain the odd chromosome or the larger of two unequal hetero- chromosomes, while the eggs which are already male can be ferti- lized only by the other class of spermatozoa. Ifa general appli- cation of the results obtained in insects were to be made, the second supposition would certainly cover more cases, but any such general application is premature until adequate evidence is at hand to prove that the sex character is represented in the chromosomes. Further study of the phenomena of synapsis and synizesis in the Coleoptera indicates the existence of at least two distinct types. In the first, which I have called the loop type, synizesis seems to be a prolonged telophase of the last spermatogonial mitosis, the sper- matogonial number of chromosomes appearing as short loops crowded together at one end of the nucleus. After a time the loops straighten and the free ends unite in pairs and the pairs unite to form a spireme. In some cases the synapsis stage is very distinct, in others, synapsis and union to form a spireme occur nearly or quite simultaneously. The second, or spireme type of synizesis is preceded by synapsis which may form a distinct stage as in Photinus pennsylvanicus and Limoneus grisens, or it may occur in the anaphase or telophase of the last spermatogonial mitosis, and a closely wound spireme follow immediately. In this type, the heterochromosomes are usually distinguishable in the synizesis stage outside of the massed spireme, while in Type I they are not seen until after the spireme is formed. Bryn Mawr College March 4, 1908 Further Studies on the Chromosomes 113 LITERATURE CITED Borinc, A. M. ’07—A Study of the Spermatogenesis of ‘Twenty-two Species of Membracide, Jassidz, Cercopide and Fulgorida. Jour. Exp. Zool., vol. iv. McCuune, C. E. ’02—The Spermatocyte Divisions of the Locustide. Kans. Univ. Quart., vol. xi. . McCracken, I. ’06—Inheritance of Dichromatism in Lina and Gastroidea. Journ. Exp. Zool., vol. 1i1. Nowuin, W. N. ’06—A Study of the Spermatogenesis of Coptocycla aurichalcea and Coptocycla guttata, with Especial Reference to the Problem of Sex Determination. Journ. Exp. Zodl., vol. ili. Stevens, N. M. ’o5—Studies in Spermatogenesis with Especial Reference to the “Accessory Chromosome.” Carnegie Institution, Washington, Pub. 36. ’o6—Studies in Spermatogenesis. II. A Comparative Study of the Het- erochromosomes in Certain Species of Coleoptera, Hemiptera and Lepidoptera, with Especial Reference to Sex Determination. Carnegie Institution, Washington, Pub. 36, no. 2. *°o8—A Study of the Germ Cells of Certain Diptera, with Reference to the Heterochromosomes and the Phenomena of Synapsis. Jour. Exp. Zodl., vol. v. *08—The Chromosomes in Diabrotica vittata, Diabrotica soror and Dia- brotica 12-punctata: A Contribution to the Literature on Heterochromosomes and Sex Determination. Journ. Exp. Zool., vol. v. Witson, E. B. ’05—Studies on Chromosomes. II. The Paired Microchromo- somes, Idiochromosomes and Heterotropic Chromosomes in the Hemiptera. Journ. Exp. Zool., vol. 1. ’06—Studies on Chromosomes. III. Sexual Differences of the Chromo- somes Groups in Hemiptera, with some Considerations on the Determination of Sex. Jour. Exp. Zodl., vol. ii. °o7—Note on the Chromosome Groups of Metapodius and Banasa. Biol. Bull., vol. xii. °07—The Supernumerary Chromosomes of Hemiptera. Science, n. s., vol. xxvi. DESCRIPTION OF PLATES The figures were drawn with camera lucida, and the magnification multiplied with the aid of a drawing camera, by 13 or by 2, as follows: Figs. 1, 4, 5 to 10, 22 to 51, 53 to 100, 106—Zeiss2 mm.,120c. X 2. Figs. 2 and 3, 13 to 16, 18 to 20—Zeiss 2 mm.,120c. X 1}. Figs. 11 and 12, 17, 101 to 104, 105 to 107—Zeiss 2 mm.,60c. X 2. Fig. 1o8—Zeiss 2 mm., 60c. X 1%. Figs. 21, 52—Zeiss 1.5 mm.120c. X 2. The plates were reduced one-half. Lettering on Plates % = an unpaired heterochromosome. h, = the larger of an unequal pair of heterochromosomes. h, = the smaller of an unequal pair of heterochromosomes. Pirate I Ellychnia corrusca (Fam. Lampyride) Fig. 1 Spermatogonium, metaphase, nineteen chromosomes. Photinus pennsylvanicus (Fam. Lampyride) Fig. 2 Odgonium, metaphase, twenty chromosomes. Fig. 3 Spermatogonium, metaphase, nineteen chromosomes. Fig. 4 Somatic cell o’, metaphase, nineteen chromosomes. Figs. 5 and 6 Synapsis stages. Fig. 7 Synizesis stage, spireme type. Figs. 8 and 10 Growth stages. Figs. 11 and 12 First spermatocytes, metaphase. Fig. 13 First spermatocyte, metakinesis, polar view. Fig. 14 First spermatocyte, metakinesis, side view. Figs. 15 and 16 First spermatocytes, anaphase. Fig. 18 Second spermatocyte, metaphase. Fig. 19 Second spermatocyte, anaphase. Fig. 20 A pair of spermatids with nine and ten chromosomes. Photinus consanguineus (Fam. Lampyride) Fig. 21 Ovarian follicle cell, metaphase, twenty chromosomes. Fig. 22 Spermatogonium, metaphase, nineteen chromosomes. Fig. 23. Synizesis stage, loop type. Figs. 24 and 25 Synapsis stages. Fig. 26 Growth stage. CHROMOSOMES OF COLEOPTERA PLATE I N. M. Stevens Tue JourNAL or ExPERIMENTAL ZOOLOGY, VOL. VI, NO. I Prate II Photinus consanguineus (continued) Figs. 27 and 28 First spermatocytes, metaphase. Fig. 29 First spermatocyte, anaphase. Figs. 30 and 31 Second spermatocytes, metaphase. Fig. 32 Second spermatocyte, daughter plates. Fig. 33. Second spermatocyte, anaphase Limoneus griseus (Fam. Elateride) Fig. 34 Spermatogonium, metaphase, seventeen chromosomes. Figs. 35 and 36 Synapsis stages. Fig. 37 Synizesis stage, spireme type. Figs. 38 to 40 Growth stages. Figs. 41 and 42 First spermatocytes, metaphase. Fig. 43 First spermatocyte, telophase. Fig. 44 First spermatocyte, daughter plates. Fig. 45 Second spermatocyte, brief rest stage. Fig. 46 Second spermatocyte, anaphase. Figs. 47 and 48 Second spermatocytes, daughter plates containing eight and nine chromosomes. Necrophorus sayi (Fam. Sylphide) Fig. 49 Spermatogonium, metaphase, thirteen chromosomes. Fig. 50 First spermatocyte, prophase. Figs. 51 and 52 First spermatocytes, metaphase. Fig. 53 First spermatocyte, daughter plates. Fig. 54 Second spermatocytes, equatorial plates. CHROMOSOMES OF COLEOPTERA PLATE II N. M. Stevens ® ee @ e 8 8 ) % : x ® BEE? Ode he gm @° ve iO ss j a ‘ @ e@@ @e® $f Dts 52 53 54 Tue JouRNAL or ExperRIMENTAL ZOOLOGY, VOL. VI, NO. I Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. 62 Prater III Chrysomela similis (Fam. Chrysomelide) Synapsis stage. Growth stage. First spermatocyte, equatorial plate. First spermatocyte, metakinesis. First spermatocyte, early anaphase. Second spermatocyte, equatorial plates containing eleven and twelve chromosomes. Developing sperm heads. Listotrophus cingulatus (Fam. Staphylinide) Spermatogonium, metaphase, twenty-six chromosomes. Figs. 63 and 64 First spermatocyte, metaphase. Figs. 65 and 66 Second spermatocytes, metaphase. Fig. 67 Figs. 68 and 69 First spermatocyte, metaphase. Staphylinus violaceus (Fam. Staphylinide) Growth stage. Figs. 70 and 71 Second spermatocyte, metaphase. Tetraopes tetraophthalmus (Fam. Cerambycide) Fig. 72 and 73 Spermatogonia, twenty chromosomes. Figs. 74 and 76 First spermatocytes, metaphase and metakinesis. Fig. 77 First spermatocyte, anaphase. Cylene robinia (Fam. Cerambycide) Figs. 78 and 79 First spermatocytes, metaphase. Epicauta cinerea (Fam. Meloide) Fig. 80 Spermatogonium, twenty chromosomes Figs. 81 and 82 First spermatocyte, metaphase. Fig. 83 Second spermatocyte, equatorial plates of two types. Penthe obliquata (Fam. Melandryide) Fig. 84 Spermatogonium, sixteen chromosomes. Fig. 85 First spermatocyte, metaphase. Fig. 86 Second spermatocytes, metaphase. CHROMOSOMES OF COLEOPTERA PLATE III N. M. Stevens Tue JourNat or ExperiMENTAL ZoOLoGy, VoL. VI, NO.I Pirate IV Cicindela vulgaris (Fam. Cicindelide) Fig. 87 Spermatogonium, metaphase, twenty-two chromosomes. Fig. 88 First spermatocyte, metaphase. Lema trilineata (Fam. Chrysomelide) Fig. 89 Spermatogonium, thirty-two chromosomes. Fig. 90 Growth stage. Fig. 91 First spermatocyte, metaphase. Fig. 92 Second spermatocytes, equatorial plates. Doryphora clivicolis (Fam. Chrysomelide) Fig. 93 First spermatocyte, metaphase, seventeen bivalents, 4, and hg the unequal pair of hetero- chromosomes. Fig. 94 First spermatocyte, late prophase. _ Figs. 95 and 96 Second spermatocytes, equatorial plates. Chrysochus auratus (Fam. Chrysomelide): Fig. 97 First spermatocyte, metaphase. Fig. 98 First spermatocyte, anaphase. Fig. 99 First spermatocyte, equatorial plate. Haltica chalybea (Fam. Chrysomelide) Fig. 100 Spermatogonium, metaphase, twenty-two chromosomes. Fig. 101 First spermatocyte, prophase. Fig. 102 First spermatocyte, later prophase. Figs. 103 to 106 First spermatocytes, metaphase. Fig. 107 First spermatocyte, anaphase. Fig. 108 A pair of spermatids, early stage. CHROMOSOMES OF COLEOPTERA N. M. Stevens ‘ PLATE IV H\~. eee 4 \) {i | \ - Vea . & = \ Se AS reer i Hi] j 102 Tue JourNAL or ExPERIMENTAL ZOOLOGY, VOL. vi, NO. I AN UNPAIRED HETEROCHROMOSOME IN THE APHIDS BY N.M. STEVENS Wirn Two Pirates In two previous papers on the germ cells of aphids (’05 and ’o6) the spermatocytes have been described as having no _hetero- chromosome of any kind. In all of the aphids which have been studied, there is, however, one peculiar lagging chromosome in the first spermatocyte mitosis (’05, Pl. IV, Figs. 37 and 38;’06, PI. Webie. 12, G1. 11, Pies. 33; 34, 42, 40; Pl Il, Figs: oo and 78; Ek IV, Figs. 100, 101, 111, 112). [his chromosome appeared to be a bivalent which separated very late, giving second spermatocytes with equal series of chromosomes (’05, Pl. IV, Fig. 39; ’06, PI. by Bigs 25): : The results recently obtained by Morgan (’08) in the study of the germ cells of Phylloxera (sp. ?), which has a similar lagging chromosome, have led to a reinvestigation of the matter in the aphids. In Phylloxera (sp.?) Morgan finds six chromosomes in somatic cells of female embryos and five in male embryos. ‘The first spermatocytes contain three chromosomes. ‘The lagging chromosome, though it appears about to divide as in the aphids, does not do so, but remains in the larger of the two second sper- matocytes, the cytoplasm dividing very unequally. ‘The smaller cells, containing two chromosomes, degenerate, while the three chromosomes of the larger cells all divide, giving only one kind of spermatozoa; 1. e., such as can fertilize female-producing eggs. The Phylloxerans, therefore, fall into the same category with the other Hemiptera homoptera described by Boring (’07), and with the Hemiptera heteroptera, Coleoptera and Orthoptera, which have an unpaired heterochromosome, the only important Tue JourNAL or ExPERIMENTAL ZOOLOGY, VOL, VI, NO. I. 116 N. M. Stevens difference being that one-half of the second spermatocytes degener- ate, leaving only one kind of spermatozoa—those destined to unite with female-producing eggs. These facts, which are described as quite clear for Phylloxera, are exceedingly obscure for the aphids, and not until I had gone over all of my material three times, was I able to convince myself that the lagging chromosome does not divide in the first spermatocyte, but in the telophase goes into the larger of two second spermato- cytes, the smaller cell sooner or later degenerating. ‘The earlier stages shown in previous figures indicate equal division of chromo- somes and cytoplasm; and in the final division of the cell there is every possible variation of inequality. ‘There is also great varia- tion in different species in the smaller second spermatocytes. In some cases the chromosomes massed in the anaphase of the first spermatocyte, never separate again, and one finds these rounded clumps of chromatin with little cytoplasm among the dividing second spermatocytes, while in the youngest cysts of spermatids they have completely disappeared. In other cases the chromo- somes separate and show no signs of degeneration in the second spermatocyte cysts, and they may even divide, mitosis going as far as a late anaphase, but not ending in cell division. In the latter case degeneration occurs in the younger cysts of spermatids. A few of the most convincing cases will now be described with figures, the different aphids being designated by their host plant as in previous papers. In the green rose aphid the anaphases (Figs. 1,2 and 3) would not even suggest the possibility that the lagging chromosome is not equally divided between the two second spermatocytes, yet in the prophase of the second division (Fig. 4) there is a double chromosome which can be no other than the lagging one of the first division with the two parts folded together. Fig. 5 is a smaller second spermatocyte containing six chromosomes and lacking the double one. Figs. 6 and 7 are first and second sper- matocyte equatorial plates, and Fig. 8 is a side view of a first spermatocyte spindle, showing the heterochromosome (x) not yet divided, while the other chromosomes are in early anaphase. In the star cucumber aphid there is great variety in the ana- Unpaired Heterochromosome in A phids 117 phases. Early stages (Fig. 9) do not indicate any inequality. Later stages (Fig. 10) show some inequality in the size of the two cells, but the lagging chromosome still seems destined to divide. Here again, however, one finds in the prophase of the second division a double chromosome (Figs. 11 and 12) which can be accounted for only by supposing that the lagging chromosome finally pulls back into the larger cell and folds together. In this species the chromosomes can be counted in a few second spermato- cyte prophases and metaphases of the smaller cells, as degeneration occurs mainly in the spermatid cysts. Figs. 12 and 13 show pro- phases of the two kinds of second spermatocytes, Fig. 13 being a Smaller cell lacking the double chromosome. Figs. 14 and 15 are the corresponding metaphases. Among my preparations of aphids collected from Solidago altis- sima, is one, rather lightly stained with iron hematoxylin, which shows one chromosome black while the others are gray, in the first spermatocyte prophases and metaphases. In the prophase stage (Fig. 16) which was mentioned in my ’o5 paper as a possible synizesis stage (Pl. IV, Fig. 34), the dark staining chromosome is the isolated one, and in the metaphase (Figs. 17, 18 and 19) it is single while the other five are double. ‘There are no anaphases in this preparation. ‘The typical anaphases found on other slides are shown in Figs. 20, 21 and 22, Fig. 22 indicating a shifting over to one of the pair of second spermatocytes of the whole of the lagging chromosome and of considerable additional cytoplasm. The prophases of the second division in this species (Fig. 18) do not show the heterochromosome (x) double, and the anaphases look as though it was simply pulled out at right angles to its original long axis (Figs. 20 and 21) and later (Fig. 22), the pulling being relaxed at one end, was returning to its original form. Figs. 24 and 25 are second spermatocytes in metaphase, show- ing five and six chromosomes, Fig. 25 being one of the very few cases where it was possible to count the chromosomes in the smaller cells. In the reddish brown aphid from the beach goldenrod, there are anaphases of the first division in which the lagging chromo- some is distinctly divided as in Fig. 26, others where it is wholly 118 N. M. Stevens within one prospective second spermatocyte, (Fig. 27) and all intermediate stages. In the first spermatocytes there are four large chromosomes of about equal size and two small ones (Fig. 28). In the prophase of the larger second spermatocytes one of the four large chromosomes is larger than the others (Fig. 29). This is as far as the evidence goes for this species. On restaining a pale slide of the Harpswell willow aphid, one cyst of second spermatocytes in metaphase was found. Here the two sizes of cells could be distinguished and chromosomes counted (Figs. 30 and 31). Fig. 32 shows daughter plates of a second spermatocyte in anaphase. Figs. 33 and 34 are prophases of spermatogonial mitoses from a male embryo. Only five chromo- somes could be counted. It was impossible to determine the number in metaphases in the same embryo. ‘This material, if it could be secured in abundance, should give perfectly clear and decisive results on all points connected with the heterochromo- some, but I found it only on one small willow at South Harpswell Me., and have discovered nothing like it anywhere else. In the black aphid on the common milkweed there are four chromosomes in the first spermatocytes, the third largest staining darker in pale iron hematoxylin preparations (Fig. 35). Fig. 36 is an early anaphase showing two double chromosomes and the single elongating heterochromosome (x). Fig. 37 1s a pro- phase from the same preparation. Figs. 38, 39 and 40 are dif- ferent stages in the division of the first spermatocyte. “The num- ber in the small second spermatocytes was not clear, but in a similar aphid from the garden nasturtium these cells have three chromosomes while the larger ones have four (Fig. 41). The woolly aphid from the beech shows well how deceptive the anaphase of the first spermatocyte can be. Fig. 42 is the stage most often seen, while Fig. 43 shows the final result of division. Figs. 44 and 45 are from the Saranac willow aphid, in which both cell and heterochromosome simulate equal division, but in the final stages of mitosis the whole heterochromosome goes over to one cell and the other cell becomes only slightly smaller. In my ’o6 paper, Pl. IV, a few figures (110-113) were given for Unpaired Heterochromosome in Aphids 119 a maple aphid having sixteen chromosomes. Fig. 47 shows a very common appearance of the telophase in this ‘aphid, the chromosomes being massed at one pole and well separated at the other. The massed chromatin is destined to degenerate. An earlier stage (Fig. 46) shows the chromatin massed at both poles, the heterochromosomes (x) dividing, and the cell apparently about to divide equally or nearly so. Fig. 49 shows more inequality in the two cells, but the heterochromosome divided and one-half in each prospective cell. Fig. 48 is quite a different case. Both parts of the heterochromosome (x) are in the nucleus of the larger cell and the chromatin of the smaller cell is densely massed. In Fig. 49 we have another variation: the two parts of the hetero- chromosome have run together and are in the larger cell, but the chromosomes of the smaller cell are not massed together. A careful comparison of first and second spermatocytes in meta- phase brings out the fact that while in the first spermatocyte there is one chromosome considerably larger than the others, in the second spermatocyte equatorial plate there are two larger than the others and nearly equal. One of these is the division product of the large chromosome of the first spermatocyte and the other is the undivided heterochromosome (x) which is second in size in the first spermatocyte (Figs. 50 and 51). In the (nothera aphid, whose male germ cells were described in my first paper on aphids (’05), the failure of the lagging chromo- some to divide in the first spermatocyte is more difficult to demon- strate than in any of the other species. Ona slide where there are dozens of telophases like Fig. 54, only one case could be found where the two parts of the heterochromosome were in one nucleus, leaving only four chromosomes in the other (Fig. 55). The lagging chromosome is the second in size and inseveral prophases of the second mitosis it appears double as shown in Fig. 56, though not so conspicuously so as in the green rose and star cucumber aphids (Figs. 4, 11 and 12). Smaller cells con- taining only four chromosomes and lacking this double chromo- some can also be found (Fig. 57). In another preparation, which has unfortunately been lost, anaphases of the smaller second spermatocytes were seen, and such may also be distin- guished among the degenerating spermatids. 120 N. M. Stevens With the exception of those cited for the Harpswell willow aphid (Figs. 33 and 34)» I am unable to find any spermatogonia or male somatic cells in my aphid material, where the number of chromosomes can be satisfactorily counted. It is perfectly cer- tain that the earlier parthenogenetic eggs and embryonic cells contain an even number of chromosomes—a complete double series of maternal and paternal chromosomes. ‘The question is: Where does the mate of the unpaired heterochromosome (x) of the spermatocytes disappear? With no direct evidence at hand, my present opinion is that the two heterochromosomes must pair before the maturation of the male-producing eggs and separate in that mitosis while the other chromosomes divide longitudinally. I have never been able to find a polar spindle in male eggs; that is, in such cases as that of the Gnothera aphid where the males and females are produced by different mothers. ‘There is one peculiar case, however, mentioned in my ’06 paper, which may be a case in point. In one parthenogenetic individual of the orange milkweed aphid, two eggs in different embryos had only seven chromosomes (Fig. 52) in the equatorial plate of the matu- ration spindle, while all others had eight (Fig. 53). “The two plates were very much alike, each having a large chromosome in the center, evidently corresponding to the two largest in other plates united. Unfortunately I never found any males of this species, although I continued to collect the aphids at short intervals until the plants were killed by frost; but it is possible that, as in the brown rose aphid, only a few scattering males and sexual females appear, while the parthenogenetic female generations go on until destroyed by freezing or starvation. I have found nothing else of this kind in looking over preparations of parthenogenetic individuals col- lected with the sexual generation, but good equatorial plates of polar spindles, cut and stained so that the chromosomes can be satisfactorily counted, are always rare in aphid material. If my surmise as to the maturation of the male-producing aphid eggs should prove to be correct, it would seem probable that these eggs develop into males because a dominant female sex-chromosome has been removed and for that reason only, since the same par- thenogenetic mother aphid may contain embryos of three kinds, Unpaired Heterochromosome in A phids $21 parthenogenetic female, sexual female and male, of approxi- mately the same age and therefore developing under the same con- ditions of temperature and nutrition. Thus Castle’s (’03) theory in regard to the appearance of males among parthenogenetic insects would be realized by a slightly different method of matura- tion. ‘This would bring us one step nearer the conclusion that sex and sexual characters are really represented, in the germ cells of insects at least, by the heterochromosomes. As to the fact that the lagging chromosome of the aphids is a heterochromosome intimately connected with the phenomenon of sex determination, the present reinvestigation of the male germ cells, I think, leaves no doubt. ‘The question as to how and when the number of chromosomes of the parthenogenetic female genera- tions is reduced to that of the male individuals will be further investigated as soon as suitable material can be obtained. The discovery of Morgan that only female-producing sper- matozoa develop in Phylloxera, and the above corroborating facts for the aphids at once suggest the idea that in the bee and the ant an unpaired male heterochromosome may be left in the male- producing eggs after maturation, and that, as in the phylloxerans and aphids only female-producing (containing a male sex-chromo- some) spermatozoa develop. ‘The complete reduction in number of chromosomes occurring during the maturation of the egg instead of in the spermatocytes might prevent the detection of such a heterochromosome in these forms, but study of other hymenopterous insects such as Nematus, where in some species only females come from the parthenogenetic eggs, in other species only males, and in still others both males and females, may throw light on the problem. In all of the other cases familiar to me, where an unpaired heterochromosome is present in the spermatocytes, if it is destined to divide in the first maturation mitosis, it is attached to mantle fibers from only one pole of the spindle. This is true for many Hemiptera homoptera (Stevens ’06; Boring ’07); for the Orthoptera (McClung and others) and for several of the Coleop- tera (Stevens ’06 and ’08). On the other hand, when the odd chromosome is attached to mantle fibers from both poles of the 122 N. M. Stevens first spindle, it divides in the first division, as in Anasa and other Hemiptera heteroptera (Wilson ’o5, 06) and in Photinus (Stevens ’08). In the latter case the division products of the odd chromo- some make connection with only one pole of the second spindle. In the aphids the unpaired heterochromosome is connected with fibers from both poles of the first spindle and, in most cases, until a very late anaphase or telophase, appears to be about to divide equally. Both cell and heterochromosome begin to do one thing, and finally do something quite different. One can hardly fail to be impressed with the idea that we have here a case where the karyokinetic apparatus is imperfectly adjusted to the demands made upon it by the cell as a whole or by the chromo- somes, but that the end is finally attained in spite of this imperfect working of the spindle. This peculiarity in the first maturation division, it would seem, must have come in with a change from purely sexual reproduction to parthenogenetic, involving the sup- pression of the male-producing spermatozoa. The apparent equal division of the lagging chromosome in anaphases of the first spermatocyte division together with the facts that there is no condensed heterochromosome in growth stages of the spermatocytes and that all of the chromosomes of the second spermatocytes certainly divide longitudinally (’05, Pl. IV, Fig. 41; ’06, Pl. II, Fig. 50; Pl. IV, Figs. 108 and 109) led to the conclusion that there was no heterochromosome in the aphids. The late shifting over of the whole lagging chromosome and a variable amount of cytoplasm into one of each pair of second sper- matocytes, together with the significance of the smaller degener- ating cells were overlooked until the preparations were reéxamined in the light of Morgan’s results on Phylloxera. ‘This experience with the germ cells of the aphids indicates the probability that a sex-determining differentiation of chromosomes in the male germ cells may exist in other cases where it has not yet been detected because of a large number of small chromosomes, or of some unexpected peculiarity in the behavior of the heterochromo- somes. I desire to express here my indebtedness to Prof. T. H. Mor- gan for urging a reinvestigation of my aphid material, and also Unpaired Heterochromosome in Aphids 138 for allowing this paper to appear in advance of his more elab- orate paper on the germ cells of the phylloxerans. Bryn Mawr College May 20, 1908 LITERATURE CILrED, Borinc, A. M. ’07—A Study of the Spermatogenesis of Twenty-two Species of Membracide, Jassida, Cercopida and Fulgoride. Jour. Exp. Zool., vol. iv. CasTLe, W. E. ’03—The Heredity of Sex. Mus. Comp. Zodl., Harvard, No. 137. Morean, T. H. ’08—The Production of two kinds of Spermatozoa in Phylloxerans- Functional ‘Female Producing” and Rudimentary Spermato- zoa. Proc. Soc. for Exp. Biol. and Med., vol. v, no. 3. Stevens, N. M. ’o5—A Study of the Germ Cells of Aphis rosae and Aphis ceno- there. Jour. Exp. Zool., vol. 11. ’o6—Studies on the Germ Cells of Aphids. Carnegie Inst., Wash., Pub. 51. °o6—Studies in Spermatogenesis. II. A Comparative Study of the Heterochromosomes in Certain Species of Coleoptera, Hemip- tera and Lepidoptera, with Especial Reference to Sex Determina- tion. Carnegie Inst., Wash., Pub. 36, no. 2. ’08—The Chromosomes in Diabrotica vittata, Diabrotica soror and Dia- brotica 12-punctata. Journ. Exp. Zodl., vol. v, no. 4. *o8—Further Studies on the Chromosomes of the Coleoptera. Jour. Exp. Zool., vol. vi, no. I. Witson, E. B. ’o5—Studies on Chromosomes. II. The paired Microchromo- somes, Idiochromosomes, and Heterotropic Chromosomes in the Hemiptera. Jour. Exp. Zodl., vol. 11. ?06—Studies on Chromosomes. III. Sexual Differences of the Chromo- some Groups in Hemiptera, with some Considerations on Deter- mination and Inheritance of Sex. Jour. Exp. Zodl., vol. 11. * DESCRIPTION OF PLATES The figures were all drawn with a Zeiss 1.5 mm. oil immersion objective and a Zeiss compensating ocular 12. The magnification was doubled with a drawing camera, and the plates reduced one-half, giving a magnification of 2000 diameters. Lettering on Plates p = a plasmosome 4 = the unpaired heterochromosome Prate I Green rose aphid Figs.1 to 3 First spermatocyte, anaphase. Fig. 4 Second spermatocyte, large, prophase, seven chromosomes. Fig. 5 Second spermatocyte, small, prophase, six chromosomes. Fig. 6 First spermatocyte, metaphase. Fig. 7 Second spermatocyte, metaphase. Fig. 8 First spermatocyte, metakinesis. Star cucumber aphid Fig. 9 First spermatocyte, early anaphase. Fig. 10 First spermatocyte, telophase. Figs. 11 and 12 Second spermatocyte, large, prophase, five chromosomes. Fig. 13 Second spermatocyte, small, prophase, four chromosomes. Fig. 14 Second spermatocyte, large, metaphase, five chromosomes. Fig. 15 Second spermatocyte, small, metaphase, four chromosomes. Aphid from Solidago altissima Fig. 16 First spermatocyte, prophase, possibly a synizesis stage. Fig. 17 First spermatocyte, metaphase. Figs. 18 and 19 First spermatocyte, lateral view of chromosomes in metaphase. Fig. 20 First spermatocyte, anaphase. Figs. 21 and 22 First spermatocyte, telophase. Fig. 23 Second spermatocyte, large, prophase, six chromosomes. Fig. 24 Second spermatocyte, large, metaphase, six chromosomes. Fig. 25 Second spermatocyte, small, metaphase, five chromosomes. Aphid from beach goldenrod Fig. 26 First spermatocyte, anaphase. Fig. 27 First spermatocyte, telophase. Fig. 28 First spermatocyte, metaphase. Fig. 29 Second spermatocyte, large, prophase, six chromosomes. UNPAIRED HETEROCHROMOSOME IN APHIDS PLATE I N. M. Srevens 25 Tue JouRNAL oF ExpERIMENTAL ZOOLOGY, VOL. VI, NO. 1 Pirate II Harpswell willow aphid Fig. 30 Second spermatocyte, large, metaphase, three chromosomes. Fig. 31 Second spermatocytes, small, metaphase, two chromosomes. Fig. 32 Second spermatocyte, daughter plates. Fig. 33 Spermatogonium nucleus, prophase. Fig. 34 Similar prophase, drawn at two foci in same section, five chromosomes. Black milkweed aphid Fig. 35 First spermatocyte, metaphase. Fig. 36 First spermatocyte, metakinesis. Fig. 37 First spermatocyte, prophase. Fig. 38 First spermatocyte, anaphase. Figs. 39 and 40 First spermatocyte, telophase. Fig. 41 Second spermatocytes, small, of a similar nasturtium aphid, three chromosomes. Woolly beech aphid Fig. 42 First spermatocyte, anaphase. Fig. 43 First spermatocyte, telophase. Saranac willow aphid Figs. 44 and 45 First spermatocyte, telophase, two stages. Maple aphid Fig. 46 First spermatocyte, anaphase. Figs. 47 to 49 First spermatocyte, telophase, three common types or stages. Fig. 50 First spermatocyte, metaphase. Fig. 51 Second spermatocyte, metaphase. Orange milkweed aphid Fig. 52 Equatorial plate of maturation spindle of parthenogenetic egg, exceptional case with seven chromosomes. Fig. 53 Usual case with eight chromosomes. - - Aphis cenothere Fig. 54 First spermatocyte, an early telophase. Fig. 55 First spermatocyte, later telophase, with the divided heterochromosome («) in one nucleus. Fig. 56 Second spermatocyte, large, prophase, five chromosomes. Fig. 57 Second spermatocyte, small, four chromosomes. UNPAIRED HETEROCHROMOSOME IN APHIDS PLATE $II N. M. Srevens SW Se may o- We oh "cr CH | oe oy | a 4 ae err ON en 33 : 30 Tue JourNAL or ExPERIMENTAL ZOOLOGY, VOL. VI, NO. I tit SEPEECT OF A CENTRIFUGAL, FORCE > UPON THE DEVELOPMENT AND SEX OF PARTHENO- GENERIC EGGS OF HYDATINA SENTA BY DAVID DAY WHITNEY Wirn One Pirate it} il iiategoYahirentorls Aa oncRStie co cs DOES C GO SC ODE Ie ache TRO MER RAE bp a aa 125 II Relation of the first cleavage plane to the stratification of the material of eggs centrifuged Wwhenitie permunalwvesicleisintachys.= 0 see eet tiie oe ei tion ee eae Tees 126 III Relation of the first cleavage plane to the stratification of the material of eggs centrifuged Wwhenthe maturation spind levs|presentateas)-rrsoeieeet sitet ee eee eee 127 IV _ Condition of the young animals which developed from centrifuged eggs................... 128 V_ The proportion of male and female producing females in the first, eeu and third gen- eLationsraitencentribu pin ga store yes ssisaie ove ate eta ohare va Nae vaateiarel a eaAere ee eer evOe 130 Wilt General Aiscussi Omer. cere cette tierce a cay anh oto T ANS Soe eee oie eee lags Ne En a 133 AV Die SS Urns aay eye ebpe pat Wokeys ckscrcsote ero bold wfsvers cites Gicisre cies oan avser feaevau ale Peerereeie Meee eee 134 AVAL IL tera Gunes ae tect. satin eaveretoreisie Bsr ets < osciatetse Sevan ede ioc enaljer egos coe a) eporaatmelent alee peepee 135 I. INTRODUCTION The eggs of Hydatina senta are very favorable for experiments with the centrifugal machine. ‘The adult females which contain eggs in stages up to and including the first maturation spindle may be centrifuged at the rate of twenty thousand revolutions in two to three minutes without any apparent injury. The animal is so transparent that the eggs can be seen immediately after cen- trifuging and their condition recorded. ‘The materials in the egg are separated into three distinct zones, a pink zone, a middle clear zone and a gray zone. ‘This strat fied material only becomes partly redistributed in the egg before cleavage and sometimes scarcely any redistribution takes place. After the egg is laid the polar body is formed and the first cleavage appears in thirty- five 10 forty-five minutes. The eggs develop within forty-eight to seventy-two hours and produce in most cases normal embryos. Tue JourNAL or ExPERIMENTAL ZOOLOGY, VOL. VI, NO. I. 126 David Day Whitney The following experiments and observations were made at the suggestion and under the supervision of Prof. T. H. Mor- gan. >) II RELATION OF THE FIRST CLEAVAGE PLANE TO THE STRATIFEI- CATION OF THE MATERIAL OF EGGS CENTRIFUGED WHEN THE GERMINAL VESICLE IS INTACT If several hundred animals are centrifuged at the same time and are examined immediately many can be seen to contain eggs, the contents of which are sharply differentiated into three zones. As the animals are thrown down during the centrifuging process in any position the stratification of the eggs may lie in any relation to the median longitudinal axis of the female. The pink zone may be toward the head or foot of the animal, toward the side next the stomach or on the opposite side, or in any other inter- mediate position. ‘The gray zone is always on the opposite side of the egg from the pink zone and the middle zone between the two. The germinal vesicle is found at the junction of the clear middle and pink zones and can be easily recognized owing to its large size (Fig. 1). From the various lots of females, centrifuged twenty thousand revolutions, eight were selected containing the pink zone toward the head of the animal and the gray zone toward the foot. The germinal vesicle was plainly visible near the line between the clear middle and pink zones. Each of these eggs was carefully watched while it remained in the oviduct of the female as well as after it was laid. In every ege the first cleavage appeared at that end of the egg which con- tained the pink pigment. ‘The first division was unequal as it is in the normal egg. In some cases all the pink zone was cut off in the smaller cell while in other cases only a part of it was included in this cell. ‘The gray zone was always included in the larger cell. | Six other females which had the pink zone of the egg toward the foot of the animal and the gray zone toward the head were isolated and the further history of each egg was carefully followed. Effect of a Centrifugal Force upon Development and Sex 127 The first cleavage in all these eggs appeared in the end which contained the pink material. It is hardly possible that in these fourteen eggs the pink material and the germinal vesicle were driven in each case to the animal pole. If this assumption is correct these two lots of eggs indicate that the position of the first cleavage is determined by the induced position of the germinal vesicle at eitherend of the egg. When, however, the germinal vesicle is carried to the side of the egg with the pink material, that is, when it comes to lie midway between the two ends of the egg, it moves later toward one of the ends of the egg since the first cleavage in such cases is at one end. III RELATION OF THE. FIRST CLEAVAGE PLANE TO THE STRATIFI- CATION OF THE MATERIAL OF EGGS CENTRIFUGED WHEN THE MATURATION SPINDLE IS PRESENT Some of the centrifuged females were found containing eggs stratified into the three characteristic zones but showing the maturation spindle apparently at the side of the egg, where it normally forms, in the clear middle zone (Fig. 2). In cases where the pink zone was in the end of the egg toward the head of the animal the first cleavage sometimes appeared in that end of the egg and cut off the pink zone in the small cell. At other times it appeared at the opposite end of the egg and included the gray zone in the small cell. In other cases where the gray zone was inthe end of the egg toward the head of the animal the first cleavage sometimes appeared in that end but at other times it appeared in the opposite end of the egg. When the eggs were centrifuged so that the stratification was from side to side, instead of from end to end the first cleavage usually cut off a part of each of the three zones in each cell. Apparently the first cleavage appeared at that end of the egg nearest to the maturation spindle irrespective of the stratified zones of the egg. 128 David Day Whitney IV CONDITION OF THE YOUNG ANIMALS WHICH DEVELOPED FROM CENTRIFUGED EGGS If the materials of the egg which are differentiated into zones by the centrifugal force are important formative substances their displacement and their different location in the first cleavage stage ought to cause abnormalities in the embryos and in the mature animals which develop from such eggs. The following experiments which are only a few of those carried out will serve to indicate to what extent the embryos form centri- fuged eggs were affected by the displacement of the materials. Experiment I. January 11, 1908, at 8:30 p.m., many females were centrifuged about twenty thousand revolutions and then placed in a watch glass containing tap-water and allowed to lay their eggs. At I0 p.m., thirty-seven eggs were isolated. January 13, at 9:30 a.m., there were in the dish thirty-two apparently normal young females, three winter eggs and two dead parthenogenetic eggs. Experiment V (Control). January 21, at 10 p.m., two hun- dred and eighty eggs that had been laid by females which had not been centrifuged were isolated in a watch glass containing tap- water. January 22, at 8 p.m., two hundred and forty normal young animals were removed from the dish. January 23, at I1 a.m., six normal young removed and of the thirty-four unhatched eggs which remained thirty-one were winter eggs and three were dead parthenogenetic eggs. Experiment XI. March 14, 7:45 p.m., several hundred females were taken from the culture jar and placed in a watch glass con- taining tap-water. No eggs had been laid by 8 p.m. These animals furnished material for the following experiments. Lot A. At 8:35 p.m., twenty-five eggs which had been laid since 8 p.m., were isolated and centrifuged twenty thousand revolutions. March 15, at 12:40 p.m., sixteen young normal females were taken out of the dish and nine unhatchedeggsremained. At g p.m. the nine eggs were still unhatched. March 16, at 10 a.m., two of Effect of a Centrifugal Force upon Development and Sex 129 the eggs had hatched. Of the seven remaining eggs one was a winter egg, three were dead parthenogenetic eggs and the other three contained living embryos which were apparently unable to break through the egg envelope. Lot B. March 14, at 8:40 p.m., twenty eggs were isolated and allowed to remain undisturbed. At 9:10 p.m., these eggs were centrifuged in the same way and treated as Lot A. March 15, at 1:30 p. m., seventeen young normal females were in the dish and three unhatched eggs which contained living embryos. At 9 p-m., two of these eggs had produced normal females but the other was still unhatched. Lot GC. March 14, at 8:45 p.m., thirty eggs were isolated and allowed to remain undisturbed. At 9:45 p.m., these eggs, in some of which the first cleavage and in others the second and third cleavage had appeared, were centrifuged and treated as Lot A. March 15, at 1:40 p.m., fourteen normal young females were -removed from the dish and sixteen unhatched eggs remained At 9 p.m., seven other normal young females were removed. March 16, at 10 a.m, two other normal young females were removed. Seveneggs remained. One was a winter egg, two were dead parthenogenetic eggs and four contained living embryos. Lot D (Control). March 14, at 10:30 p.m., thirty-three eggs were isolated. March 15, at 12:30 p.m., thirteen normal young females and two normal young males were removed from the dish. At p.m., seventeen other normal young females and one normal young male were remov d. In Experiment I the eggs were centrifuged when they contained the germinal vesicle or else the maturation spindle. In Expen- ment XI, Lot A, they were centrifuged in the stage when the polar body is forming or just after it was formed. In Lot B the eggs we e centrifuged just before the first cleavage and in Lot C dur- ing and after the first cleavage. In some of these experiments the young animals which de- veloped from the centrifuged eggs seem to die sooner if not fed than do the animals developing from normal eggs. Also there 130 David Day Whitney are more cases of the embryos unable to get out of the egg after they are fully formed. They can be seen writhing and twisting inside the egg and sometimes lived there as long as seven days. These cases of abnormalities were due perhaps to the food material in the egg being displaced and therefore unable to nourish certain muscles which are well supplied with food material in the normal embryo. Consequently the young animal was weaker in certain parts of its body. In a few cases the eggs began to develop but apparently soon ceased and never produced embryos which showed any ciliary movement. In normal embryos the ciliary movement around the head can be seen several hours before hatching. Whether the early death of such eggs is due to abnormal cleavage, mis- placed egg substance, or misplaced chromosomes in division is not clear. It is nevertheless apparent from these experiments that a very high percentage of normal young animals develop from eggs that have been centrifuged in the various stages of their early development. \ THE PROPORTION OF MALE AND FEMALE PRODUCING FEMALES IN THE FIRST, SECOND AND THIRD GENERATIONS AFTER CENTRIFUGING If the dislocation of the egg substances has any influence on sex it should become evident by following the history of individual eggs in which the zones of stratification are differently arranged in their relation to the first cleavage plane. The following data give the result of experiments carried out to examine the question. Experiment XXXI. March 5, at 2:15 p.m., a female contain- ing a large egg was centrifuged twenty thousand revolutions. At 3:10 p.m., the egg had been laid and was in the first cleavage stage. ‘The pink zone was entirely included in the smaller cell (Fig. 3). On March 6, at 11 a.m., a normal young female was swimming about in the dish. Food was then added. She produced eggs Effect of a Centrifugal Force upon Development and Sex 131 and on March 8, at 10 a.m., four young daughter-females were present in the dish. This female produced twenty-nine eggs which developed into females. ‘Iwo of these daughter-females matured and produced males and twenty-seven matured and produced females. Experiment XX XIII. The conditions, size of egg, and the arrangement of egg material in the first cleavage were approxi- mately the same as in Experiment XXXI. A normal young female developed from the egg, matured and produced males. Experiment XX XVII. March 5, at 2:15 p.m., a female con- taining a large egg was centrifuged. At 4:30 p.m., the egg that had been laid was in the first cleavage stage. The small cell included portions of the pink and clear zones, Fig. 4. On March 6, at 11 a.m., a normal young female was swim- ming about in the dish. Food was added. ‘This female grew to maturity and produced twenty-five eggs, all of which developed into females. One of these daughter-females matured and pro- duced males and twenty-four matured and produced females. Experiment XLIV. The conditions, size of egg, and the ar- rangement of the egg material in the first cleavage were approxi- mately the same as in Experiment XXXVII. A normal young female developed from the egg, matured and produced twenty- five eggs, all of which developed into female-laying females. In later generations males appeared. ExperimentX XXII. March 5, at 2:15 p.m.,a female containing a large ege was centrifuged. At 3:30 p.m., an egg had been laid and was in the first cleavage stage. The small cell included por- tions of the three zones (Fig. 5). On March 6, at 9 a.m., a normal young female was present in the dish. Food was added. ‘This female matured and produced only three eggs. One of these eggs developed into a male-laying female and the other two developed into female-laying females. This small production of eggs was due to the scanty amount of food given to the female. Experiment L. March to, 1 p.m., a female containing a large ege was centrifuged. At 3 p.m., the egg had been laid and was in the first cleavage stage. The small cell contained about two- 132 David Day Whitney thirds of the gray zone and a portion of the clear zone (Fig. 6). On March 11, at II a.m., a normal young female was swim- ming about in the dish. Food was added. ‘This female matured and produced fifteen eggs which developed into females. One of these daughter-females produced males and the other fourteen produced females. Experiment LV. ‘The conditions, size of egg and the arrange- ment of the egg material in the first cleavage were approximately the same as in Experiment L. A normal young female developed from the egg, matured and produced fourteen eggs all of which developed into female-laying females. In later generations males appeared. Several small male eggs were centrifuged in the same manner as the largeeggs. Insome of these the first cleavage plane appeared so as to cut off all the pink zone in the small cell and in others it cut off some of each of the three zones. In both cases apparently normal males were produced. None of the females in the above experiments produced the normal number of eggs, which is forty to fifty, because of poor food conditions. In former experiments it has been shown that the percentage of male-laying females in a family of daughter-females may vary from 0 to 50 per cent and also that the percentage of male-laying females in one generation is no indication what it may be in the next generation. Experiments XX XI to LV. In these experiments the appear- ance of male-laying females from the various forms of centri- fuged eggs does not seem to be markedly different from normal cases. In the experiments where the daughter-females of a family were all female-laying females males always appeared in the later generations, thus showing that no pure female-laying female strains were produced. Moreover, large (female) eggs never produced male animals nor did small (male) eggs ever produce female animals. Effect of a Centrifugal Force upon Development and Sex 133 GENERAL DISCUSSION Lyon has found in the eggs of the sea-urchin, Arbacia, that the first cleavage plane is always at right angles to the plane of stratif- cation of the egg material. Lillie has shown that the stratification, caused by centrifuged force, of the material in the egg of Chztopterus plays no part in determining the position of the polar lobe. When the germinal vesicle is still intact the egg has a well defined polarity, that 1s, one end is the animal and the other end the vegetative pole. If the germinal vesicle is driven to the vegetative pole the polar spindle which develops from it always migrates to the animal pole and there forms the polar bodies. In later work on sea-urchin’s eggs Morgan and Lyon show that, “while the cleavage conforms strictly to the induced stratification, the gastrulation does not conform to the symmetrical arrangement of the materials. ‘The exceptional cases show that there is no necessary relation between stratification of the materials as such and the embryonic axes.” However, in the eggs of Cumingia Morgan has shown that the stratification of the egg material does not influence the position of the first cleavage plane. He says, “This difference is due to the shifting of the nucleus in the egg of the sea-urchin, while the spindle in Cumingia retains its original orientation.” It must be borne in mind that the above results were obtained from eggs centrifuged at different stages in their maturation. The eggs of Chztopterus were centrifuged when in the germinal vesicle and maturation spindle stages, while the eggs of arbacia were centrifuged when in the female pronucleus stage. ‘The eggs of Cumingia were also centrifuged when in the maturation spindle stage and those of Hydatina in the germinal vesicle stage as well as in the maturation spindle stage. When the eggs of these different animals are centrifuged in the maturation spindle stage the spindle is not usually moved from its original position and consequently the first cleavage take place precisely as it does in normal eggs. When the eggs of Chatopterus and Hydatina are centrifuged 134. David Day Whitney in the germinal vesicle stage the later histories of the germinal vesicles differ. In the egg of Chztopterus the maturation spindle which develops from the germinal vesicle, according to Lillie, migrates to the animal pole if it does not happen to be located at that pole, while in Hydatina the maturation spindle never migrates from the end of the egg into which the germinal vesicle is driven by the centrifugal force. In the egg of Arbacia the female pro- nucleus may be so oriented by centrifugal force that the direction of the first cleavage plane is due rather to its location and follows in consequence the stratification of the materials in the egg. None of the previous workers have reared the embryos from centrifuged eggs to maturity because the forms upon which they worked were not suitable for such experiments but Hydatinais an exceptionally favorable form for such work. Eggs were centrifuged in various stages of maturation so that the zones of egg materials were differently arranged in their relation to the first cleavage plane, thus making it possible that in some cases the pink material of the egg would be included in the cells that make up the anterior end of the embryo while the gray material would be included in the cells of the foot region or vice- versa. In other cases the material would be more or less equally distributed in the anterior and posterior regions of the embryo. In no case was the sex of the eggs changed and such eggs pro- duced a very high percentage of normal young males and females. Furthermore the young females grew to adult animals and _ pro- duced normal offspring of which the sex ratio was apparently normal. It would, therefore, seem that the effect of centrifugal force upon the eggs of Hydatina senta is not sufficient to cause any noticeable change of structure or of sex in the animals that develop from them. VI SUMMARY 1 When the unsegmented eggs of Hydatina senta are centri- fuged twenty thousand revolutions the materials in the eggs are stratified into a pink zone, a clear middle zone and a gray zone. Effect of a Centrifugal Force upon Development and Sex 135 2 Ifeggs are centrifuged a short time before maturation when the nucleus is intact the nucleus 1s carried to the top of the clear zone against the bottom of the pink zone. 3 Very little redistribution of the egg materials takes place before the first cleavage. In consequence the segmented egg retains the distribution of materials impressed on it by the centrifugal force. 4. The first cleavage plane always appears at that end of the egg at which the pink zone and germinal vesicle are located. It forms across one end as in the normal egg separating a smaller and a larger cell. 5 The egg centrifuged after the polar spindle has formed shows that the spindle does not move from its original position. Its location determines the position of the first cleavage plane in so far as this appears at the end of the egg nearest to where the spindle lies. 6 Normal animals, both males and females, develop from centrifuged eggs and these have been reared to sexual maturity. 7 The sex of animals developing from large (female) or small (male) eggs is not affected by the centrifugal force nor is the sex ratio in the descendants of females developing from centrifuged eggs altered. Zoological Laboratory Columbia University April 25, 1908 BRAG WIRE Hertwic, O. ’97—Ueber einige am befruchteten Froschei durch Centrifugalkraf hervorgerufene Mechanomorphosen. Sitzungsbere preuss. Akad. Wiss. Berlin, math-phys. Klasse. °99—Beitrage zur experimentellen Morphologie und Entwicklungs- geschichte. 4. Ueber einige durch Centrifugalkraft in der Ent- wicklung des Froscheies hervorgerufene Veranderungen. Arch. f. mikr. Anat., lit. ’04—Wietere Versuche tiber den Einfluss der Centrifugalkraft auf die Entwicklung tierischer Eier. Arch. f. mikr. Anat., xiii. Lituiz, F. ’06—Observations and Experiments Concerning the Elementary Phe- nomena of Embryonic Development in Chetopterus. Journ. Exp. Zodl., vol. iii. 136 David Day Whitney Lyon, E. P. ’06—Some Results of Centrifugalizing the Eggs of Arbacia. Amer. Journ. Physiol, xv. Morean, T. H. ’o06—The Influence of a Strong Centrifugal Force on the Frog’s Egg. Arch. f. Entw. Mech., xxii. ’o7—The Effect of Centrifuging the Eggs of the Mollusc, Cumingia. Science, n. s., vol. xxvii, no. 680, pp. 66-67. Morcan AnD Lyon ’07—The Relation of the Substances of the Egg, separated by a Strong Centrifugal Force, to the Location of the Embryo. Arch. f. Entw.-Mech., xxiv. WetzeEL, G. ’04—Centrifugalversuch an unbefruchteten Eiern von Rana fusca. Archiv. f. mikr. Anat., ]xiii. DESCRIPTION OF PLATE Fig. 1 Section of a centrifuged egg, showing the three zones and the germinal vesicle located near the boundary of the pink and the clearzones. Gilson mercuro-nitric fixing fluid and Heidenhain’s iron hematoxylin stain. Fig. 2 Section of a centrifuged egg showing the three zones and the maturation spindle in the clear zones. Yolk granules are lodged against the achromatic figure. Bouin’s fixing fluid andHeiden- hain’s iron hematoxylin stain. Figs. 3to6 Free hand semi-diagrammatic drawings of living eggs in the first cleavage stage which were centrifuged when in the oviducts of the females. The fine stippling shows the position of the pink zone, the coarser stippling shows the position of the gray zone, and the clear space indicates the clear zone. Fig. 3 Experiment XXXI—The pink zone included in the small cell and the gray and clear zones are in the larger cell. Fig. 4 Experiment XXXVII—About one-half of the pink zone and a portion of the clear zone included in the small cell and the other part of the pink zone, all the gray zone, and a part of the clear zone are included in the larger cell. Fig. 5 Experiment XXXII—Portions of each of the three zones in each of the two cells. Fig. 6 Experiment L—About two-thirds of the gray zone included in the small cell while the other third of the gray zone together with all of the pink zone and nearly all of the clear zone are included in the larger cell. EFFECT OF A CENTRIFUGAL FORCE UPON DEVELOPMENT AND SEX Daviv Day Wuuitnry 5 Tue Journat or Experimenta ZoéxoGy, vot. VI, NO. I. OBSERVATIONS ON THE MATURATION STAGES OF THE, PARTHENOGENETIC AND? SEXUAL EGGS OF HYDATINA SENTA BY DAVID DAY WHITNEY Pe Unter OP Ct Oe 5 afore eats is, cherstavexmycnare oj avenarayeies opm ec lausrsce\ele nai, ouatVohere ay HPSS Feu eh rele erin sean Ne tr oM 143 (0 DPS beste) 9 eatin COO MCIOE arlon CORMIER AC eticeh ac ioe Acad etic BG MOO pa > oo. om GreiO.0 144 WATT esi tera turers ete vere ce hecaie cere tueye te lec wic het el niscmte res ATS ers (Re ara toa a Mere tea elon Tn Recent Pak 145 I INTRODUCTION Despite the experiments that have been carried out by several workers to discover how sex is determined in parthenogenetic eggs the attempts to show that such external factors as tempera- ture or food influence the result do not appear to have been suc- cessful. Attention has turned more recently to the possibility that there are internal factors in the eggs that are all-important in producing males or females. As early as 1845 Dzierzon brought forward very strong evidence to show that the eggs of the honey-bee, Apis mellifica, always develop into males if unfertilized, but if fertilized they develop into females (queens or workers). In other words, internal rather than external agents bring about the result. “This theory has been often attacked and strongly defended, and now seems to be generally accepted. In the aphids Balbiani and Stevens find that the same female may produce both male parthenogenetic and fertilized or winter eggs. Lauterborn finds the same phenomenon in the Rotifer, Asplanchna, and Issakowitsch in a Daphnid. Whether the eggs that are fertilized are originally male eggs or develop from a dif- THe JourRNAL or ExPEeRIMENTAL ZOOLOGY, VOL. VI, NO. I. 138 David Day Whitney ferent part of the ovary is not certain, but the evidence seems to indicate that they are male eggs. In the two kinds of parthenogenetic eggs of some of the aphids Stevens finds that there is no reduction in the number of the chromosomes during the formation of the polar body. Lenssen, in a study of the parthenogenetic eggs of the Rotifer, Hydatina senta, finds that there is a reduction in the number of the chromo- somes during the formation of the polar body in the male egg but no reduction in the female egg. Weismann found that both kinds of parthenogenetic eggs of Polyphemus (Daphnid), certain Ostracods and Rotifers produced only one polar body while the fertilized eggs produced two polar bodies. Blochmann and Stevens found the same relation to hold for certain aphids. In Lisparis dispar, a parthenogenetic Lepidopteran, Platner found that two polar nuclei were formed. In the bee apis, Bloch- mann, Paulcke, Petrunckewitsch and others find that the partheno- genetic eggs which develop into male animals give off two polar bodies. In the Rotifer, Asplanchna, Mrazek, Erlanger and Lauterborn found that the female parthenogenetic egg gave off one polar body and that the male parthenogenetic as well as the fertilized egg gave off two polar bodies. In the parthenogenetic eges of Hydatina senta Lenssen thought that the male egg gave off one polar body and the female egg gave off none! At the suggestion and under the supervision of Prof. T. H. Morgan the following work upon the eggs of Hydatina senta was done with the view of obtaining more light upon the maturation stages and their relation to the determination of sex. I am also indebted to Prof. E. B. Wilson for many valuable suggestions and criticisms. IT MATERIAL AND METHODS The Rotifers were collected and reared in cultures as described in a former paper. The first maturation spindle is formed before the egg is laid and in order to study the early maturation stages animals con- Maturation of Parthenogenetic and Sexual Eggs 139 taining eggs were killed and fixed in masses of thousands and sectioned in toto. Many eggs were found in the desired stages, but as the eggs are filled with yolk granules of various sizes it was exceedingly difficult to find many sections in which the yolk granules were not mingled with the chromosomes. Hot sublimate acetic, Bouin’s fluid, strong Flemming, Gilson, Carnoy, and alcohol acetic, were used as killing and fixing fluids. Some good preparations were obtained by each method, but alcohol acetic gave the best results in obtaining equatorial plates; for it coagulated the cytoplasm of the egg in such a way as to embed the yolk granules in its meshes, thus leaving the spindle and its chromosomes free from yolk granules. Thousands of animals were sectioned and about three hundred good slides were made. Sections were cut 5 1n thickness in 51° to 52° C. paraffine. Many parthenogenetic females were also isolated separately and the sex of their offspring determined, for those eggs first laid, before the females were killed and sectioned. The general nature of the maturation stages of such eggs was determined before a more detailed study was made of the eggs in the mixed slides. After the eggs are laid the envelope around them is so thin and at the same time so exceedingly impervious to fixing fluids that the eggs usually collapse in the process of fixation. Sometimes a few do not collapse in alcohol acetic but, however careful one may be, by the time the eggs are embedded they have shrunken. In such eggs the yolk granules are so crowded in among the chromosomes and stain so darkly that no satisfactory results can be obtained. In order to free the spindle from these granules the eggs were first centrifuged. In sections of such eggs the maturation spindle remained in the clear middle zone of the egg and was often entirely free from yolk granules. As only a few sections of these eggs were made no good stages were found in which the chromosomes could be counted but the method gives promise of results that can not be obtained in other ways. Heidenhain’s iron hematoxylin was used chiefly and gave the best results although many other stains were tried. In order to see the polar bodies the eggs, some time after they 140 David Day Whitney were laid, were put into Schneider’s aceto-carmine for about thirty seconds and then into a water-glycerine solution (1 drop in 5 cc. of water). [he blastomeres become separated and the polar bodies can be readily seen. III FEMALE EGG The female egg is easily distinguished from the male egg by its larger size and is never mistaken for the winter egg which may be of equal size, but has a much thicker envelope around it, besides containing the conspicuous sperm nucleus. In the female parthenogenetic egg the number of chromosomes was never definitely determined but many spindles in metaphase were seen in side view, containing numerous chromosomes (20 to @ ad a ae et , 3 ne |e as Pa B A @ Fig. 1 Female parthenogenetic egg. 4, equatorial plate of the polar spindle, showing twenty- three to twenty-five chromosomes; B, prophase of polar spindle, showing twenty-two chromosomes. 30). In one polar view of a metaphase twenty-five chromo- somes were seen, Fig. 1, 4. In a prophase twenty-two dumb-bel shaped chromosomes were seen in one section (Fig. 1, B) and in the adjoining section there were four other dumb-bell shaped chromosomes together with one that was not constricted. No anaphase or telophase stages were found although hundreds of eggs were examined. Lenssen found the chromosomes somewhat scattered about on the equator of the maturation spindle and con- cluded they were in an early anaphase but since he considered the unreduced number to be ten or twelve chromosomes the twenty or more chromosomes that he saw were probably in an early meta- Nore—The drawings of the chromosomes were made as carefully as possible with a camera under a 1.5 mm. Zeiss apochromatic and compensation ocular 6. They were then enlarged with a drawing camera about three times, corrected by comparison with the objects, and reduced by one-third in repro- duction. Maturation of Parthenogenetic and Sexual Eggs. 141 phase instead of in anearly anaphase. He never saw a telophase stage and decided without any evidence that the chromosomes never separated beyond the early anaphase stage and that later all the chromosomes form the segmentation nucleus. This is probably not the case because one polar body can always be seen near the periphery of the egg after the first cleavage, in total mounts prepared by the method already described. Some- times a constriction can be seen across the middle of the polar body giving it the appearance of being divided into two parts. In the two-cell stage of the egg after the two blastomeres separate the polar body is always found in the space between the two cells (Fig. 2, 4-B). In the four-cell stage it is seen at the point of junc- ture of the four cells (Fig. 2, C). y Fig. 2 Female parthenogenetic egg. 4, B, eggs in the two-cell stage, showing one polar body; C, egg in the four-cell stage, showing one polar body at the intersection of the two cleavage planes. LIN MALE EGG The male egg is much smaller than the other two kinds of eggs and has a thin envelope around it similar to that of the female parthenogenetic egg. The maturation spindle was seen several times when the chromosomes were in metaphase, anaphase and telophase stages. In two cases of telophase ten and fourteen chromosomes respectively were counted on one end of the spindle (Fig. 3, C-D). Polar views of the metaphase stage showed clearly eleven to thirteen chromosomes (Fig. 3, 4-B). ‘They were always less in number and larger in size than the chromosomes in the metaphase stage of the female parthenogenetic eggs. 142 David Day Whitney Three polar bodies are to be found near the periphery of the egg close to the line of meeting of the blastomeres. One was usually larger than the other two and often at a little distance away from them (Fig. 4, 4-B), although in one instance the three polar bodies were close together and seemed to be of the same size (Fig. 4, C). Wd '@ & eB TAN os ° e ° if boas 2 : e “x LP ‘ A B Gi D Fig. 3 Male parthenogenetic egg. 4, B, equatorial plates of the polar spindle, showing twelve to thirteen chromosomes; C, D, polar spindle in telophase, showing ten to fourteen chromosomes. Lauterborn states that in the male parthenogenetic egg of Asplanchna the first of the two polar bodies which was extended usually divided. Lenssen concluded that only one polar body was formed because he saw the maturation spindle in the telophase stage. He did not follow the history of the chromosomes in the later stages and conse- quently never saw any polar bodies. ( Fig. 4 Male parthenogenetic egg. A, B, eggs in the four-cell stage, showing three polar bodies, two of which are smaller than the other; C, egg in the two-cell stage, showing three polar bodies of nearly the same size. Vv WINTER EGG The fertilized or winter egg has a very thick envelope. An oval shaped small body which is probably the sperm nucleus is always found near the egg nucleus. The chromosomes were seen in sections on the maturation spindle (side view) in all stages but in only two anaphase stages (Fig. 5, Maturation of Parthenogenetic and Sexual Eggs 143 C-D), could they be counted because of being too closely crowded together. The polar view of the metaphase in the alcohol acetic fixation gave the best results. Fourteen chromosomes were seen in several sections of different eggs (Fig. 5, 4-B). The chromo- somes were of about the same size as those in the metaphase of the male parthenogenetic egg and were much larger in size and less in number than those in the metaphase of the female partheno- genetic egg. 6\ y @ i , j z{ a) legi\ * ere fj erik 4 i ae fy e@ @. 5 Hl Sie ii |e sole yey) @ eo oo ene RY gone urge oF, - @ oF \0 1; w! Lig! ~WXe ~eS ay is © 09 Se @® Fe WATE A B te Fig. 5 Winter or fertilized egg. A, B, equatorial plates of the polar spindle, showing fourteen chromosomes; C, D, anaphases of the polar spindle, showing twelve to fourteen chromosomes on each end of the spindle. On account of the thick and opaque egg envelope, the polar bodies were never seen. VI GENERAL DISCUSSION Although Lenssen was mistaken in regard to the number of the chromosomes nevertheless he was firmly convinced that the num- ber in the maturation stages of the male parthenogenetic and the fertilized egg were the same and that the number in the female parthenogenetic egg was greater. By comparing my Figs. 1, 3 and 5, it will be seen that this conclusion is confirmed. The greatest number of chromosomes seen in an equatorial plate of the male egg was possibly thirteen and the number seen in an equatorial plate of a winter egg was fourteen. ‘The chromosomes of both eggs in the same stages were of the same size. In the female parthenogenetic egg the greatest number of chro- mosomes seen was twenty-five (Fig. 1). “The chromosomes were 144 David Day Whitney very much smaller than in the other two kinds of eggs and usually were so crowded together that it was impossible to count them except in a very few cases. These observations show that there is probably a reduction in the number of chromosomes in the male parthenogenetic and winter egg but no reduction in the female parthenogenetic egg. The former case would be similar to what occurs in the honey- bee. In the aphids Stevens found that there is no reduction in the number of chromosomes in either of the male or female par- thenogenetic eggs but only in the fertilized egg. It appears that in different animals parthenogenetic eggs vary in the number of polar bodies that they give off. The male egg of Asplanchna, Hydatina and Apis gives off two polar bodies while the male egg of aphids gives off only one. If it is true that the male egg when fertilized becomes the winter ege which develops into a female it seems evident that the reduction in the number of chromosomes and the formation of the second polar body is not in itself the factor that determines the ultimate sex of the ege. The sperm would seem to introduce a factor that determines the sex of the embryo. This idea is strongly suggested by the evidence that Meves has brought forward in the case of the honey- bee in which he finds that only one kind of functional sperm is produced. Morgan also finds a similar phenomenon for certain Phylloxerans. If the same process occurs inthe sperm of Hydatina the cause of the change in sex of the male egg may be at least surmised. Vil SUMMARY 1 In the female parthenogenetic egg of Hydatina senta there is no reduction in the numberof chromosomes during maturation. One polar body is extruded. 2 Inthe male parthenogenetic egg there is a reduction in the number of chromosomes during maturation. Two polar bodies are formed, one of which subsequently divides. 3 In the winter egg, that becomes fertilized, there is a reduc- tion in the number of chromosomes during maturation, and since Maturation of Parthenogenetic and Sexual Eggs 145 a similar process of reduction takes place in the parthenogenetic egg that becomes a male it would seem to follow that the sex of the embryo from this egg is changed by the spermatozo6n. Zoélogical Laboratory Columbia University April 25, 1908 LITERATURE CITED Bavsiant, E. G. *69—72—Meémoire sur la génération des aphides. Am. Sc. Nat. per. 5; Zool., Urs, 809; ls rane 7Os bays 1h 72. Biocumann, F. ’88—Ueber die Richtungskérper bei unbefruchtet sich entwick- elnden Insekteneiern. Verh. naturh. med. Ver. Heidelberg, N. F., vol. iv, no. 2. ’8g9—Ueber die Zahl. der Richtungskérper bei befruchteten und unbe- fruchteten Bieneneiern. Verh. naturh. med. Ver. Heidelberg, IN: F., vol. iv, pp. 239-41; fr. R. Mic. Loc.; 1889: Brauer, A. ’94—Zur Kenntniss der Reifung des parthenogenetisch sich entwick- elnden Eies von Artemia salina. Arch. mikr. Anat., vol. xi. CastLe, W. E. ’03—The Heredity of Sex. Bull. of the Mus. of Comp. Zodl. Harvard College. vol. xl, no. 4. Dz1erzon, J. ’45—76—[For a complete list of the writings of Dzierzon, see Biblio- theca Zoologica, ii, p. 2270.] ERLANGER U. LAUTERBORN ’97—Ueber die ersten Entwickelungsvorgange im parthenogenetischen und _ befruchteten Raderthierei. Zool. Anz, Vol. xX: IssakowitscuH, A. ’05—Geschlechtsbestimmende Ursachen bei den Daphniden. Biol. Centralb., xxv. LauTerzorn, R. ’98—Ueber die cyclische Fortpflanzung limnetischer Rotatorien. Biol. Centralbl., xviii. LENssEN ’98—Contribution a l'étude du developpement et de la maturation des ceufs chez Hydatina senta. La Cellule, xiv. Maupas, M. ’g1—Sor le déterminisme de la sexualité chez |’Hydatina senta, Ehr. CR: Ac sce Paris, exit Meves, F. ’07—Die Spermatocytenteilungen bei der Honigbiene. Arch. mikr. Anat., lxx. Moraean, T. H. ’08—The Production of Two Kinds of Spermatozoa in Phylloxerans. Functional “Female Producing” and Rudimentary Sperma- tozoa. Proceedings of the Society for Experimental Biology and Medicine, vol. v, no. 3. 146 David Day Whitney Mrazek, At. ’97—Zur Embryonalentwickelung der Gattung Asplanchna. Jahresb. bohna Ges., 2, pp. I-II. Nussspaum, M. ’97—Die Entstehung des Geschlechts bei Hydatina senta. Arch. mikr. Anat., xlix. Pautcke, W. ’99—Zur Frage der Parthenogenetischen der Drohnen. Anat. Anz., vol. xvi. PerruNKeEwiItscu, A. ’o1—Die Richtungskorper und ihr Schicksal im befruch- teten und unbefruchteten Bienenei. Zool. Jahrb., vol. xiv. Puituies, E. F. ’°03—A Review of Parthenogenesis. Amer. Phil. Soc., vol. xlii. Piatner, G. ’88—Die erste Entwickelung befruchteter und parthenogenetischer Eier von Lisparis dispar. Biol. Centrabl., vol. viii, no. 17. ’89—Ueber die Bedeutung der Richtungskérperchen. Biol. Centrabl. vol. viil. Stevens, N. M.’04—A study of the Germ Cells of Aphis rosze and Aphis cenotherz Journ. Exp. Zool., vol. 1. WEISMANN AND IscHIKAwa '88—Weitere Untersuchungen zum Zahlengesetz der Richtungskérper. Zool. Jahrb., vol. 11. Weismann, A. ’80—Parthenogenesis bei den Ostracoden. Zool. Anz., iil. Wuirtney, D. D. ’07—Determination of Sex in Hydatina senta. Journ. Exper. Zool., vol. v. Wison, E. B., °05—Studies on Chromosomes. Jour. Exp. Zodl., vols. ii and ii. STUDIES ON CHROMOSOMES V THE CHROMOSOMES OF METAPODIUS. A CONTRI- BUTION. TO THE HYPOTHESIS OF THE, GENEDIC CONTINUITY OF CHROMOSOMES’ BY EDMUND B. WILSON Wirn One Prate AND THIRTEEN FIGURES IN THE TEXT The genus Metapodius (Acanthocephala), one of the coreid Hemiptera, shows a very exceptional and at first sight puzzling relation of the chromosome-groups which has seemed to me worthy of attentive study by reason of its significance for the hypothesis of the “individuality” or genetic continuity of the chromosomes. The most conspicuous departure from the relations to which we have become accustomed lies in the fact that different individuals of the same species often possess different numbers of chromo- somes, though the number in each individual is constant. An even more surprising fact is that in all of my own material every male individual possesses at least 22 spermatogonial chromosomes, including a pair of unequal idiochromosomes like those of the Pentatomidz, while in Montgomery’s material of M. terminalis every male has but 21 spermatogonial chromosomes, one of which is a typical odd or “accessory” chromosome (unpaired idiochro- mosome).? The present paper presents the results of an investigation of these relations that has now extended over nearly four years, in the course of which serial sections of more than sixty individuals 1 Part of the cost of collecting and preparing the material for this research was defrayed from a grant of $500 from the Carnegie Institution of Washington, made in 1906. I am indebted to Rev. A. H. Manee, of Southern Pines, N. C., for valuable codperation in the collection of material, and to Dr. Uhler, Mr. Heidemann, Mr. Van Duzee, and Mr. Barber for aid in its identification. 2 By Professor Montgomery’s courtesy I have been enabled to study thoroughly his original prep- arations and to satisfy myself of the correctness of his account (Montgomery ’o6). I also owe to him a number of unsectioned testes of the same type. Tue JouRNAL or ExPERIMENTAL ZOOLOGY, VOL. VI, NO. 2. 148 Edmund B. Wilson have been carefully studied. ‘These individuals belong to three well marked species—M. terminalis Dall. and M. femoratus Fab. from the Eastern and Southern States, M. granulosus Dall. from the Western—all of which show a similar numerical variation.’ My first material, including sections of two testes of M. terminalis (Nos. 1, 2) from the Paulmier collection, long remained a complete puzzle and led me to the suspicion that the material was patho- logical. This possibility was eliminated by the study of additional material of the same type; but the contradiction with Montgom- ery’s results on the same species suggested that his specimens were not correctly identified (Wilson ’o07a). Continued study at length convinced me that this supposition too was probably un- founded. If the identification was correct, as I now believe it was, M. terminalis is a species that varies not only in respect to the individual chromosome number but also in respect to the sex- chromosomes, certain individuals having an unpaired “accessory” chromosome, while others have an unequal pair of idiochromo- somes. ‘The latter condition alone has thus far been found in M. femoratus and M. granulosus. ‘The essential facts, and the general history of the spermatogenesis, are otherwise closely similar in the three species. The range of variation in the number of chromosomes is in M. terminalis from 21 to 26,in M. femoratus from 22 to 27 or 28, and in M. granulosus from 22 to 27, the particular number (or its equivalent in the reduced groups) being a characteristic feature of the individual in which it occurs. I do not mean to assert that there is absolutely no fluctuation in the individual. In this genus, as in others, apparent deviations from the typical number fre- quently are seen, and real fluctuations now and then appear; but the latter are so rare that they may practically be disregarded. That the number may be regarded as an individual constant (subject to such deviations as are hereafter explained (p. 185) is abundantly demonstrated, not only by the agreement of large numbers of cells from the same individual but perhaps even more 3 A complete list of the individuals examined, arranged by localities, is given in the Appendix at p.-202, Each individual is there designated by a number by which it is referred to in the text and description of figures. Studies on Chromosomes 149 convincingly by the definite correlation of the spermatocyte- groups with those of the spermatogonia of the same individual. This is shown in the following table, which summarizes the facts thus far observed.‘ SUMMARY Somatic aes First spermatocyte terminalis | ~—‘femoratus granulosus So meaopont or fitigicon | | | | ovarian cells) CE ed let oa a glee Ps apis Poe | | | i, wary PNT ald (OPORTO OR ROE RRR a | Il 9) | ° ° ° ° ° DO eters fase. siche = ateyoiein alts 12 3 4 3 fc) I ° DBR laters tetee) Semone elaves 13 5 2 ° 2 2 fe) 271 CCRC CREE OLEH 14 | 3 3 2 I 4 fo) Dees 5.ossierailaei de vrataieis steic| 15 2 2 ° ) I I OLR ry AB ccialetecaieiete /aajer'= 16 I ° 24 | I 4 2 Garesocewee css cued 17 ° ) opened I ° 13( 0 )10) Yaoriagr SAC Reroe fe) fo) I co) ) = = Distribution in the whole group Total somatic number Number of males Number of females Totals EAN PRPs asters ena) Ss beaters 9 ° 9 05 oS I EEE TCC erie 7 4 Il MloooddosansousokeobecansoGcr 7 4 | II 2h ES A neces oy OETA SR ee 9 4 13 DEsoscobsvesoHpbovosuDDHOboes a 3 | 6 7A ACC ECCI OI CE ROC RCO CE 7 3 IO (Ci) Sogemea dncenenernsser ands I ° I 1S (Crap eis Gran seein coors ) I I ANo\ital lle icant Oe occ 43 19 62 4 The somatic numbers of the males are in each case determined from the dividing spermatogonia. Those of the female are from dividing cells in various parts of the ovary—mainly from the region just ‘above or below the end-chamber—some of them undoubtedly folicle-cells, others probably young nutri- tive cells or odgonia. The chromosome-groups from different regions differ considerably in size, but otherwis show the same general characters. With a very few exceptions the number of chromosomes has been determined by the count of several groups from the same gonad, in many cases by the count of a very large number. In many individuals hundreds of perfectly clear equatorial plates may be seen and the evidence is entirely demonstrative. In seven of the males (owing to lack of mitoses, or to defec- tive fixation) the somatic number has been inferred from that shown in the spermatocyte divisions, or vice versa; but with a single exception both numbers have been directly observed in other individuals of the same type. Iam therefore confident that the numbers are substantially correct as given. In case of the female, only the somatic numbers can be given, since the maturation-divisions are not available for study. 150 Edmund B. Wilson The material of terminalis is from New Jersey, Pennsylvania, Ohio, North Carolina, South Carolina and Georgia; that of femor- atus from the three states last named; that of granulosus from Arizona. ‘The variation of number 1s independent of locality, and individuals of the same species showing different numbers were often taken side by side on the same food plants. It is equally independent of sex, as the table at once shows. I am unable to find any constant correlation between the number of chromosomes and any other visible structural characters of the adult animals. Such an astonishing range of variation in the chromosome num- ber in the same species seems at first sight to present a condition of chaotic confusion. But, as I shall endeavor to show, the first impression thus created disappears upon more critical examination. Detailed study of the facts proves that the variation 1s not indis- criminate but affects only a particular class of small chromo- somes that are, distinguishable from the ordinary ones both by size and by certain very definite peculiarities of behavior. ‘These chromosomes are absent in all of Montgomery’s material; in my own they are sometimes present, sometimes absent, the total num- ber varying accordingly. [he chromosomes in question are the ones which in earlier papers I have called the “supernumeraries. ’’ In behavior they show an unmistakable similarity to the idiochro- mosomes; and for reasons given beyond I believe them to be noth- ing other than additional small idiochromosomes, the presence of which has resulted from irregularities of distribution of the idio- chromosomes in preceding generations. ‘The relations seen in Montgomery’s material form the converse case, the small idio- chromosome having disappeared or dropped out. I shall try to show that both cases are probably due to the same initial cause. 5 Wilson ’07a, ’o7b. I first discovered this phenomenon in the pentatomid species Banasa calva (osb) describing the single supernumerary as a ‘‘heterotropic chromosome.” Later (’o7a) a single supernumerary was found in certain individuals of Metapodius terminalis, and other numerical varia- tions in this species and in femoratus and granulosus were briefly recorded; but at that time I did not yet fully understand the facts. Banasa calva is the only form oustide the genus Metapodius, in a totalof more than seventy species of Hemiptera I have examined, in which supernumerary chromosomes have been found. Miss Stevens (’o8b) has recently found in the coleopteran genus Diabrotica a condition that is in some respects analogous to that seen in Metapodius. Studies on Chromosomes I51 A GENERAL DESCRIPTION Since the phenomena as a whole are somewhat complicated, I have thought it desirable to bring the most essential facts together for ready comparison in a preliminary general account illustrated by a limited number of selected figures (Figs. 1, 2). The funda- mental type of the genus is, I believe, represented by individuals that possess 22 chromosomes in the somatic groups of both sexes, and in which no supernumeraries are present (Fig. 1, d-7/). “Two of the chromosomes are a pair of very small m-chromosomes, like those of other coreids; two are a pair of idiochromosomes consist- ing in the male of a large and a small member, in the female of two large ones; while the remaining 18 are ordinary chromosomes or “autosomes.”” These chromosomes have in the spermato- genesis the same general history as in other Hemiptera heteroptera. In the first division the idiochromosomes are separate univalents, their position being typically (but not invariably) outside a ring formed by the nine larger bivalents within which lies the small m-chromosome bivalent (Fig. 1, d, Photo 2). This division accordingly shows 12 separate chromosomes (one more than the reduced or haploid number.) In the second division, as des- cribed beyond, they are always united to form a dyad or bivalent, composed of two unequal halves, and the number of separate chromosomes is 11. The spermatogonial groups possess 22 chro- mosomes (Fig. 1, ¢) of which the small idiochromosome may often be recognized as the smallest of the chromosomes next to the m-chromosomes; but it does not differ sufficiently in size fromthe other chromosomes to be always certainly distinguishable.* In the growth period the idiochromosomes, as usual, have the form of condensed deeply-staining chromosome-nucleoli, while the other chromosomes are in a vague, faintly staining condition. They are usually in contact but not fused (Fig. 1, 7, Photo 25), thus form- 6 In considering the relative size-relations it is important to bear in mind that the apparent size, as seen in polar view, varies considerably with the degree of polar elongation. Still more important is the fact (which I have emphasized in a preceding paper) that in the first division univalent chromosomes always appear relatively much smaller than they do in the spermatogonia. This is the case with the idiochromosomes and the supernumeraries, which are always readily recognizable in the spermatocyte- divisions, but are often difficult to distinguish in the spermatogonia. 152 Edmund B. Wilson EXPLANATION OF FIGURES Fic. 1 About one-fourth of the figures were drawn upon enlarged photographs by the method described in a preceding paper (Wilson ’og). The others are from camera lucida drawings. In all cases the form, size, and grouping of the chromosomes are represented as accurately as possible. The form, size, and general appearance of the spindles are shown, but no attempt has been made to represent the exact details of the fibrille. Figs. 1 and 2 are enlarged about 3300 diameters, the others a little less than 3000 diameters. Lettering, in all the Figures I, large idiochromosome or odd chromosome; /, small idiochromosome; m, m-chromosome; p, plas- mosome; s, supernumerary chromosome. In cases where s and / are both present and of equal sizeit is impossible to distinguish between them. In such cases I have as a rule designated as 7 the one lying nearest to J; but this is quite arbitrary. It should be noted also that J cannot always be distinguished from the smaller of the ordinary bivalents. j Studies on Chromosomes 153 Fic. 1 M. terminalis a-c (No. 3), 21-chromosome form; a, first spermatocyte metaphase; b, spermatogonial metaphase; ¢, nucleus from the growth period. d-f (No. 19), 22-chromosome form, stages corresponding to above. g-i (No. 20, Photo 4), 23-chromosome form, one large supernumerary. jl (No. 43), 23-chromosome form, one small supernumerary. 154 Edmund B. Wilson ing a very characteristic bipartite body; but in a good many cases they are separate (Fig. 6, c,d, Photo 26). A large and very dis- tinct plasmosome is also present. ; Such a group of 22 chromosomes may be regarded as the type of which all the other forms may be regarded as variants, and probably as derivatives. In forms having more than 22 chromo- somes the increase in number is due to the presence of from one to six supernumeraries. [hese vary in number and size in different individuals, but both are constant in a given individual. Their maximal size is equal to that of the small idiochromosome (in which case they are indistinguishable from the latter); such forms will be called “large supernumaries.” ‘Their minimal size, (“small supernumeraries’’) is about the same as that of the m- chromosomes; but from the latter they are always distinguishable, in the male, by a quite different behavior in the maturation pro- cess. When a single supernumerary is present it may be either large or small, its size being (with slight variation) constant in the individual. When more than one is present all may be of the same size (the most usual condition) or they may be of different sizes, the relation being again an individual constant. Whatever their number or size their behavior is essentially the same as that of the idiochromosomes. In the growth-period they have a con- densed form and are typically united with the idiochromosomes to form a compound chromosome-nucleolus, the components of which are often distinctly recognizable and vary in number with the number of the supernumeraries. In the first division they divide as separate univalents, and this division accordingly shows as many chromosomes above 12 as there are supernumeraries— i.e., 1f the spermatogonial number be 22 + , the number in the first division is typically 12 + ». ‘Their typical position in this division is, like that of the idiochromosomes, outside the ring of larger bivalents, though there are many exceptions. In the sec- ond division they are, as a rule, again associated with the idio- chromosomes to form a compound element, though not infre- quently one or more of them may be free from the others. A definite correlation thus appears in each individual between the number and relative sizes of the chromosomes seen in the Studies on Chromosomes 155 maturation-divisions and in those of the spermatogonia; and it also appears in the number and size of the components of the chromosome-nucleoli when these can be distinctly recognized. Figs. 1 and 2 illustrate this correlation and epitomize the most essential facts. [These figures have been selected from a much larger number to show the clearest and most typical conditions. Some of them are enlarged from the photographs reproduced in Plate I. Many others, with an account of secondary variations, are given beyond. Each horizontal row of figures represents three stages of the same type which, with two exceptions, are all from the same individual. ‘The left hand figure in each row shows the typical arrangement of the chromosomes in the metaphase of the first spermatocyte-division, the middle figure a spermato- gonial group, and the right hand one a nucleus from the growth period, to show the chromosome-nucleolus together with some of the diffused ordinary chromosomes. Fig. 1, a—-c (terminalis, No. 3), represent these three stages in an individual of the 21-chromosome type (Montgomery’s material) showing 11 chromosomes in the first division, 21 in the sper- matogonia, and a single chromosome-nucleolus in the growth period. (Additional figures of this individual in Fig. 3.) Fig. 1, d-f (terminalis, No. 19), show the 22-chromosome type, with a small idiochromosome present in addition to the large one. The small idiochromosome (z) is distinguishable in Fig. 1, e. (Additional figures in Figs. 4-6.) Fig. 1, g-z (terminalis, No. 20), show the 23-chromosome type, with one large supernumerary. In the spermatogonial group (h) this chromosome and the small idiochromosome are probably rep- resented by the two designated asz and s. ‘The nucleus from the growth-period (7), shows the plasmosome (p) and a tripartite chromosome-nucleolus formed by the idiochromosomes and the supernumerary attached in a row (cf. Photo 27; additional figures in Figs. 7-8). Fig. 1, ;-/ (terminalis, No. 43), show a 23-chro- mosome group with one small supernumerary. This clearly appears in the spermatogonial group (s); and the small idiochro- mosome (7) 1s also distinguishable. In the nucleus from the growth-period (/), the supernumerary and small idiochromosome 156 Edmund B. Wilson are united (7, s) the large idiochromosome (/) being separate. (Additional figures in Figs. 7, 8.) Fig. 2, a-c (terminalis, No. 21), show the corresponding stages in an individual of the 24-chromosome type, with two large super- numeraries. ‘Their identification in the spermatogonial group is somewhat doubtful. (Additional figures in Fig. ro.) Fig. 2, d, e (terminalis No. 34), show a 25-chromosome type with three large supernumeraries. ‘The growth-period (f) 1s from an individual of granulosus (No. 54) that is possibly of the 26- chromosome type. (Additional figures in Fig. 12.) Fig. 2, g, h (femoratus No. 42), and 7 (granulosus, No. 60) show the 26-chromosome type with four large supernumeraries. (See Photo. 28, additional figures in Figs. g, 10.) Fig. 2, 7-1 (femoratus, No. 40), are from a very interesting indi- vidual of the 26-chromosome type, with two large and two small supernumeraries (additional figures in Figs. 9, 10). “The sperma- togonia of this individual (&) uniformly show 26 chromosomes, including four very small ones (two m-chromosomes, two small supernumeraries), but the large supernumeraries and the small idiochromosomes are doubtful. No case was found in which all of the six components of the chromosome-nucleolus could be seen; 1 shows five of them, including the two small ones. B ADDITIONAL DESCRIPTIVE DETAILS I will now give a somewhat more detailed and critical account of the facts. Taken as a whole, the series (including nearly 300 slides of serial sections) presents a profusion of evidence on many cytological questions that could not be adequately described save in a large monograph; but I will here limit the account mainly to the numerical and topographical relations of the chromosomes. The clearness of the preparations is such that nearly all the prin- cipal phenomena might have been illustrated by photographs (of which upwards of 200 have been prepared). ‘Thirty of these are reproduced in Plate I, less for the purpose of giving ;the evidence in detail than of illustrating its character to those not directly familiar with this material. Studies on Chromosomes @ © ¢i Bhs) Ge ; h oY eth rae @ oe: @ | u 7), k ErGez a-e, M. terminalis; f, 7, granulosus; g—h, j-/, femoratus. a-c (No. 21), 24-chromosome form, two large supernumeraries. d-e (No. 34), 25-chromosome form, three large supernumeraries. f (No. 54), growth-period, 25- or 26-chromosome form. g-h (No. 42 Photo 8), 26-chromosome form, four large supernumeraries. i (No. 60), 26-chromosome form, growth-period. j-! (No. 40), 26-chromosome form, two large and two small supernumeraries. 158 Edmund B. Wilson 1 Individuals having twenty-one spermatogonial Chromosomes, including an unpaired Idiochromosome. Small Idiochromo- some and Supernumeraries absent To this group belong only the specimens, all males, collected by Montgomery at West Chester, Pa., of which I have examined nine individuals, all of which have essentially the same characters.? Montgomery (’or) originally described these forms as having 22 spermatogonial chromosomes but subsequently (’06) corrected this to 21, describing the phenomena as agreeing in all essential respects with those seen in Anasa and other coreids. A study of the orig- inal preparations has enabled me to confirm this later account in every essential point. After the synizesis or contraction phase of synapsis (as in all individuals of the genus) the ordinary chromo- somes appear in the form of rather delicate spireme-like threads, longitudinally split. In later stages of the growth-period they shorten, become irregular, lose their staining capacity, and assume the vague, pale condition characteristic of so many other forms. In the early prophases of the first division they become more defi- nite, stain more deeply, and appear as coarse longitudinally split rods that often show an indication of a transverse division at the middle point, or in the form of the double crosses as described by Paulmier in Anasa (’99). In the later prophases they condense still further to form nine compact bivalents which finally arrange themselves in a more or less regular ring. ‘The equatorial plate of the first division always shows in polar view 11 chromo- somes (Fig. 3, a,b, Photo 1). In the most typical case the univa- lent idiochromosome lies outside this ring, but it sometimes lies in or inside it. “The small m-chromosome bivalent is always near the center of the ring. In side view the larger bivalents are either dumb-bell shaped or more or less distinctly quadripartite, in the 7 These were taken from magnolia trees. In the summer of 1907 I collected in the same locality two males and three females, all from blackberry bushes. To my disappointment, these differ from Montgomery’s specimens, one male having 22 spermatogonial chromosomes, the other 23; while the ovarian cells have in one female 23 and in the other two 24 chromosomes. It is possible that a different species fell into Montgomery’s hands, perhaps an introduced form; but both the structure of the testis and the character of the chromosome-groups agree so exactly with my own material that I now believe that Montgomery’s identification was probably correct. Studies on Chromosomes 159 latter case appearing dumb-bell shaped as seen in polar view. The eccentric idiochromosome is of nearly the same size as the smallest of the large bivalents and is often indistinguishable from the latter except by its position. All these chromosomes divide equally in this division, the m-chromosomes usually leading the way in the march towards the poles, while the idiochromosomes often lag slightly behind the others. The second division likewise shows 11 chromosomes in polar view (3, c, d); but the regular grouping characteristic of the first division 1s now usually lost, the ring formation being often no longer apparent, while either the m-chromosome or the idiochro- mosome may now occupy any position.’ In this mitosis all the chromosomes divide except the idiochromosome which lags behind the others and finally passes undivided to one pole (Fig. 3, e-A, Photos 14, 15) as Montgomery described. The nucleus formed at this pole thus receives 11 chromosomes, the sister nucleus but 10, precisely as in Anasa, Narnia, Chelinidea or Leptoglossus. This is proved beyond all doubt by polar views of the anaphases, showing the sister groups lying one above the other in the same section (Fig. 3, 4). In the particular example figured the idio- chromosome lies eccentrically, but this is quite inconstant. The spermatogonia (Fig. 3, 7, 7) always show 21 chromosomes, a largest and a smallest pair iiee always distinguishable. ‘The unpaired idiochromosome cannot be distinguished from the others. The m-chromosomes are usually equal, but sometimes appear slightly unequal. In the growth-period the m-chromosomes and the idiochromo- some have the same history as in other coreids. ‘The former are 8 The regrouping of the chromosomes in the second division, first described by Paulmier (’99) in Anasa tristis, is characteristic of the Coreide generally, an eccentric position of the idiochromosome being a nearly constant feature of the first division but not of the second. Failure to recognize this fact in the case of Anasa tristis seems to have been one of the main sources of error in the entirely mis- taken conclusions of Foot and Strobell (’o7a, ’07b) regarding this species. (Cf. Lefevre and McGill, ’o8.) Demonstrative evidence on this point is given by polar views of rather late anaphases in which every chromosome of each daughter plate may be seen in the same section. Such views, of which I have studied many, both in Anasa and in other genera, show that oneof the chromosomes may indeed occupy an eccentric position, and may there divide; but in such cases the odd chromosome is always found elsewhere in the group, lying either in or near one of the daughter-groups and not in the other. When the odd chromosome is eccentric it is found in one of the daughter groups but not in the other. 160 Edmund B. Wilson typically separate, and at first diffuse (as in Anasa or Alydus). Later they condense to from two spheroidal bodies that conjugate in the late prophase to form the central small bivalent and are almost immediately separated again by the division. The idio- & e @ 2° @ e @ e $ eo. a eo °e Pd e. TY 34 2S Sam Fic. 3 M. terminalis (Montgomery’s material (Nos. 3-11), 21-chromosome form) a, b, first division, polar view (Photo 1); c-d, second division; e, f, g, side views of second division (Photos 14, 15); 4, sister-groups from the same spindle, in one section, anaphase second division, one showing 10 chromosomes the other 11. i-], Spermatogonial groups, 21 chromosomes; k-/, early and late growth-period. chromosome has throughout the early and middle growth-period the form of a single spheroidal or ovoidal intensely staining chro- mosome-nucleolus, which shows in brilliant contrast to the other chromosomes (Fig. 3, k, /, Photo 24). This body is sometimes slightly constricted in the earlier period. Later it is always con- Studies on Chromosomes 161 stricted, assuming the bipartite form in which it enters the equa- torial plate to form the eccentric chromosome. ‘Throughout the growth-period a large plasmosome is also present, usually separate from the chromosome-nucleolus. In properly stained sections these two bodies differ so markedly in staining reactions that they cannot for a moment be confused. In haematoxylin preparations the chromosome-nucleus is intensely black, the plasmosome pale yellowish, bluish or gray. In Montgomery’s safranin-gentian preparations (though now somewhat faded) the former is bright _ red, the latter bluish or nearly colorless. There are no females in Montgomery’s material; but in view of the relations known in many other related forms it may safely be concluded that the 11-chromosome spermatozoa are female-pro- ducing, and that the female somatic number in this race Is 22. 2 Individuals with twenty-two Chromosomes in the somatic Groups of both Sexes including a pair of unequal Idiochromosomes in the Male, and a Pair of equal large ones in the Female This condition has been found in seven males and four females, all three species being represented. The three species closely agree in all the phenomena. To the males of this type precisely the same description applies as to the foregoing case except ae a small idiochromosome 1 1s present in addition to the “odd” or “accessory” chromosome. The latter 1s now indistinguishable an a “large idiochromosome, ”’ and the identity of these two forms of chromosomes, on which I haye laid stress in former papers, is thus fully demonstrated. ‘This appears most clearly in the maturation divisions. In the first division the chromosomes show the same grouping as in the 21- chromosome forms, but a small idiochromosome accompanies the “accessory,” frequently lying beside it outside the principal ring, though sometimes being in or inside the latter (Fig. 4, a-7, Photos 2, 3). This chromosome is always recognizable as the smallest of all the chromosomes except the 77-chromosomes, and it is in general about half the size of the large idiochromosome or slightly less. All the chromosomes now divide equally (Fig. 4, /, 162 Edmund B. Wilson Photo 11), 12 chromosomes passing to each pole. The second division immediately follows without the intervention of a “resting stage,’’ and the chromosomes undergo the same regrouping as that described for the 21-chromosome forms. As this takes place, the two idiochromosomes conjugate to form an unequal bivalent (precisely as in Lygaeus or Euschistus); so that when the equato- rial plate reforms but 11 (instead of 12) chromosomes appear in polar view (Fig. 5, a-c, Photo 12). The idiochromosome-biva- lent now usually lies near the center of the group (contrasting with the first division), and the m-chromosome is usually not far from it. Such views are almost indistinguishable from those of the 21-chromosome individuals, since the small idiochromosome is covered by the large one and only appears in side view. In the course of the division the idiochromosome bivalent separates into its two components, which pass to opposite poles, while all the other chromosomes divide equally. The idiochromosomes at first separate more rapidly than the other daughter-chromosomes(Fig. 5,7» 2), as in other genera, but as the division proceeds the reverse condition prevails, so that the two idiochromosomes are seen lag- ging on the spindle between the diverging daughter groups (Fig. 5 i-l). In the later stages one passes to each pole. ‘There is much variation in this process. Often the two move at the same tate so that in the late anaphases one may be seen entering each pole (Fig. &, /, Photo 17). Not uncommonly, however, one or the other lags behind upon the spindle (usually the large one, though Fig. 5, ;, shows the reverse case) giving a condition that exactly resembles that seen in the 21-chromosome forms (Fig 5, m, n), but earlier anaphases in the same cysts at once show the difference. It is no less conclusively shown by polar views of the late anaphases, in which each daughter-group is seen to consist of 11 chromosomes, ten of which are duplicated in the two while the the eleventh is in one case the large, in the other the small idio- chromosome (Pig. 5.09507 15502): The difference between the two types is shown with almost equal clearness by the chromosome-nucleoli of the growth-period. In the 21-chromosome type, as already stated, this body is single. A similar appearance is sometimes given in the 22-chromosome indi- Studies on Chromosomes 163 Fic. 4 22-chromosome forms a-l, first division; a,b, term. No. 19, typical (Photo 2); c, term. No. 12 (Photo 3); d, e, fem. No. 29 f. term. No. 12; g, k, term. No. 19, idiochromosomes united; i, fem. No. 29, same condition; j, gran. No. 47; k, fem. No. 29, first division, side view, idiochromosomes united; /, fem. No. 46 (Photo 11), first divi- sion, anaphase, division of both idiochromosomes. m-q, spermatogonial groups; m, term. No. 19; 1, 0, fem. No. 46; p, q; fem. No. 29. r-t, ovarian groups; r, term. No. 24:5, term. No. 44;1, term. No. 23, exceptional form and grouping. 164 | Edmund B. Wilson ». viduals, owing to close union of the two idiochromosomes. But in very many cells of this period the chromosome-nucleolus con- sists of two very distinct unequal moieties, in contact (Fig. 6, a, b, Photo 25), or not infrequently widely separated (Fig. 6, c, d. Photo 26). When in contact they form a double body closely similar to the 1diochromosome-bivalent of the second division. There can be no question of confusing either of these bodies with the plasmosome, since the latter, showing its characteristic stain- ing reactions, is also present. In the late prophases of the first division the idiochromosomes, if previously united, almost invariably part company to divide as separate univalents, as in other Hemiptera; but they usually remain near together outside the principal ring. Only very excep- tionally do they divide together. The spermatogonial groups (Fig. 4, m-q) uniformly show 22 chromosomes, and in some cases the small idiochromosome may be recognized by its small size (m, g). This is, however, not nearly so marked as in the first division, since 1t now appears rela- tively twice as large, owing to the univalent character of the other chromosomes, and often it cannot certainly be distinguished from the smaller of these (7, p). These facts make it clear that if the small idiochromosome be supposed to disappear, the entire series of phenomena would be- come identical with those shown in the 21-chromosome individuals, the large idiochromosome now appearing as the odd or “‘acces- sory’? chromosome. The unreduced female groups of this type (ovarian cells) are closely similar to those of the male (Fig. 4, rt) but a small idio- chromosome can never be distinguished. ‘The absence of this chromosome cannot be so convincingly shown in Metapodius as in such forms as Lygaeus or Euschistus, owing to its greater rela- tive size. Nevertheless, after the detailed study of many female groups I am convinced that this chromosome is not present, and that all the chromosomes may be equally paired. Apart from analogy, therefore, I think the conclusion reasonably safe that in Metapodius, as in other forms, the unequal 1diochromosome- pair of the male is represented in the female by a large equal pair, Studies on Chromosomes 165 e @ \\ @ @ & i Ny i]! ° A e (>) Pt ar fi IN Al ® ® i aa aas Oe af Fic. 5 22-chromosome forms a-c, second division, polar view; a, fem. No. 19; b, fem. No. 28; c, gran., 47 (Photo 12). a-p, second division, side view; d-h, fem. No. 29, metaphases, separation of idiochromosomes; Lope term. No. 19, anaphases, lagging of one idiochromosome; k-m, gran., No. 47, late anaphases (Photo 17); n, term., No. 19, late anaphase, lagging large idiochromosome; 0, fem., No. 46, exceptional condi- tion, both idiochromosomes passing to one pole (Photo 18); p, term. No. 19, similar form; q, r, term., No. 19, sister anaphase groups, from the same spindle; s, t, fem., No. 29, the same. 166 Edmund B. Wilson 4 and that, accordingly, the usual rule holds in regard to fertiliza- tion. . Exceptional conditions. ‘There are two conditions, rarely seen, that are of interest for comparisons with other species. Now and then the idiochromosomes fail to separate for the first division, but remain in more or less close union to form an asymmetrical bivalent, which in side view is seen to form a tetrad (Figs. 4, i, k, Photo 3). This bivalent undergoes an equation division, in this respect agreeing with the conditions uniformly seen in Syro- Fic. 6 M. femoratus (No. 29) 22-chromosome form Four nuclei from growth-period showing diffused ordinary chromosomes, condensed chromosome- nucleoli and plasmosome; in a and b the two idiochromosomes are united to form double chromosome- nucleoli (Photo 25); in c and d they are separate (Photo 26). mastes (Gross ’04, Wilson ’og), and differing from that occurring in the Coleoptera or Diptera (Stevens ’06, ’o8a). A rarer but more interesting deviation from the type is the failure of the idiochro- mosomes to separate in the second division, both passing together to the same pole (Fig. 5, 0, p, Photo 18). Since the other chromo- somes divide equally it may be inferred that in this case one pole receives 12 chromosomes and the other but 10. This has been seen in only three cells and is doubtless an abnormality. It may however, possess a high significance as forming a possible point Studies on Chromosomes 167 of departure for the origin of the whole series of relations observed in the genus. 3 Individuals possessing twenty-three Chromosomes, one ; Supern umerary This condition exists in all three species and has been found in seven males and four females. In four of these males the super- numerary is large (of approximately the same size as the small idiochromosome, as in Fig. 1, g-z); 1n three it is no larger than the m-chromosomes (as in Fig. 1, ;-/), and is indistinguishable from the latter save in behavior. In each case, as already described, the spermatogonia show 23 chromsomes and the first division 13; and in those showing a small supernumerary in the first division the spermatogonia always show three very small chromosomes. The grouping in the first division, though conforming to the same general type, shows many variations of detail, as may be seen from Fig 7, a—/, Photos 4-6. It is a curious fact that the form of grouping is to some extent characteristic of the individual. For example, the typical arrangement, with both i1diochromosomes and supernumerary outside the ring, is very common in Nos. 43 (Fig. 1, 7-1) and 20 (7, a-c), very rare in Nos. 1, 2 (Fig. 7, 2) and 49 (Fig. 7, f-h). In No. 49, very many of the first division meta- phases show both supernumerary and small idiochromosome lying inside the ring (Fig. 7, g-h). I am unable to suggest an explanation of this. In this division all the chromosomes divide equally (Fig. 7, m-p), so that each secondary spermatocyte receives 13 chromosomes. The usual regrouping now takes place, and the idiochromosomes couple as usual to form an asymmetrical bivalent. ‘The super- numerary sometimes remains free (1. e., not attached to any other), in which case 12 chromosomes appear in polar view (Fig. 8, b,d). Much more frequently the supernumerary attaches itself to the idiochromosome bivalent to form a triad element, polar views now showing but 11 chromosomes (8, a, c), one of which is compound. The three components of such triads usually lie in a straight line, the supernumerary being attached sometimes to the small idio- 168 . Edmund B. Wilson Fic. 7 23-chromosome forms, one supernumerary a-h, first division, polar views, one large supernumerary; a-c, term., No. 20, typical grouping; d-e, gran., No. 48; f, g, , gran., No. 49 (Photo 5). i-l, first division, polar views, one small supernumerary; i, term. No. 1 (Photo 6); j-/, term. No. 43 typical grouping in k. m-p, first division, side-views; m and n (term. No. 43) show division of I, 7, m, and small s; 0, term., No. 20, division of I, 7, and large s; p, term., No. 43, division of m, i, and small s. q-s, spermatogonial groups from individuals with one large supernumerary; q, r, term., No. 20; s, gran., No. 49. t-y, spermatogonial groups from individuals with one small supernumerary; ¢, u, term., No. 43; v-y, term., No. 2 (Photo 29). Studies on Chromosomes 170 Edmund B. Wilson chromosome, sometimes to the large, or not infrequently lying between the two (Fig. 8, g, 4, o-q). ® e e @ & ef? / © & 00;,/@ 09 V6 mee see ome e 2 ® ? © e © 4ee a m? @ ~ ee e*o; 23-chromosome. forms, one supernumerary a-f, polar views, second division; a, gran., No. 49, large supernumerary attached; b (same cyst)super- numerary free; cd, similar views of terminalis, No. 43, with small supernumerary; e-f (No. 43), sister groups from same spindle, pclar views. g-m, side-views, second division, from gran., No. 49, with large supernumerary, free in j, attached in the others. n—u, similar views from individual (term., No. 43) with small supernumerary; in w the supernumer- ary is free.” Studies on Chromosomes I7I In the ensuing division, if the supernumerary lies free it passes without division as a heterotropic chromosome to one pole (8, w). When connected with the idiochromosome bivalent it passes to one pole attached to one or the other of the idiochromosomes (Fig. 8, k-m, p-t). In either case one pole receives 11 chromosomes and one 12 (Fig. 8, e, 7); but since the supernumerary may accompany either 1diochromosome four classes of spermatid nuclei are formed, namely: @) 10] 211 @)io-bek ss — 12 (3) toti=11 (44) 10+ 7 +s=12 As described in an earlier paper (’07a), there is a tendency for the supernumerary to be associated more often with the small idiochromosome than with the large, and classes 1 and 2 are accord- ingly more numerous than 3 and 4. I was formerly inclined to attribute importance to this as pointing to the more frequent occurrence of the supernumerary in the male than in the female. The larger series of data now available leads me to doubt whether it has much significance; for if (leaving the 21-chromosome forms out of account) the whole series of forms be taken together, one or more supernumeraries are found in 27 out of 34 males, and in 15 out of 19 females—about 80 per cent in each case. It appears therefore that in the long run the supernumeraries are distributed between the two sexes with approximate equality. Figs. 7, g-s show spermatogonial groups from individuals with one large supernumerary, but in none of them can this chromosome or the small idiochromosome be certainly distinguished. Fig. 7, t-y are from individuals with one small supernumerary, each showing three very small chromosomes. In ¢ and wu the small idiochromosome is doubtful. Fig. 7, v-y, on the other hand, are from an individual (terminalis, No. 2), showing great numbers of very fine spermatogonial groups, in almost all of which the small idiochromosome is at once recognizable. The same is true of a second individual from the same locality. These two individuals, from the Paulmier collection, were the first material I examined and found so puzzling until the examination of another similar individual, No. 43, cleared up the nature of the second division. 172 Edmund B. Wilson 4. Individuals with twenty-six Chromosomes; four Su pernumeraries It will be convenient to consider this type before the 24- and 25- chromosome forms, since the material is more favorable for an account of the remarkable phenomena occurring in the second division. Of these individuals there are seven males and three females, all three species being represented. Unfortunately very few perfectly clear spermatogonial groups are shown; but the spermatocyte-divisions and cells of the growth-period are particu- larly well shown and in large numbers of cells. In all but one of these individuals the four supernumeraries are large and of nearly equal size. In one (femoratus No. 40) two are large and two small. ‘The latter case, already shown in Fig. 2, 7-/, is further illustrated by Fig. 9, h, 7, 7, n, 0. ‘Two of these (h and 7) show but three supernumeraries in the first division, a common appear- ance in this individual (see p. 186). Fig. 9, a—/, show varying arrangements of the 16 chromosomes that appear in the first division, the most typical ones being k and/. In 9, a-c, k,/, both idiochromosomes and the four supernumeraries lie outside the ring. In 9, g, all but the large idiochromosome are inside the ring. In some of these slides the compound chromosome-nucleoli are shown with great distinctness in many cells of the growth-period. This body usually has the form of a flat plate that lies next the nuclear wall (Fig. 10, g, r) so that a clear view of all the compo- nents can only be had in tangential sections. ‘Thus viewed (Fig. 10, s—u, Photo 28) it may often be seen to consist of six components one of which (the large idiochromosome) is about twice the size of the others and is usually at one side or end of the group. The other five evidently represent the small idiochromosome and the four supernumeraries. In side view (Fig. 10, q,r) not more than three or four of the components, can as a rule be recognized. In a considerable number of cases these six chromosomes are not ageregated to form a single body but form two or more simpler bodies. The second division in these forms presents an extraordinary Studies on Chromosomes ye: Fic. 9 26-chromosome forms, four supernumeraries a-g, first polar, supernumeraries large and equal; a-d, fem., No. 42; e, gran., No. 55; f, gran., No. 59; g, gran., No. 60. h-j first polar, from (fem., No. 40, with two large supernumeraries and two small; all of these are shown in J, (cf. Fig. 2, 7), while in k and i one is missing (see p. 186). k, first polar, term., No. 36; / from same individual (Photo 9). m-o, spermatogonia groups; m, fem., No. 42, abnormal group with 27 chromosomes; 1, 0, fem., No. 40. showing two small supernumeraries. p-q, Ovarian groups, gran. No. 61. iy a Edmund B. Wilson appearance which I[ at first thought must be due to an artificial clumping together of the chromosomes through defective fixation; but the study of very many of these figures convinced me that such is not the case. As in the preceding types, ten of the chromosomes, including the m-chromosomes, have the form of symmetrical dumb-bell shaped bodies which are equally halved in the ensuing division. ‘The remaining chromosomes are usually aggregated to form a compound element (Fig. 10, h-/, Photos 22, 23) in which may be very clearly distinguished the same components as those that appear in the chromosome-nucleoli of the growth-period; and the size-relations make it evident that one of them 1s the large idiochromosome, one the small, while four are the supernumer- aries. In other words, these six chromosomes, which divide as separate univalents in the first division, have now again conju- gated to form a hexad group. ‘This compound element almost always lies near the center of the group. Polar views of this divi- sion accordingly show typically 11 chromosomes, of which. the central one is compound (Figs. 10, a—g, Photo 13). Not infre- quently, however, one or more of the supernumeraries may be sep- arate from the others (Fig. 10, f, g), the apparent number in polar view varying accordingly. In side views the grouping of the components of the hexad element is seen to vary considerably though the large idiochro- mosome is more frequently at one end of the group. In the ensu- ing division the other ten chromosomes divide equally, while the hexad element breaks apart into two groups that pass to opposite poles (Fig. 10, /-p). The distribution of the various elements is dificult to determine exactly, since they always lag behind the others in the anaphases and are scattered along the spindle in such a way as often to give confusing pictures. ‘The study of many such anaphases leads me to conclude, however, that at least one of the smaller components always passes to the opposite pole from the larger one, while the other four undergo a variable distribution. In Fig. 10, /, the group is just separating into three toward each pole; in 10 m, 1t 1s quite clear that three of the small ones are pass- ing to one pole, while the large one and two small ones are passing to the other, and Fig. 10, n, is probably a similar case. In these Studies on Chromosomes 175 cases it seems clear that each pole receives 13 chromosomes, as follows : A TOs! se 3G 118) b 190+i+2=13 Fig. 10, 0, on the other hand, shows a perfectly clear case in which the hexad element has separated into a 2-group and 4-group: Fig. 10, p, shows what 1s probably a later stage of the same type. In both these cases one pole appears to receive 12 and one 14 as follows: A Wear Il ap Bo ap 11 b 1o+t+s5=12 one pole receiving but one supernumerary, and the other three. The cases in which all of the components may be clearly recog- nized in the anaphases are comparatively rare, and in the greater number of them the distribution of the supernumeraries appears to be symmetrical. Of their unsymmetrical distribution in some cases there can be no doubt (and the same is true of the 14-chromo, some form, as described beyond). ‘The few undoubted cases of this all show one to one pole and three to the other (as in Fig. ro, o- p), and I have never found a case in which all four pass to the same pole. It seems, therefore, probable that in the 26-chromosome type there are at least six classes of spermatozoa, as follows: (1) 190 +2 +25= 13 (2) 1o+it+2s=13 (3) 1o+ 7+ s=12 (4) 1o+7+35= 14 (5) 10+ 1+ 35= 14 (6) 10+7+ s=12 It is possible that the following four additional classes may be produced : (7) 10+ 1+ 45 = 15 (8) 10+7 = 11 (9) 10+ 7 = 11 (70) 10+7+45= 15 Perfectly clear spermatogonial figures of this type were rarely found, though many of them show approximately 26. The nor- mal group of fem., No. 42,1s shown in Fig. 2, h. “Iwo groups from fem. No. 40 (with two small and two large supernumeraries) are shown in Fig. 9, 1, 0, each having 26 chromosomes including four small ones (cf. Fig. 2, k). “Two ovarian groups from gran., No. 61, 176 Edmund B. Wilson Fic. 10 26-chromosome forms a-g, second division, polar, d from fem. No. 40, the others from fem. No. 42; a, (Photo 13) b, c, show a single central hexad; ine and g the components are more loosely united; in d and f one supernumer- ary is free. h-p, side-views, second division, from fem. No. 42 (Photos 22, 23) explanation in text. q-u, growth-period, gran.. No. 60; g and r show the compound chromosome-nucleolus in oblique and side-view, s,t, u, en face. Studies on Chromosomes 77] 178 Edmund B. Wilson Fic. 11 24-chromosome forms, two supernumeraries. a-e, term., No. 21, first polar, showing various groupings; g, the same, gran., No. 52 (Photo 7). h, term., No. 21, second polar, tetrad element near center. i-0, somatic groups from individuals with two large supernumeraries; 7—/, spermatogonial groups from term. No. 21; m, n, ovarian groups from fem. No. 31; 0, ovarian group, fem., No. 45. p-r, spermatogonial groups from fem., No. 22, with one large supernumerary and one small; Photo 30). s-w, second division, side-view; s, term., No. 21; t-w, gran., No. 52 (Photo 21). Bio Studies On Chromosomes Fic. 11 180 Edmund B. Wilson are shown in Fig.g, p, g. Fig. 9, , shows a spermatogonial group from fem., No. 42, that is abnormal in showing with perfect clear- ness 27 instead of 26 chromosomes (cf. Fig. 2, /). 5 Individuals with twenty-four Chromosomes, two Supernumerartes The material for these individuals and those of the 25-chromo- some class, is less satisfactory than in the preceding case, but the relations are undoubtedly quite analogous to those just described. The 24-chromosome class is represented by 9 males and 4 females, and occurs in all three species. In one of the males one of the supernumeraries is large (of the same size as a small idiochromo- some) and one small; in all the others both are large. Additional figures of the first division, showing variations in the grouping, are given in Fig. 11, a—g; of spermatogonial groups in Figs. 11, 7—r. Of particular interest is the male, term., No. 22, shown in Photo 30 and in Fig. 11, p-r. ‘This individual was, unfortunately, immature showing only spermatogonia and cells in the growth-period; but many perfectly clear spermatogonial groups are shown. ‘These groups uniformly show 24 chromosomes, of which three are very small, while in many cases two others are slightly but distinctly smaller than the others. ‘The latter are evidently the small idio- chromosome and the larger supernumerary, while the three small ones represent the m-chromosomes and the small supernumerary. In the second division the two idiochromosomes and the super- numeraries are frequently united to form a tetrad element, various forms of which are shown in Fig. 11, s-w. ‘The distribution of these four components 1s not so well shown in this material as in that of the 26-chromosome class, described above. It is, however, clear that this distribution is inconstant. In cases like those shown in Fig. 11, s, ¢, it is probable that the tetrad divides in the middle, so that each idiochromosome is accompanied by a supernumerary, and each pole receives 12 chromosomes. ‘The cases shown in Fig. 11, v, w, prove however that this.is not always the case; for in w the large idiochromosome is seen passing to one pole while both supernumeraries, attached to the small idiochromosome, Studies on Chromosomes 181 are passing to the other. In this case one pole receives 11 chromo- somes, the other 13. It is evident that in this form there is the possibility of forming six classes of spermatozoa, as follows: (1) 10+ 7J=11 (2) 1o+f+25= 13 (3) 1o+ T+ s=12 (4) lotit s=12 (s)) tol 2s — 13 (6) 1o+% =11 In none of these individuals is the material very favorable for the study of the chromosome-nucleoli. ‘They are always evidently compound, but only in a few cases can the components be clearly recognized (as in Fig. 2, c). 6 Individuals with twenty-five Chromosomes; three Supernumerartes No individuals of this type were found in M. femoratus. The other two species are represented by three males and three females but here again the material does not admit of exhaustive study. In one of the females, two of the supernumeraries are large and one small, the ovarian cells showing 25 chromosomes, of which three are very small (Fig. 12, »-&), a condition seen in every group of which a clear view can be had. ‘The two larger supernumer- aries cannot, however, be certainly identified in any of these. In all the other individuals the supernumeraries are of the larger form. Fig. 12, a, b, show the first division in one of these cases (term., No. 34); c-g are spermatogonial groups from the same indi- vidual; h, an ovarian group of the same type. Fig. 12, m—p, are from a doubtful case in which nearly all the first division figures show three supernumeraries (7, 0), but a single case (m) shows distinctly four. 7 Individuals with twenty-seven Chromosomes, five Supernumeraries. This class is represented by a single very interesting male of granulosus (No. 57), in which only the first division can be satisfac- torily examined. Many polar views of this division show 17 chromosomes (Fig. 13, a-1, Photo 10), of which two are always 182 ‘Edmund B. Wilson smaller than the others. One of these, always central in position, is evidently the m-chromosome bivalent. Of the remaining six, one is in most cases decidedly smaller than the others—a relation — # . Es & ® : el ~ oo of LAD fe ss % ° . re Sorgsn : o ae . M N 0 P 25-chromosome forms, three supernumeraries a, b, first polar, term., No. 34; c-g, spermatogonial groups from same individual; h, term., No. 38. ovarian group; i-k, ovarian groups from term., No. 27, with two large supernumeraries and one small; /, gran., No. 58, ovarian groups, three large supernumeraries. m, n, 0, first division, p, second division from gran., No. 54, with three or four supernumeraries. of which the constancy is attested by the nine figures given of this division. It is evident that in this individual there are four large supernumeraries and one small; and although nospermatogonia Studies on Chromosomes 183 are clearly shown it may be inferred that the somatic number is 27. The chromosome-nucleoli in this individual are evidently com- pound, but in no case can all the components be clearly recognized. The second division shows, as a rule, 11 elements in polar view, the central one being compound (Fig. 13, 7-k), but the distribu- tion of the compound element could not be determined. @7 @B @ ey 6,2 Go & © ©... e ve ay & re e es @ ee > es B 3 @ @° 4 e@ tig ee & @ ee &g @ @ &°:: @ Oy 04 a b C d ‘ e ° @/ o%S, ocee S'S, 880 @®° @e @e: © 6. © & e = e 7% @,°’ : i @a2 Og ee @ 8,0 J @> @ YA \ @ og 9 g z On AN Sao ee" 382 | 48") . & ee \\\\) | y, e °@ @ . . \ \) Vb ° . / i Wh / Fic. 13 27 and (?) 28-chromosome forms a-i, first division, from gran., No. 57, having four large supernumeraries and one small j (polar) and k (side-view), second division, same individual (Photo 10). /, ovarian group from fem., No. 33, having three large and two or three small supernumeraries; in this group appear 28 chromosomes. 8 Individuals with twenty-eight (?) Chromosomes; six Supernumerartes. The last case to be considered is that of a single female of femora- tus (No. 33), in which the number is either 27 or 28. A single perfectly clear ovarian group, shown in Fig. 13, /, shows beyond 184 Edmund B. Wilson doubt 28 chromosomes, including five smallest ones and three or four next smallest. A few other less clear groups were seen in which appear but 27 chromosomes, the missing one being one of the smallest. In these cases one of the’small ones may be hidden among the larger ones; but it is also possible that the 28-group is an abnormality. In this individual there are probably three larger supernumeraries and either two or three small ones. C SUMMARY AND CRITIQUE 1 In the genus Metapodius the number of chromosomes is constant in the individual but varies in different individuals from 21 to 27 or 28. The number 21 appears only in the males of M. terminalis (Montgomery’s material). 2 The number is independent of sex and locality, and is not correlated with constant differences of size or visible structure in the adults. 3. The variation affects only a particular class of chromosomes. 4 The 22-chromosome forms represent the type from which all the others may readily be derived. These forms possess a pair of unequal idiochromosomes which show the same behavior as in Lygeus or Euschistus, all the spermatozoa receiving 11 chromo- somes, and half containing the large idiochromosome, half the small. 5 In the 21-chromosome forms the small idiochromosome has disappeared, leaving the large one as an “‘odd” or “accessory” chromosome. Half the spermatozoa accordingly r receive II chro- mosomes and half Io. 6 Numbers above 22 are due to the presence of from one to five or six additional small chromosomes which show in every respect the same behavior as the idiochromosomes, and are probably to be regarded as additional small idiochromosomes. In the growth period they have a condensed form and are typically associated with the idiochromosomes to form a compound chromosome- nucleolus. In the first division they divide as separate univalents. In the second, they are typically (though not invariably) again associated with the idiochromosomes to form a compound element. The components of this element undergo a variable distribution Studies on Chroniosomes 185 to the spermatid nuclei. All the spermatid nuclei receive the haploid type-group of 11 chromosomes, half including the small idiochromosomes and half the large; but in addition each may receive one or more supernumeraries. The total number of chromosomes in the sperm nuclei is therefore variable in the same individual. | 7 Both the number ef the supernumeraries and their size, indi- vidually considered, are constant in the individual. The first question that the foregoing report of results will raise is whether the number and size relations of the chromosomes in each individual are really as constant as I have described them. I have for the most part selected for illustration and description the more typical conditions; but, granting the accuracy of the figures, does such a selection really give a fair presentation of the actual conditions? It is almost needless to say that very many cases might have been shown that would seem to give conflicting results. By far the greater number of these discrepancies are, I believe, only apparent. Numerical discrepancies of this kind are very often evidently due to mere accidents of sectioning or to the super- position or close contact of two or more chromosomes. Again, apparent discrepancies in the size relations of the chromosomes, as seen in polar views, very often arise through different degrees of elongation (particularly in the maturation divisions). But apart from such apparent variations, real deviations undoubtedly occur in almost all of the relations described. Now and then, for exam- ple, a spermatogonial or ovarian group 1s found that clearly shows one chromosome too many (as in Fig. 9, ™m),° and the same is true of the first spermatocyte-division, but such cases are very rare. The former case 1s probably a result of an abnormality in the forma- tion of the chromosomes from the resting nucleus, the latter not improbably to a failure of synapsis. Again, both spermatogonial and spermatocyte-cysts are occasionally found in which the num- ber of chromosomes is doubled or quite irregular. These are * A perfectly clear case of this has been found in the pyrrochorid species Largus cinctus (a particu- larly fine form for study): In this form the normal male number is 11, the female 12; but in one female three cells were found each of which shows with all possible clearness 13 chromosomes, very many other cells showing the normal number. 186 Edmund B. Wilson probably due to an antecedent nuclear division without cell divi- sion, or to multipolar mitoses such as now and then occur in both spermatogonia and spermatocytes. As regards the chromosome-nucleoli of the growth-period, the contrast between those of the 21 and 22-chromosome forms, or between either of these forms and those with higher numbers is usually at once apparent; but in very many cases where more than one supernumerary is present the number of components can only here and there be clearly seen. Contrary to what might be expected from their compact form, the compound chromosome nucleoli seem to be rather difficult of proper fixation, their components often clumping together or breaking up more or less when they coagulate. I infer this from the fact that different slides differ materially in the clearness with which these bodies are shown. Two discrepancies, apparent or real, should be especially men- tioned. One is the difficulty of recognizing the larger supernu- meraries in the somatic groups. As already explained, these chro- mosomes, like the idiochromosomes, appear relatively much larger in the somatic groups than in the first maturation division (owing to their univalence in the latter case); but we should expect to recognize them more clearly, at least in the female groups, than is actually the case. This is perhaps due to their undergoing a ereater degree of condensation than the others during the growth- period; but I am not sure that this explanation will sufice. A second discrepancy, which may involve an important conclusion, is that in perfectly clear views of the first division, the number of supernumeraries is often less than would be expected from the spermatogonial groups. ‘This is notably the case with femoratus, No. 40 (Fig. 9, A-7), which has clearly 26 spermatogonial chro- mosomes, but very rarely shows 16 in the first division, the usual number being 15. A similar discrepancy has been noted in other individuals, and in several of the types. Since the typical num- ber in all these cases appears in some or many of the first sperma- tocytes, I long supposed the occasional deficiency to result from an accident of sectioning. I now incline to believe, however, that in some cases one (or easel more) of the supernumeraries may really disappear (by degeneration?) during the growth-period, Studies on Chromosomes 187 and that this may be one way in which their progressive accumu- lation in number in successive generations is held in check. For the foregoing reasons it cannot be said that any of the rela- tions described appear with absolute uniformity or fixity. The condition typical of each individual must be discovered by the study and comparison of large numbers of cells. I will only say that prolonged and repeated study has thoroughly convinced me that the relations, as described, may be regarded as being on the whole individual constants. ‘This judgment is based primarily on the exhaustive study of a few of the best series of preparations of individuals of the 21, 22, 23, and 26-chromosome types, in which the facts are quite unmistakable and have given the point of view from which the less favorable material of other cases may fairly be examined. D DISCUSSION OF RESULTS The principal significance of these phenomena seems to me to lie in their bearing on the general hypothesis of the “individual- ity” or genetic continuity of the chromosomes; but they are also of interest for a number of more special problems which I will first briefly consider. T he Relation of the Chromosomes to Sex-production in Metapodius The conditions seen in this genus seem to be irreconcilable with any view that ascribes the sexual differentiation to a general quanti- tative difference of chromatin, whether expressed in the number or the relative size of the chromosomes. In all known cases of constant sexual differences in the chromosomes it 1s invariably the female that possesses the larger number of chromosomes or the greater quantity of chromatin,!® and this has naturally suggested the view that this difference per se may be the sex-determining factor. As I have pointed out before (’0g), such a view is inapplicable to cases like Nezara or Oncopeltus, where the idiochromosomes are of equal size and no quantitative sexual differences are visible; yet the phenomena in these genera are otherwise so closely similar 4 10 See review in Wilson ’og. 188 Edmund B. Wilson to those seen in other insects that I cannot doubt their essential similarity also in respect to sex-production. In Metapodius the facts are still more evidently opposed to the quantitative interpretation. The number of chromosomes has here no relation to sex-production; and, as will be seen from the table at p. 149, in the forms with supernumeraries the relative . frequency of high numbers and of low is nearly equal in the two sexes. If my general interpretation of the chromosomes in this genus be correct, a like conclusion applies to the total relative mass of chromatin in the two sexes; for all individuals alike possess the type-group of 22 chromosomes (Montgomery’s form excepted) while the supernumeraries represent the excess above this amount. I have endeavored to determine whether this appears in direct measurements, independently of my general interpretation; but have found this impracticable for several reasons. Very consider- able differences in the apparent size of the chromosomes are pro- duced by different degrees of extraction; but this will not account for the considerable differences seen in the same slide when the extraction is uniform. It is evident that the actual size of the chromosomes varies with the size of the cells; for example, both in Metapodius and in many other genera, the chromosomes in the larger spermatogonia near the tip of the testis are larger (in many cases much larger) than those of the smaller spermato- gonia of other regions. How great the differences are may be appreciated by a comparison of the figures. For example, in the spermatogonial groups of No. 2 (23 chromosomes, Fig. ? v-x), the chromatin mass is obviously much greater than in those of No. 21 (24 chromosomes, Fig. 11, 7-/). In the 25-chromo- some female groups shown in Fig. 12, :-k (No. 27), the chromatin mass is evidently much less than in the 21-chromosome male group shown in Fig. 1, 6, or in the 23-chromosome male groups of Fig. 7, v-x. Conversely, the 22-chromosome female group of No. 44 (Fig. 4, s) shows a much greater chromatin mass than in the corre- sponding male group of No. 46 (Fig. 4, 0), or the male 24-chromo- some group shown in Fig. 11, 7. Evidently, therefore, the relative mass of chromatin can only be determined by means of accurate measurements of both the Studies on Chromosomes 189 chromosomes and the mass of protoplasm, but I have found the errors of measurement of the cell size to be too great to give any trustworthy result regarding the relative chromatin mass. Despite the difficulties in the way of an accurate direct deter- mination, I believe the facts on the whole warrant the conclusion that the relative chromatin mass shows no constant correlation with sex. The most probable conclusion is that the male-produc- ing spermatozoa in Metapodius are distinguished by the same characters as in other forms having unequal idiochromosomes, the former class being those that receive the large idiochromosome, the latter those that receive the small one, irrespective of the super- numeraries that may be present in either class. For reasons that I have elsewhere stated, I believe that if the idiochromosomes be the sex-determinants their difference is probably a qualitative one, and since the small idiochromosome may be lacking it would seem that the large one must in every case play the active role— perhaps as the bearer of a specific substance (enzyme !) that calls forth a definite reaction on the part of the developing individual. If this be so, we can comprehend the fact that the presence of additional small idiochromosomes (supernumeraries) in either sex does not affect the development of the sexual characters in that Sex. b The possible Origin of the unpaired Idiochromosome (“odd” or “accessory Chromosome) and of the Supernumeraries The explanation of the unpaired idiochromosome offered in the second and third of my “Studies on Chromosomes” (’05, ’06) was suggested by the fact that various degrees of inequality exist in the paired idiochromosomes, there being an almost continuous series of forms connecting those in which the idiochromosomes are equal (Nezara, Oncopeltus) with those in which they are so very unequal that the small one appears almost vestigial (Lygzus, Tenebrio). It is evident that by the further reduction and final disappearance of the small member of this pair the large one would be left without a mate, and its history in the maturation process would become identical with that of an “odd” or “accessory” 190 Edmund B. Wilson chromosome. I still believe that this explanation may be applic- able to many cases; but a different one seems more probable in the case of Metapodius and perhaps may be more widely applicable. This was suggested by the observation (p. 166) that in a very few cases, in 22-chromosome individuals both idiochromosomes were seen passing to the same pole in the second division. ‘The rare- ness of this occurrence shows that it is doubtless to be regarded in one sense as abnormal. But even a single such event in an original 22-chromosome male, i} the resulting spermatozoa were functional, might give the starting point for the whole series of relations ob- served in the genus, including the establishment of an unpaired idio- chromosome. ‘The result of such a division should be a pair of spermatozoa containing respectively 10 and 12 chromosomes. The former might give rise at once to a race having an unpaired idiochromosome and the somatic number 21 in the male (as in Montgomery’s material). ‘The latter might similarly produce an individual having in the first generation a single supernumerary chromosome and in succeeding generations an additional number. This appears from the following considerations: 1 If a to-chromosome spermatozoon, arising in the manner indicated, should fertilize an egg of the 22-chromosome class (hav- ing 11 chromosomes after reduction) the result should be a male containing 21 chromosomes, the odd one being the large idiochro- mosome derived from the egg. Such an individual would be in no respect distinguishable from those of Montgomery’s material, and would similarly form male-producing spermatozoa containing 10 chromosomes and female-producing ones containing 11 (includ- ing the unpaired idiochromosome). A single such male, paired with an ordinary 22-chromosome female, would suffice to establish a stable race identical with the form found by Montgomery at West Chester, Pa., the males having 21 chromosomes, the females having 22, precisely as in Anasa or Leptoglossus. “This seems to me the most probable explanation of the conditions found in Montgomery’s material; and possibly it may explain the origin of the unpaired idiochromosome in other cases as well. 2 The result of fertilizing the same type of egg by a spermato- zoon from the 12-chromosome pole would be an individual having Studies on Chromosomes IgI 23 chromosomes (egg II + spermatozo6n 12) including two large idiochromosomes—hence presumably a female—and one small. ‘The eggs produced by such a female should after matura- tion be of two classes, having respectively 11 and 12 chromosomes. The 12-chromosome class would contain both a large and a small idiochromosome, and if fertilized by ordinary 11-chromosome spermatozoa would produce individuals with 23 chromosomes, male or female according to the class of spermatozoon concerned. Such females would, as before, contain two large idiochromosomes and one small. ‘The males would contain one large and two small. and would accordingly produce spermatozoa having either If or 12 chromosomes. Now, such an additional small idiochromosome in the male would be indistinguishable from a single “supernumerary chro- mosome’”’ as it appears in the 23-chromosome individuals in my material. The resemblance is shown not only in size but also in behavior; for, as I have shown, the supernumerary, like the idiochromosome, forms a chromosome-nucleolus during the growth period, it divides as a univalentin the first division, and in the sec- ond is usually associated with the idiochromosome bivalent. A single such supernumerary chromosome, once introduced into the race would lead to the presence of additional ones in succeeding generations. ‘Thus, :2-chromosome eggs fertilized by 12-chromo- some spermatozoa would give individuals (male or female) with 24 chromosomes, including two supernumeraries; and from these might arise, through irregularities of distribution such as I have described, gametes with 11, 12, or 13 chromosomes, giving in the next generation 22, 23, 24, 25 or 26 chromosomes according to the particular combination established in fertilization." If this 1 Since the presence of an unpaired idiochromosome in some individuals and of supernumeraries in others is assumed to be traceable to the same initial cause, we should naturally expect to find the two conditions coexisting side by side, and in approximately equal numbers; but in point of fact the former is very rare and was only found in one locality, while the latter is very common. This may constitute a valid objection to my interpretation. It should be borne in mind, however, that abnormal divi- sions of the kind assumed to form the starting point are very rare, and that an extremely minute propor- tion of the total number of spermatozoa produced ever actually enter the eggs. The chances against fertilization by either class of the original modified spermatozoa are therefore very great. Since only sixty individuals have been examined it need not surprise us that one of the two conditions in question 192 Edmund B. Wilson interpretation be correct, the origin of an unpaired chromosome in certain individuals of this genus has been owing to the same cause that has produced the supernumeraries. Since both condi- tions coexist in the same species, along with that which may be regarded as the original type (22 chromosomes) it may be conclu- ded that Metapodius is now in a period of transition from the sec- ond to the third of the types distinguished in my last study. It seems quite possible that other species of coreids that now have constantly an unpaired idiochromosome may have passed through a similar condition, though in all of them thus far examined both the small idiochromosome and the supernumeraries have dis- appeared. In Metapodius, accordingly, the supernumeraries may be regarded as on the road to disappearance. ‘That such is the case is rendered probable by the fact that their number does- not pass a certain limit, and is rarely more than four. ‘The very small chromosomes of this kind, so often observed, are perhaps degenerating, or even vestigial in character. But aside from this, attention has already been called to the probability that one or more of the supernumeraries may be lost during the growth-period (p.186); and while this is not certain, it may well be that both methods are operative in their disappearance. The foregoing interpretation of the supernumeraries enables us to understand why variations in their number are not accom- panied by corresponding morphological differences in the soma- tic characters; for they are but duplicates of a chromosome already present and hence introduce no new qualitative factor. It can hardly be doubted that some kind of quantitative difference must exist between individuals that show different numbers, but none has been more frequently met with. Another objection might be based on the different relations that occur in Syromastes. In this form (see Wilson ’og) the passage of both idiochromosomes to one pole without separation is a normal and constant feature of the second division, yet no supernumeraries appear in any of the individuals, and it is probable that the female groups contain two pairs of idio- chromosomes like the single pair that appears in the male. We have no data for a conjecture as to how such a condition can have arisen; but evidently the small idiochromosome does not in this case become an erratic supernumerary but retains a definite adjustment to the other chromosomes. Still, I do not consider this an obstacle to my interpretation of Metapodius, for it is now evident that the history of the idiochromosomes in general has differed widely in different species and families, even among the Hemip- tera. We have thus far only made a beginning in their comparative study. [See Addendum, p. 200.] Studies on Chromosomes 193 has thus far been discerned. Such a difference does not appear in the size of the animals, for there are large individuals with no supernumeraries and small individuals that possess them. An interesting field for experiment seems here to be offered. ¢ The “Individuality” or Genetic Continuity of the Chromosomes It is in respect to this much debated hypothesis that the facts observed in Metapodius seem to me most significant and important. It is evident that the whole series of relations are readily intelligi- ble if the fundamental assumption of this hypothesis be accepted. Without such explanation they seem to me to present an insoluble puzzle. The disposition to reject this hypothesis that appears in a considerable number of recent papers on the subject will doubtless lead to more critical and exhaustive observation of the facts; but when it goes so far as to deny every principle of genetic continuity in respect to the chromosomes, it is, I believe, a backward step. This reaction perhaps reaches a climax in the elaborate and apparently destructive criticism of Fick (07) who considers the hypothesis to be thoroughly discredited, and believes his analysis to justify the conclusion: “Dass weder theoretisch noch sachliche Bewetse fiir die Erhaltungslehre vorliegen, sondern dass 1m Gegen- theil unwiderlegliche Beweise gegen sie vorhanden sind, so dass es im Interesse der Wissenschaft dringend zu winschen ist, dass die Hypothese von allen Autoren verlassen wird” (’07, p. 112, italics in original). [ incline to think that this sweeping judgment would have carried greater weight had Professor Fick, in certain parts of his able and valuable discussion, taken somewhat greater pains in his presentation of facts and shown a more judicial temper in their analysis.” To some of the objections and difficulties 2 T will give two specific examples of this. The experimental results of Moenkhaus (’04), on hybrid fishes, which evidently form a strong support to the continuity hypothesis, are unintentionally but com- pletely misrepresented in the statement at p. 75: ‘‘So berichtet Moenkhaus bei Fundulus-Monidia- kreuzung (sic), dass sich die beiderlei (zuerst sehr verschiedenartigen) Chromosomen in der Regel schon nach der zweiten Teilung nicht mehr unterscheiden lassen.”’ But Moenkhaus’s explicit statement, based on the examination of ‘“‘many thousand cells,” is that even in the late cleavage ‘‘ Nuclei showing the two kinds of chromosomes mingled together upon the spindle are everywhere to be found” (op. cit., p- 48). Fick evidently had in mind the fact that the paternal and maternal chromosomes do not as a rule retain their original grouping after the first two or three cleavages. His actual statement, however, 194 Edmund B. Wilson brought forward in this critique reply has already been made by Boveri (’07), Strasburger (’08), Schreiner (’08) Bonnevie (’08) and others. Some of the difficulties are real, but an attentive study of the matter will show that a large part of Fick’s critique is directed against the strict hypothesis of individuality and offers no adequate interpretation of the essential phenomenon that requires explanation. It may be admitted that many of the facts seem at present difficult to reconcile with the view that the identity of the chromosomes as actual individuals is maintained in the “resting”’ nucleus; and I have myself indicated (The Cell, 1g00, p. 300) that the name “individuality” was perhaps not the best that could have been chosen. Certainly we have as yet comparatively little evi- dence that the chromosomes retain their boundaries in the “rest- ing” nucleus. It is evident that the chromosomes are greatly diffused in the nuclear network, and it may be that the substances of different chromosomes are more or less intermingled at this time. Fick’s “‘manceuvre-hypothesis,’ which treats the chromosomes of the dividing cell as temporary “tactic formations,” may there- fore be in some respects a more correct formulation of the facts than that given by the hypothesis of “individuality” in the strict sense of the term. But the last word on this question has by no means yet been spoken. A new light is thrown on it by the recent important work of Bonnevie (’08) which brings forward strong evidence to show that in rapidly dividing cells (cleavage stages of Ascaris, root-tips of Allium), although the identity of the orig- (both here and in the later passage at p. 98) will wholly mislead a reader not familiar with Moenkhaus’ work, in regard to one of the most significant and important discoveries in this whole field of inquiry. Hardly less misleading is Professor Fick’s report of my own observations on the sex-chromosomes of insects, which are stated as follows: ‘‘Wilson’s Unterschungen beweisen eben sicher nur soviel, dass bei einigen Insektengattungen constante Beziehungen zwischen dem Geschlecht und dem Vorhandensein eines besonderen Chromosomenpaares bestehen, bei anderen Gattungen nicht” (p.go). Iam confident that those who are familiar with the researches referred to will not accept this as a fair statement of the results. The fact is that in one form or other the sex-chromosomes are present in all of the forms that I have examined (now upwards of seventy species) and that with various modifications all conform to the same fundamental type. It is true that in two genera (Nezara and Oncopeltus) the sex-chromosomes are equal in size, and hence afford no visible differential between the somatic groups of the two sexes; but I especially emphasized the fact (cf. ’06, pp. 17, 34) that these chromosomes are in every other respect identical with those of other forms in which the size-difference clearly appears, and are connected with the latter by a series of intermediate gradations that leaves no doubt of the essential uniformity of the phenomena. Studies on Chromosomes 195 inal chromosomes is lost in the “ resting’ nucleus after each mito- sis, each new chromosome nevertheless arises by a kind of endog- enous formation within and from the substance of its predecessor. In this way an individual genetic continuity of the chromosomes can be directly followed through the “resting period” of the nu- cleus. “Eine genetische Kontinuitat der Chromosomen nacheinan- der folgender Mitosen konnte in der von mir untersuchten Objekten teils sicher (Allium, Amphiuma) teils mit iberwiegen- der Wahrsheinlichkeit (Ascaris) verfolet werden. Es ging aber auch hervor, dass eine /dentitat der Chromosomen verschiedener Mitosen nicht existiert, sondern dass jedes Chromosom in einem fruher existierenden endogen entstanden ist, um wieder am Ende seines Lebens fiir die endogene Entstehung eines neuen Chromo- soms die Grundlage zu bilden” (op. cit, p. 54). Whether this particular conclusion will also apply to more slowly dividing cells remains to be seen. But apart from this direct evidence it seems to me that a denial of every form of genetic continuity between the chromosomes of successive cell-generations—which, despite certain qualifications, seems to be the position of Fick and anum- ber of other recent writers—is only possible to those who are ready to ignore some of the most obvious and important of the known facts, especially those that recent research has brought to light among theinsects. ‘The most significant of these are: 1 In Metapodius the specific number varies, while in the indi- vidual both the number and the size-relations of the chromosomes are constant. 2 In all species where the somatic chromosome-groups show sexual differences in regard to the number and size-relations of the chromosomes, exactly corresponding differences exist between the male-producing and the female-producing spermatozoa. Both these series of facts demonstrate that the “tactic forma- tion” of a fixed number of chromosomes of particular size is not a specific property of a single chromatin-substance as such, of the species. It has been assumed by some writers that departures from the normal specific number, such as appear in merogonic, parthenogenetic, double-fertilized or giant (double) eggs, are the result merely of departures from the normal quantity of chroma- 196 Edmund B. Wilson tin.” If attentively considered the facts summarized above will, I think, clearly show the inadequacy of such an explanation. Why should a given quantity and quality of chromatin always reappear in the same morphological form as that in which it entered the nucleus? Why, for example, in Metapodius should the minute fraction of chromatin represented by a single small supernumerary always reappear in the form of such a chromosome, showing specific peculiarities of behavior, rather than as a corre- sponding enlargement of one of the other chromosomes? Why should a larger excess always appear as a group of two, three, or more supernumeraries that differ definitely in behavior from the others and show constant size relations among themselves ? Specifically, 1 in individual No. 40, why should two small supernu- meraries and two large ones always appear, rather than three large ones! In species where a constant quantitative chromatin-differ- ence exists between the sexes, why should the excess in the female always appear in the same form as that which appears in the female- producing spermatozoa—in one case as a large idiochromosome instead of a small (Lygzeus), in another as an additional chromo- some of a particular size (very large in Protenor, small in Alydus, of intermediate size in Anasa), in a third case as three additional chromosomes (Galgulus) ? To these and many similar questions which the facts compel us to consider, [ am unable to find any answer on the merely quantitative hypothesis. Each of them receives a simple and intelligible reply under the view that it is the number, size, and quality of the chromosomes that enter the nucleus that determine the number, size, and mode of behavior of those that issue from 13 Fick’s treatment of these cases is worth citing. ‘‘Esmussvon vornherein als wahrscheinlich be- zeichnet werden, dass unter den abnormen Umstinden, da einmal die Zahl der‘Chromatin-Mané6verein- heiten’ (im Sinne meiner Manéverhypothese gesprochen) in der Zelle erhéht ist, diese Zahl sich erhalt” (p. 96). Why should the number be maintained? Because, we are told, “Die Erhaltung der erhéhten Zahl und ihre regelmassige Wiederkehr bei den folgenden Teilungungen muss bei dem nun einmal uber die Norm erhéhten Chromosomenbestand der Zelle als der einfachere, leichter verstandliche V or- gang erscheinen, als es ein besonderer, ein ‘‘Regulation” auf die Norm hervorbringender Akt ware.” To most readers this will seem like an argument for, rather than against, the hypothesis of genetic continuity. But since it is obviously not thus intended I can discover no other meaning in the passage than that with a given “bestimmte Chromatinmanéverart” characteristicof the species (p.115) the num- ber of chromosomes formed is proportional to the quantity of chromatin-substance. Studies on Chromosomes 197 it. But such an answer implies the existence of a definite indi- vidual genetic relation between the chromosomes of successive cell- generations; and it is this relation, I take it, that forms the essence of the hypothesis of genetic continuity, whether or not we include in the hypothesis the assumption that the chromosomes persist as “individuals” in the resting nucleus where their boundaries seem to disappear. We might, for instance, assume that the chromo- somes are magazines of different substances (e. g., enzymes or the like) that differ more or less in different chromosomes, that are more or less diffused through the nucleus in its vegetative phase, but are again segregated out in the original manner when the chromosomes reform." We have, admittedly, but an imperfect notion of how such a re-segregation may be effected, though the conclusions of Bonnevie already referred to, constitute an impor- tant addition to the earlier ones of Boveri (see ’07, p. 232) in this direction. However this may be, in my view the most practicable, indeed the almost necessary, working attitude is to treat the chromo- somes as i} they were actually persistent individuals. ‘The facts in Metapodius, which at first sight seem to present so chaotic an aspect, fall at once into order and become intelligible if regarded as due to the presence in the species of a certain number of erratic chromosomes, one or more of which may be introduced into the zygote at the time of fertilization and which in some sense retain their identity throughout the development. The particular combina- tion established at the time of fertilization is the result of the chance union of two particular gamete combinations. Since the distribu- tion of the supernumeraries to the spermatid nuclei is variable, different gamete combinations occur in the spermatozoa of the same individual; and the same is probably true of the eggs. More- over, adults of the same species live side by side on the same food- plants and presumably may breed together. Different combinations may thus be produced in the offspring of a single pair, whether the parents possess the same or different numbers. Metapodius thus fulfills the prediction of Boveri, written nearly twenty years ago. ““Wenn bei einer Spezies einmal sehr viele und verschieden- 14 A view similar to this is suggested by Fick himself in his earlier discussion (’o5, p. 204), but it does not reappear in his later one. 198 Edmund B. Wilson artige Irregularitaten vorkamen, diese sich wohl auf lange hinaus erhalten missten, so dass unter Umstanden Falle mit ausserordent- lich grosser Variabilitat der Chromosomenzahl zur Beobachtung kommen konnten, ohne dass selbst diese das Grundgesetz umstos- sen yermochten, welches lautet: Es gehen aus jedem Kerngeriist so viele Chromosomen hervor als in die Bildung derselben eingegangen sind” (go, p. 61). ‘To the earlier expression of this ‘‘Grundgesetz’” Boveri has recently added the statement that the chromosomes that emerge from the nucleus are not merely of the same number but also show the same size-relations as those that entered it. ‘Was durch den kurzen Ausdruck “ Individuali- tat der Chromosomen”’ bezeichnet werden soll, ist die Annahme dass fiir jedes Chromosoma, das in einen Kern eingegangen ist, irgend eine Art von Einheit im ruhenden Kern erhalt, welche der Grund ist, dass aus diesem ruhenden Kern wieder genau ebenso viele Chromosomen hervorgehen und dass dieses Chromosomen iiberdies da, wo vorher verschiedene Grossen unterschieden waren, wieder in den gleichen Gréssenverhaltnissen auftreten” (’07, p.229). The facts seen in Metapodius and other insects are thoroughly in accord with the foregoing statement, and justify the additional one that the chromosomes conform to the same principle in respect to their characteristic modes of behavior. In the Hemiptera heteroptera generally the idiochromosomes and supernumeraries, the m-chromosomes, and the “ordinary chromosomes” or “auto- somes” show each certain constant peculiarities in respect to the time of synapsis and behavior during the growth-period, and assume a characteristic (though not entirely constant) mode of grouping in the first spermatocyte. Perhaps the most obvious of these facts is the very early condensation of the idiochromo- somes and supernumeraries in the growth-period as contrasted with the other chromosomes; and in the case of Pyrrochoris I have shown (’og) that the idiochromosome never assumes a diffuse con- dition after the last spermatogonial division. But even more significant are the definite differences shown in the couplings of the various forms of chromosomes that take place in the course of the spermatogenesis. Nothing in these phenomena is more striking than the accuracy with which these couplings take place. Studies on Chromosomes 199 As Montgomery and Sutton have shown, the ordinary paired chromosomes of the spermatogonia give rise to bivalents of corre- sponding size at the time of general synapsis. ‘The actual coupling of the ordinary chromosomes at this time is still a matter of dispute; but no doubt can exist in regard to the couplings that occur at a later period in case of the m-chromosomes, the idiochromosomes, and the supernumeraries. ‘These characteristic couplings are not determined merely by the size of the chromo- somes. The union of the unequal idiochromosomes after the second division takes place with the same regularity as that of the equal m-chromosomes in the prophases of the first. A small supernumerary that is indistinguishable from the m-chromosomes in the spermatogonia never couples with the latter in either divi- sion, but with the much larger idiochromosomes. ‘The couplings are equally independent of the original positions of these chromo- somes, either in the spermatogonia or in the growth-period, as is seen with especial clearness in case of the m-chromosomes. ‘These phenomena naturally suggest the conclusion that the couplings result from definite affinities among the chromosomes. ‘The possi- bility no doubt exists that the couplings are produced by extrinsic causes (such as the achromatic structures) but the evidence seems on the whole opposed to such a conclusion. I consider it more probable that they are due to intrinsic qualities of the chromosomes and that the differences of behavior shown by different forms may probably be regarded as due to corresponding physico-chemical differences. ‘This conclusion is in harmony with Boveri’s experi- mental results, though based on wholly different data. While it does not seem worth while to attempt its wider development here, I may express the opinion that all the chromosomes may con- sist in the main of the same material basis, differing only in respect to certain constituents; and further that the degree of qualitative difference may vary widely in different species. ZoGlogical Laboratory Columbia University August 10, 1908 15 See for example, Meves (’07, pp. 453-468) who, like O. Hertwig, Fick and others, rejects the theo Pp PP- 453-4 g J ry of “ individuality.” 200 Edmund B. Wilson ADDENDUM The probability in regard to the female groups of Syromastes, expressed in the footnote at p.19g2 was first stated in my preced- ing paper (’09, p. 73 ) after a study of the male only. Since the present paper was sent to press I have had opportunity to ex-— amine females of this form. The facts are exactly in ac- cordance with my prediction, the female groups containing 24 chromosomes, while the male number is 22. It now seems clear, however, that the two idiochromosomes of Syromastes do not correspond respectively to the large and the small idiochromo- some of Metapodius or Lygzus but are equivalent, taken together, to the large idiochromosome or to the odd chromosome of Anasa, etc. October 25, 1908. WORKS REFERRED TO Bonnevik, K. ’08—Chromosomenstudien. I. Arch. f. Zellforschung, i, 23. Boveri, Th. ’90—Zellenstudien. III. Ueber das Verhalten der chromatischen Kernsubstanz bei der Bildung der Richtungskérper und bei der Be- fruchtung. Jena, 18go. °o7—Zellenstudien. VI. Die Entwicklung dispermer Seeigel-Eier, etc. Jena, 1907. Fick, R. ’05—Betrachtungen iiber die Chromosomen, ihre Individualitat, Reduc- tion, und Vererbung. Arch. Anat. u. Phys., Anat. Abth., Suppl. 1905. °o7—Vererbungsfragen, Reduktions-und Chromosomenhypothesen, Bas- tard-Regeln. Merkel und Bonnet’s Ergebnisse, xvi, 1906. Foot, K. anp STROBELL, E. C. ’07a—The “Accessory Chromosome”’ of Anasa tristis. Biol. Bull., xii. °o7b—A Study of Chromosomes in the Spermatogensis of Anasa tristis. Am. Journ. Anat., vii, 2. Gross, J. ’04—Die Spermatogenese von Syromastes marginatus. Zo6l]. Jahrb., Anat. Ontog., xii. LeFEvrE, G.anp McG111, C. ’08—The Chromosomes of Anasa tristis and Anax junius. Am. Journ. Anat., vil. 4. Moenxuaus, W. S. ’04—The Development of the Hybrids between Fundulus heteroclitus and Menidia notata, etc. Am. Journ. Anat., ili. Studies on Chromosomes 201 Meves, Fr.—Die Spermatocytenteilungen bei der Honigbiene, etc. Arch. Mikr. Anat., xx. Montcomery, T. H. ’o1—A Study of the Germ-cells of Metazoa. Trans. Am. Phil. Soc., xx. *06—Chromosomes in the Spermatogenesis of the Hemiptera Heteroptera. ibids: xxi, 3) PautmigrR, F. C. ’99—The Spermatogenesis of Anasa tristis. Journ. Morph., xv, Suppl. SCHREINER, K. E. anp A. ’08—Gibt es eine parallele Konjugation der Chromo- somen? Videnskebs-Selskabets Skrifter. i. Math-Naturw. Klasse, 1908, no. 4. Stevens, N. M. ’06—Studies in Spermatogensis. IJ. A Comparative Study of the Heterochromosomes in Certain Species of Coleoptera, Hemiptera and Lepidoptera, etc. Carnegie Institution. Pub. 36, i. ’o8a—A Study of the Germ-cells of Certain Diptera, etc. Journ. Exp. ZLOON- 5. Ve, ie ?08b—The Chromosomes in Diabrotica vittata, etc. Ibid., v, iv. STRASBURGER, E. ’08—Chromosomenzahlen, Plasmastrukturen, Vererbungstrager und Reduktionsteilung. Jahr. wiss. Bot., xlv, iv. Witson, E. B. ’o5a—Studies on Chromosomes, I. Journ. Exp. Zodl., 11. ’o5b—Studies, etc. II. Ibid., ii, iv. ’o6—Studies, etc., III]. Ibid., iii, 1. *o9—Studies, etc., IV Ibid., vi, 1. ’07a—Note on the Chromosome-groups of Metapodius and Banasa. Biol. Bull.,.xu, 5. o7b—The Supernumerary Chromosomes of Hemiptera. Report of May Meeting. N. Y. Acad. Sci. Science, n. s., xxvi, 677. ’o7c-—The Case of Anasa tristis. Science, n. s., xxv, 631. Edmund B. Wilson APPENDIX List of individuals examined, arranged according to locality 2.02, No. Species I terminalis 2 terminalis 3-11 | terminalis 12 terminalis 13 terminalis 14 terminalis 15 terminalis 16 terminalis 17 terminalis 18 terminalis 19 terminalis 20 terminalis 21 terminalis 22 terminalis 23 terminalis 24 terminalis 25 terminalis 26 terminalis 27 terminalis 28 femoratus 29 femoratus 30 femoratus 31 femoratus 32 femoratus 33 femoratus 34 terminalis 35 terminalis 36 terminalis 37 terminalis 38 terminalis 39 femoratus 40 femoratus 41 femoratus 42 femoratus 43 terminalis 44 terminalis 4O 40 4O 40 Ay Q Ww 4 4 4 WO Ay AQ Q Q wo Q 4 Ww WH QA QQ YQ; HAAHDAAAAAAG Locality Madison, N. J. (Paulmier) Madison, N. J. (Paulmier) _ West Chester, Pa. (Montgomery) West Chester, Pa. (Wilson) West Chester, Pa. (Wilson) | West Chester, Pa. (Wilson) | West Chester, Pa. (Wilson) West Chester, Pa. (Wilson) Mansfield, Ohio Mansfield, Ohio Raleigh, N.C. Raleigh, N. C. Raleigh, N.C. | Raleigh, N.C. | Raleigh, N.C. | Raleigh, N. C. | Raleigh, N. C. Raleigh, N.C. Raleigh, N. C. Raleigh, N.C. Raleigh, N. C. | Raleigh, N.C. Raleigh, N.C. Raleigh, N.C. Raleigh, N. C. | Southern Pines, N.C. Southern Pines, N. C. Southern Pines, N. C. | Southern Pines, N. C. Southern Pines, N. C. Southern Pines, N. C. Southern Pines, N. C. Southern Pines, N. C. Columbia, S.C. Charleston, S. C. Charleston, S. C. Supernumeraries 1 small 1 small absent absent 1 large 1 large 2 large 2 large absent absent absent 1 large 2 large 1 large, 1 small absent absent 1 large 2 large 2 large, 1 small absent absent 1 large 2 large 4 large 3 large 2-3 small 3 large 3 large 4 large 2 large 3 large 2 large 2 large, 2 small 1 large 4 large I small absent Somatic * No. 23 23 21 22 23 23 24 a4 22 22 22 23 24, 24 22 22 23 24 25 22) 22 23 24, 26 27-8 25 25 26 24 25 24 26 23 26 23 22 No. in first div. 12 15 15 16 14 14 16 16 13 Studies on Chromosomes List of individuals examined, arranged according to locality—Continued Species “femoratus femoratus granulosus granulosus granulosus granulosus granulosus granulosus granulosus granulosus granulosus granulosus granulosus granulosus granulosus granulosus granulosus granulosus oO WAM OW AAAA™AAYA AA AA \™ w Locality Charleston, S.C. Savannah, Ga. Tucson, Arizona Tucson, Arizona Tucson, Arizona Tucson, Arizona Tucson, Arizona Tucson, Arizona Tucson, Arizona Tucson, Arizona Tucson, Arizona Tucson, Arizona Tucson, Arizona Tucson, Arizona Grand Canyon, Arizona Grand Canyon, Arizona Grand Canyon, Arizona Grand Canyon, Arizona Supernumaries 2 large absent absent 1 large 1 large 2 large 2 large 2 large 2 large 3-4 large 4 large 4 large 4 large, 1 small 3 large 4 large 4 large 4 large + 4 large Somatic No. 203 204 Edmund B. Wilson EXPLANATION OF PLATE I. The figures are reproduced directly from the original photographs, without retouching, at an enlarge- ment of 1500 diameters. It should be borne in mind that in the photographs considerable apparent size- variations are produced by differences of focus, and that unless the chromosomes lie exactly in one plane the photograph often gives a less accurate impression than a drawing. Drawings of most of these photo- graphs with designations, will be found among the text figures, as indicated. Y 1 M.terminalis (No. 3, Montogmery’s material), 21-chromosome form, first spermatocyte-division polar view; unpaired idiochromosome (odd or accessory) outside the ring, to the right (Fig. 3, 0). 2 M. terminalis (No. 19), 22-chromosome form, first division, polar view; the two separate idio- chromosomes at the right. (The small idiochromosome, being slightly out of focus, appears too small. Its size is correctly shown in the drawing, Fig. 4, b). 3 M. terminalis (No. 12), 22-chromosome form similar view; idiochromosomes in contact (Fig.4, f)- 4 M. terminalis (No. 20), 23-chromosome form, one large supernumerary, view similar to the pre- ceding; idiochromosomes and supernumerary to the right (Fig. 1, g). 5 M. granulous (No. 49), 23-chromosome form, one large supernumerary, which lies inside the ring with the small idiochromosome and m-chromosome (Fig. 7, g). 6 M. terminalis (No. 1), 23-chromosome form, one small supernumerary lying inside the ring with the m-chromosome and one of the large bivalents (Fig. 7, 7). 7 M. granulosus (No. 52), 24-chromosome form, two large supernumeraries (Fig. 11, g). 8 M. femoratus (No. 42), 26-chromosome form, four large supernumeraries (Fig. 2, g). 9 M. terminalis (No. 36), 26-chromosome form, similar to preceding (Fig. 9, e). 10 M. femoratus (No. 57), 27-chromosome form, four large supernumeraries and one small (Fig. 13, h). 11 M. femoratus (No. 46),.22-chromosome form, first division in side view, both idiochromosomes dividing (Fig. 4, 7). 12 M. granulosus (No. 47) 22-chromosome form, second division, polar view (Fig. 5, c). 13. M. femoratus (No. 42), 26-chromosome form; second division, polar view, showing hexad ele- ment near center (Fig. 10, a). 14 M. terminalis (No. 3, Montgomery’s material) 21-chromosome form, second division side view, showing lagging idiochromosome (‘‘accessory chromosome”) (Fig. 3, f). 15 From the same cyst as the last, later stage of second division; idiochromosome entering one pole (Fig. 3, g). 16 M. femoratus (No. 29), 22-chromosome form, second division metaphase in side view, showing idiochromosome bivalent (like Fig. 5, 2). 17 M. granulosus (No. 47), 22-chromosome form, late anaphase of second division, one idiochromo- some entering each pole (Fig. 5, /). 18 M. femoratus (No. 46), abnormal late anaphase of second division, showing both idiochromo- somes passing to the same pole (Fig. 5, 0). 19 M. femoratus (No. 29), 22-chromosome form, second division showing initial separation of the idiochromosomes (like Fig. 5, f). zo M. granulosus (No. 49), 23-chromosome form, one large supernumerary, second division meta- phase, showing triad element formed by the union of the supernumerary with the idiochromosome- bivalent (like Fig. 8, 7). 21 M. granulosus (No. 52), 24-chromosome form, two large supernumeraries, second division, show. ing tetrad element consisting of the idiochromosomes and supernumeraries united in a linear series (Fig. II, u). Studies on Chromosomes 205 22 M. femoratus (No. 42), 26-chromosome form, four large supernumeraries; second division show- ing hexad element formed by the idiochromosomes and supernumeraries (Fig. 10, /:). 23 From the same cyst, similar view (Fig. 10, k). 24 M. terminalis (No. 3, Montgomery’s material), 21-chromosome form, nucleus from the growth- period, showing single spheroidal chromosome nucleolus (like Fig. 3, /). 25 M. femoratus (No. 29), 22-chromosome form, growth-period, showing double chromosome- nucleolus (idiochromosome-bivalent) and plasmosome (Fig. 6, b). 26 From the same slide, showing different ordinary chromosomes, separate chromosome-nucleoli and plasmosome (Fig. 6, c). 27. M. terminalis (No. 20), 23-chromosome form, growth-period, showing tripartite chromosome- nucleolus formed by the idiochromosomes and supernumerary (like Fig. 1, 7). 28 M. granulosus (No. 60), 26-chromosome form, growth-period, showing hexad chromosome- nucleoli from three different cells (like Fig. 10, s-w). 29 M. terminalis (No. 2), 23-chromosome form, one small supernumary; spermatogonial group, showing three small chromosomes (the supernumerary and two m-chromosomes); the small idiochromo- some distinguishable above towards the left (Fig. 7, y). 30 M. terminalis (No. 22), 24-~chromosome form, one small supernumerary and one large (Fig. 11, p) STUDIES ON CHROMOSOMES V. PLATE 1. (E. B Wilson) 26 27 28 The Journal of Experimental Zoology, Vol. VI. i) ; 4 pais aah Satay te 7%. a He THE EFFECTS OF DESICCATION ON THE ROTIFER PHILODINA ROSEOLA BY MERKEL HENRY JACOBS Wiru One Ficure iby JinagoyshuenoyeE Gano ose coadasoK DDO cov odnd oo an DN UM OUODNoOUKbObEdOGGDeOODsONSHSaSe 207 TileeElistorical eerste scstocstorser-tore elects tre icisieiteleine clelolers is beter sicielerieletetstoeiteheeatcre teste etary 210 Ae structure and natural bistonyore mil Odinga eres) -jelelsfeictels letters cisiaie lel isteritere eee ere ey rete 217 IY Ts sevperPoel eeilyeline gas g006 .0Gacddme sod Jos Do Qob OU gU DO Usd OU ODDS abanosanosoL 220 Pee behaworund ernormMa conditions cect stecclnnerta veer eeicieceiaiseeitrtsee iene seatt ere 220 ZeeBebavion atthe Onsenon GestccatlOnls-t-\ctete\ateln\oleforstatelel=lotaisielaleioteiaielete rai otetl oleate iterate 222 V_ Visible changes attending a process of desiccation. ..........eeeeee cece cence tee eeeeeees 223 Wil Depreeotdesicedtionia ttained byy Philo cinia re jarereseinpeser plela) -Ve wlohe inte lel ele/=le\sieleisl=)ol-ti-fele stot iol is 227 AME Heme Denmservel@ydelaess od adosondondoncdoobocs Smoncangssoanemsounmecec 228 2 Evidence for the view that a true desiccation occurs, 2 cae fave sage eyeteds ee ateneleieiePetohe ol etree crore 231 VII _ Effect of desiccation on the life processes of Philodina..............esesceeeee cece eeeees 236 Mpa RIA We efte CES Me SICCATIO Ler ctere)=belalayeisiale) 9) leleleyslarelelehalsteter=yayetetet stele mtet-tear tf ieee 236 2 Influence of the previous condition of the animal.................. cece eee e neces 237 3 Influence of the conditions attending the desiccation.............-.++eeeee seen eran 238 7 DACP Oh ge ADiGba sian wopooGaE Ob docs Ran dboDedopousoeSsobCeodner ao7os 238 b Effect of temperature at which drying DECUIS saiereieleca\eierencforetec siete settee toe 242 G) enectotduratronofdesiccation. cereale sleet) cle/sier= os -tajsl-foletelaeieie ele airtel 244 d Effect of alternations of moisture and dryness. .................02.0ccececeees 244 Cm PiechOmimtensibyon CestecatiOreyemrerc jeter Pelelsfereya) foto rniate -)erela erate ctetete lteter 246 f Effect of temperature on mortality.............. sees tee c secs e eee e eee e eee ees 248 4 Conditions of the life processes in a dried animal................00eeseeeeeeeeeeeee 249 5 Effectof desiccation on egg production... .. ..-.- 06. -- 25621 sews ieee eee mee 252 VALLE Siinmieny Anatosdsomeaas sedan acbunovopoccacD Dub comsunnud ead boOsPDoDosocmecdo day cc 260 IDG “1d UGFar tol shAapedcanopeonepaenaur obs o eden doDn coo coeordacuacuseaneesd sue obaueccons” 262 I INTRODUCTION It has long been known that various animals, among them certain of the rotifers, tardigrades, and the smaller nematode worms, can survive conditions fatal to most other organisms. Although normally living in water, or at least under conditions of moisture, these animals may be dried for long periods of time without serious injury. ‘They are found in countless numbers in Tue JourNAL or ExPERIMENTAL ZOOLOGY, VOL. VI, NO. 2. 208 Merkel Henry ‘facobs the dust from the gutters of roofs and other places where water is accustomed to stand at intervals, and even after prolonged drying the addition of a little water is all that is necessary to start them into activity. While in the dried condition they show no visible signs of life. All movements cease and the body frequently shrinks to a shapeless mass so that it is difficult to distinguish them from the particles of sand among which they are found. When water is again applied the body gradually regains its original form, movements appear, and after a longer or shorter time, depending on the conditions of the desiccation, the animals re- sume their normal activities apparently none the worse for the experience. Not only may they be dried naturally in the air but they may be subjected artificially to even more extreme conditions. Various observers have kept them in desiccators and vacua for long periods of time and have subjected them to temperatures at which life ordinarily is impossible, without destroying their power of again resuming their normal activities upon the application of water. The length of time the animals may remain in this state of sus- pended animation is often considerable; there are well authenti- cated cases of rotifers, whose usual period of life is probably not more than a few weeks or months at most, which have been revived after a period of desiccation extending over three or four years; one observer even claims to have revived them after fifteen years’ desiccation. In the case of the Anguillulidz even longer periods have been recorded; Baker in 1771 succeeded in reviving individuals of Tylenchus scandens which had remained in a dried condition in grains of wheat for 27 years. As to the principal facts just given there seems to be little doubt. They have been confirmed by numerous observers and anyone may repeat for himself with little trouble the experiments on which they are based. In the interpretation of these facts, however, there has been, and still is, much diversity of opinion. What is the actual effect of drying on the rotifer or tardigrade? Is the water contained in its tissues really removed or does the animal have some means of protecting itself against the loss of water, the desiccation being only apparent? If the former be the case, what Effects of Destccation on the Rotifer 209 is the condition of the dried animal? Are its life processes merely retarded or have they come to a complete standstill? In the revival of a dried animal by the application of water, are we deal- ing, as many have supposed, with a case of passage from death to life or merely with an acceleration of vital processes which have been continuing all the while but in a greatly reduced state? These and similar questions have been under discussion for many years and as yet no unanimity of opinion has been reached by zodlogists in regard to them. Every point has been affirmed and denied many times by equally capable men. Much of the discussion on the subject has been pure speculation based on neither observation nor experiment and hence is of little value; however, even the most careful observers have differed radically with each other on many points of importance, the observations of one worker being contradicted by the apparently equally accu- rate observations of another worker. ‘The result is that the ques- tion even at the present day is in a state of the greatest confusion and uncertainty and is still far from being finally settled. It was with the intention of clearing up some of these points of dispute that the present piece of research was undertaken in the fall of 1906 at the suggestion of Prof. E. G. Conklin. It gives me great pleasure to express at this point my sense of deep indebted- ness to him not only for suggesting the subject but for the inter- est he has taken in the workeand for the many helpful suggestions and criticisms he has offered. ‘The experiments were performed in the Zoological Laboratory of the University of Pennsylvania during the years 1906-1907 and 1907-1908. During the course of the work a number of points of interest somewhat off of the main line of the investigation came up; some of these points are touched upon in the present paper, others are still under investi- gation and are reserved for subsequent publication. The animal worked upon was Philodina roseola, one of the Bdelloid rotifers, and in all cases except where otherwise expressly stated it will be understood that the observations apply to this form. Philodina was chosen partly on account of the ease with which it could be obtained, making experiments on large numbers of individuals possible, and partly because it shows the phenom- 210 Merkel Henry ‘facobs enon of revival after desiccation in a particularly favorable way. A limited number of observations were made on Adineta (Calli- dina) vaga, a nearly related form, but these show practically no points of difference from those made on Philodina. II HISTORICAL In the year 1701, Anton von Leeuwenhoek in searching for new objects to examine under his microscope chanced to take some of the dry dust from the gutter of a roof and on moistening it was greatly astonished to see after a time living animals swim- ming about actively in the water. Struck with the observation he again allowed the animals to dry and on moistening them the next day with water previously boiled by way of precaution against introducing any life from outside sources, he obtained the same result. Further experiments convinced him that these ani- mals, which were rotifers, probably belonging to the species Roti- fer vulgaris, might be deprived of water for at least several months without losing the power of recovering their normal activities when water was again supplied. This was the first observation to be made on the phenomenon of desiccation with subsequent revival in animals; strangely enough it attracted little attention at the time.- Leeuwenhoek believed that the animals themselves were not truly dried but that they were protected by an impene- trable cuticle from loss of water * * * “‘cuticulas ex tam solida conflatas esse materia ut ne miniman quidem permittant exhalationem. Quod si sese aliter haberet, asserere non vereor haec animalcula * * * onni aqua destituta necessario omnia esse emoritura.”’ | The next mention of the subject was made in 1743 by an Eng- lishman, Turbervill Needham, who observed minute worms to issue from grains of wheat when water was applied to them. To use his expression they “took life” on the application of moisture. In the same year another Englishman, Henry Baker, again called the attention of naturalists to the rotifers discovered by Leeu- wenhoek, but he contented himself with merely repeating that writer’s description of them and it was not until ten years later Effects of Desiccation on the Rotifer 211 that he discussed the subject on his own account. Needham’s discovery excited considerable attention but was for the most part received with incredulity. That living organisms could be dried and after remaining apparently lifeless for a time be caused to “take life” by the application of water seemed so incredible to the scientists of the day that they preferred either to deny the animal nature of the worms altogether, calling them “filaments animés,”’ “fibres mouvantes ” or “ etuis pleins de globules mobiles” which were started into movement by the imbibition of water or to assert that they arose by spontaneous generation. It was not until the time of Fontana (1771) and Roffredi (1775) that the ani- mal nature of these anguillulids was established beyond all doubt. In the meantime other men were making similar observations. Among the earliest of these may be mentioned Trembley (1747), Baker (1753), Schaeffer (1755), Ginnani (1759), Ledermiiller (1759), Fontana (1768), Goze (1772), Corti (1774), Miller (1775) and Roffredi (1775). The list of the animals capable of enduring desiccation was also increased during this period. Leder- muller in 1759 had observed the revival of “paste eels,” Fontana in 1768 of Gordius, and Spallanzani in 1776 of tardigrades and the anguillulids found in the dust of roofs. Later observers added several new forms and increased the list of rotifers and tardigrades which show this peculiarity. The history of our exact knowledge of the phenomenon of sus- pended animation dates from the publication in 1776 of Spallanzani’s “Opuscoli di Fisica Animale e Vegetabile.”” The section of this work relating to the desiccation of animals is a model of scientific research. Spallanzani, unlike most of his contemporaries, tested his theories by actual experiments and these experiments were often carried out with great care and considerable ingenuity. If other workers had used the same amount of care the subject would be in a far less confused state than it is at the present day. Spallanzani first repeated the experiments of Leeuwenhoek and made the observation, which has been confirmed by a number of subsequent workers on the subject, that rotifers can recover their activity after a period of drying only when a certain amount of sand or moss is present; when dried on a clean glass slide they are 212 Merkel Henry ‘facobs invariably killed. He considered this result to be due to the inju- rious effect of the air on the dried animals, the sand protecting them from its action. He made many other interesting observa- tions, for example, on the effect of high temperatures on animals in different degrees of desiccation, on the number of times they may be dried without being killed and the effect on them of vari- ous chemical substances. Unlike Leeuwenhoek he believed that the rotifers could endure the withdrawal of the last traces of water from their bodies. He was able to revive them after subjecting them while in the dried state to the desiccating action of high temperatures and the vacuum. ‘There was no doubt in his mind that the process of desiccation caused an actual stoppage of the life processes and that the animals in the dried condition were to be considered as dead. He entitled his paper, “Observations and experiments on certain marvelous animals which the observer can at his will make pass from death to life.”’ Spallanzani’s work naturally excited great interest and gave rise to much discussion. ‘The statement that animals could be brought back to life after being dead for a time could not be allowed to pass unchallenged oy the physiologists of the day and many heated controversies arose between those who supported Spallanzani and those who opposed him. Among those who engaged in the discussion may be mentioned such. distinguished naturalists as Haller, Cuvier, Oken, Humboldt, Lamarck, Tre- viranus, and Johannes Miller. The controversies for the most part, however, were carried on on purely theoretical grounds and had little experimental evidence to back them. From the time of Spallanzani to the time of Doyére, a period of sixty years, although the question continued to be one of the most discussed in the whole realm of physiology, practically no new facts were added to our knowledge of the subject. On the whole, the opponents of Spallanzani seemed to gain the upper hand. ‘This was largely due to the apparently convincing arguments advanced by two men, Ehrenberg in Germany and Bory St. Vincent in France. Both of these workers not only denied that dead rotifers could be brought back to life but asserted that “recovery after true desiccation was impossible. The unde- Effects of Destccation on the Rotifer 253 niable fact that rotifers, tardigrades, and anguillulids may be obtained from apparently dry sand they explained in different ways. Ehrenberg, on the one hand, contended that actual dry- ing does not occur, the sand protecting them from loss of water “as a woolen mantle protects the Arab from the intense heat of the desert.” He believed that in this state of apparent desicca- tion all of the vital processes continued, even reproduction. Bory St. Vincent, on the other hand, believed that the animals in ques- tion could not survive a period of drying even when sand was present and that their apparent revival was due to the hatching of eggs concealed in the sand. ‘‘Nous avons quelquefois * * * retrouvé des Rotiféres * * * mais ils n’y ressuscitaient pas; ils s’y développaient comme les Daphnies et autres petits ento- mostracés, dont les ovules sont demurés dans le sol.” The next worker to take up the subject in a scientific manner was the French naturalist Doyére. So great was the authority of Ehren- berg and Bory St. Vincent and so plausible their reasoning that he had been led to doubt the accuracy of Spallanzani’s observations or at least to regard the matter as worthy of further investigation. He accordingly undertook a series of experiments, published in 1842, with the result that Spallanzani’s conclusions were in the main confirmed. He found, however, that the animals are not always killed in the absence of sand, a certain proportion of those dried on a clean slide recovering, although requiring a much longer time. Furthermore, he showed that it is not the exposure to the air that injures the rotifers and tardigrades in this case as Spallanzani had supposed, since animals dried in the air and then placed in a vacuum showed a lower mortality than those dried directly in the vacuum. He concluded that the rapidity of drying is an important factor in the effect of the desiccation. He furthermore showed that rotifers may be revived after an apparently almost perfect desiccation. He found that they could endure a sojourn of 17 days in a desiccator followed by 28 days in an air pump with a pressure of 5-6 centimeters of mercury and that after thorough drying in the sunlight a temperature of 140° C. or more could be resisted for a brief period. He concluded, therefore, that the last traces of water might be extracted without destroying the power 214 Merkel Henry “facobs of revival, and that, since life processes are impossible in the absence of water, in the dried animal we are dealing with a case of life in potentia as opposed to life zm actu. ‘These contentions were supported by observations made several years later by the physicist Gavarret, who subjected rotifers to the action of a vacu- um of only 4 mm. pressure for 51 days, sulphuric acid being present to absorb all traces of free moisture, and yet was able to revive them by the application of water. About the same time Pouchet, followed by Pennetier and Tinel, obtained results which were diametrically opposed to those of Doyére. Numerous carefully conducted experiments led these naturalists to believe that a true desiccation is invariably fatal in the case of rotifers, tardigrades, and anguillulids just as in other animals. ‘They found in all of their experiments that these ani- mals when dried on a glass slide either with or without a small quantity of sand are killed in the course of a few days even at ordi- nary temperatures and that individuals dried under more natural conditions cannot resist an hour’s exposure to a temperature of too° C. Pouchet even found that rotifers and tardigrades obtained in a dried condition from natural sources were all killed in three months when exposed to the air in a sunny place. All of these results were presumably due to the loss of water by the animals when exposed to unfavorable conditions. Since the views of Pouchet, Pennetier, and Tinel on the one hand and Doyére and Gavarret on the other, differed so widely and since both seemed to be based on equally careful experiments, it was decided, instead of wasting time in fruitless discussion, to submit the matter to arbitration. Accordingly a commission was appointed in 1859 by the Société de Biologie consisting of MM. Balbiani, Berthelot, Brown-Séquard, Dareste, Guillemin, Ch. Robin, and Broca, chairman, to hear the evidence presented by both sides, to perform any experiments of their own they might consider necessary, and to present the result of their deliberations in the form of a report to the society. This commission dis- charged its duty with the greatest thoroughness. It held forty- two regular sessions without counting the times when a few of its members met for discussion. The examination of the evidence Effects of Desiccation on the Rotifer 205 presented by both sides in the form of experiments occupied several months and the final report of the commission presented in March, 1860, covers a hundred and forty printed pages. ‘This report on the whole sustains the contentions of Doyére, not because of any inaccuracies in Pouchet’s results, but because they are largely negative in nature while Doyére’s are positive. It concludes with the words, “des animaux * * * aménés au degré de dessic- cation le plus complet qu’on puisse réaliser dans |’état actuel de la science, peuvent conserver encore la propriété de se ranimer au contact de l’eau.” Although the language employed in the report is rather guarded, it evidently was the opinion of the commission that life under these conditions could exist only 17 potentza. The results obtained by this commission as well as those of Doyere and Spallanzani are contradicted by the observations of some of the more recent workers on the subject. In 1873 H. Davis, a member of the Royal Microscopical Society and a student of rotifers published a paper which has since been largely quoted and in which he stated his conclusions that in Philodina, at least, the desiccation when followed by revival is only an apparent one, the animal being able to protect itself by means of a gelatinous secretion against the loss of its body fluids. He demonstrated how such a secretion could be effective by showing that grapes covered with a thin coating of gelatine remained in a juicy condi- tion for a long time even in the vacuum of an air pump while grapes not thus treated soon lost their water and assumed a shriveled appearance. ‘The fact that rotifers dried with a small quantity of sand are capable of recovery even after a prolonged exposure to extreme conditions while those dried on a clean slide for a much shorter time are not, had always been more or less puzzling to naturalists since it was first observed by Spallanzani. It received a ready explanation on Davis’ theory, it being assumed that in the former case where drying was slow the animal had time to protect itself by pouring out a secretion, while in the latter case drying was so rapid that the animal, not being able to form its usual protection, was killed. “This explanation was so simple and so in accord with the facts known for other animals that it was infmediately accepted by many and at the present day is still frequently quoted. 216 Merkel Henry “facobs Certain of the most recent investigators have gone even far- ther than Davis. O. Zacharias (’86), for example, on the basis of several observations made on a species of Philodina has tried to discredit the whole subject of the revivification of desiccated animals. According to him, rotifers can resist drying no better than other aquatic animals; when withdrawn from water they are invariably killed. Like Bory St. Vincent he considers the many cases of supposed revival to be due to the hatching of eggs which by virtue of their thick shells are able to resist drying the same as the eggs of insects or other animals. ‘The fact that revival is commonly supposed to occur in the presence of sand he regards as strong confirmatory evidence since it is to be supposed that concealed in the sand there might easily be many eggs which would escape the notice of the observer. He concludes his paper with the words * * * “es einzig und allein die Ever sind durch welche die continuierliche Generation ensfolge aufrecht erhalten wird.” F. Faggioli (91) as the result of a number of observations and experiments on several species of rotifers comes to the same con- clusion. Like Zacharias he regards the stories of the revival of animals submitted to conditions of desiccation as myths based on imperfect observations. Perhaps his position may best be summed up in the following quotation from Fredericg (’89) which he cites in his paper. “Les Rotiféres et les Tardigrades adults meurent sans retour quand on les desseche. Mais les oeufs qu’ils ont generalement dans le corps ne sont pas dans le méme cas. Ces oeufs conservent leur vitalité malgré l’absence d’eau. Places ensuite dans un milieu convenable et humide ils se développent avec rapidité et donnent naissance a une nouvelle génération de jeunes animaux que’lon avait a tort considérés comme résultant de la réviviscence du corps de leurs parents.” It is seen from this short historical review that the entire subject is in a very unsatisfactory state. On scarcely any point is there any general agreement. All observers, perhaps, will admit that it is possible to obtain living animals from sand that 1s apparently dry but further than this there is no consensus of opinion. One body of observers represented by Bory St. Vincent, Zacharias and Effects of Destccation on the Rotifer 217 Faggioli claim that the animals come from eggs concealed in the sand in the dead bodies of the parents, the latter having of necessity been killed by the exposure to the conditions of dryness. Another body, represented by Leeuwenhoek, Ehrenberg, Pouchet, Davis, and Hudson, although admitting that adult rotifers may survive even prolonged conditions of desiccation, consider that they are able to do so only in virtue of some effective means of preventing the escape of water from their tissues, the condition of the animal in the dried state therefore not differing in any essential respect from its normal state. Still others, among them Spallanzani, Schultze, Doyere, Gavarret and Preyer consider that the desicca- tion is a true one and may continue theoretically up to a point of absolute dryness without injury to the animal. ‘They either state specifically that the animal in this condition is lifeless or tacitly assume such to be the case. Other observers, while admitting that desiccation may proceed very far, consider that the life pro- cesses never come to a complete standstill although they may be very greatly retarded. It is the purpose of the present paper to consider the evidence for these various views and to attempt to reduce the subject to a state of greater certainty than that in which it now rests. The two points which will chiefly be considered are first, whether or not the animal suffers a true desiccation and, second, what is the state of the vital processes of the Animal in the absence of water. Before proceeding to the observations and experiments on which the conclusions of the paper are based, a brief account will be given of the structure, natural history, and behavior of Philodina to render more intelligible that which is to follow. III STRUCTURE AND NATURAL HISTORY OF PHILODINA Philodina roseola is one of the commonest of the rotifers, be- longing to the order Bdelloida of which the common rotifer, Roti- fer vulgaris, is also a member. It is widely distributed through- out the world, being frequently found in small basins in the rocks in which water periodically collects; a favorite habitat is in the stone urns in cemeteries. It is usually associated with the uni- 218 Merkel Henry “facobs cellular alga, Spzhrella lacustris (Hzmatococcus pluvialis), on which it feeds and from which it appears to derive the red color which gives it its specific name. In size, it is microscopic, being barely visible as a minute speck to the naked eye. Adult individ- uals when fully extended measure from 0.35 to 0.5 mm. in length and when contracted from 0.1 to 0.15 mm. ‘The newly hatched young are about five-eighths as long as the adults which they resemble closely except in being colorless and in having a more transparent alimentary tract. The members of the group Bdelloida are peculiar among the rotifers in their method of locomotion. In addition to their swim- ming movements which resemble those of other rotifers, they are able to creep over the surface of solid objects like a leech, hence the name applied to them. ‘These creeping movements are ren- dered possible by the telescopic structure of the body, which may readily be extended or contracted, and by the presence at its ante- rior end of a so-called proboscis by means of which adhesion to solid objects is possible. ‘The proboscis of Philodina is fairly stout and when the trochal disc has been withdrawn within the body, as it always is when the animal is creeping, it appears to form the anterior end of the body. When the trochal disc, which consists cf two ciliated lobes which the earlier observers mistook for wheels, is extended, the proboscis is drawn back so as to lie behind and dorsal to it. In creeping, the animal first fastens its foot by means cf a sticky substance secreted by the cement glands contained in it; the proboscis is then attached and the foot drawn forward, the body shortening and arching itself, like that of a leech or a “‘meas- uring worm.” ‘The foot is then again attached and the body extended. These movements may occur with considerable rapid- ity although the progress of the animal is not so fast as when swim- ming. As a rule a high temperature and fresh water favor swimming movements and low temperatures and foul water creep- ing movements. Externally Philodina is covered with a fairly thick cuticle which in the regions of the head and foot is divided into a number of seg- ments by alternate stiff and flexible portions. By an inrolling of the flexible regions the stiffer parts are allowed to fit within one Effects of Destccation on the Rotifer 219 another and a sort of telescoping of the body is thus permitted. When fully contracted the head and tail are drawn entirely within the large segment which covers the middle of the body and both ends of the latter are puckered together as if drawn in with a string. The animal in this condition is shaped like a lemon and is well fitted to resist the injurious effects of desiccation. The part of the cuticle that is outermost, that is, that which surrounds the middle of the body, is thicker than the parts that cover the head and tail. It is also somewhat different in chemical nature as is shown by the fact that when living animals are placed in a weak solution of methylene blue it is only part of the body that takes the stain. ‘his is true either of contracted animals or those that are creeping or swimming. In the latter case a blue band appears about the middle of the body, the head and foot remaining color- less. ‘The fact that that part of the cuticle which alone is exposed at the time of drying-should be of a different nature from the re- mainder is probably significant. Perhaps it may be more imperme- able to water and thus check the rapidity of evaporation. Whether or not it is a complete protection will be considered in another place. Within the cuticle and the hypodermis which secretes it are found the muscles; these are especially well developed in Phil- odina on account of its habit of contracting and extending the body. ‘They are both longitudinal and circular, the former sery- ing to draw the head and foot together at the time of contraction and the latter by exerting a pressure on the fluids of the body to cause their extension. In a rotifer dried slowly enough to retain its natural form, the position of these muscles may be observed in the form of slight thickenings beneath the cuticle. The various internal organs are much the same as those of other rotifers. [he mastax is well developed and the walls of the stomach are thick and glandular. ‘They usually appear more or less yellowish, greenish, or sometimes a deep brick red, depending on the amount of pigment present in the cells of Sphzrella on which the animals are feeding. “The nephridia are fairly conspic- uous and under the higher powers of the microscope it is easy to observe the peculiar flickering movement of the flame cells. A contractile vesicle is present into which the nephridia discharge their products. 220 Merkel Henry “facobs The method of reproduction of Philodina is of interest. “There are two ovaries, one on each side of the alimentary canal, and each is combined with a large vitellarium. ‘The eggs undergo a partial development within the body of the mother and at the time of laying are relatively large and are protected by fairly thick shells. The point of special interest in connection with the reproduction of Philodina is that there are no male individuals; this is also true of the other members of the group Bdelloida. Although these animals are among the commonest of our rotifers and have been under observation for over two hundred years, males have never been observed, and it seems reasonably certain that they do not exist. Parthenogenesis appears to be the only method of repro- duction. The eggs hatch two or three days after being laid and the young from the first behave exactly like the adults which they closely resemble in most respects. “They seem to be somewhat less resist- ant to desiccation, but there is much individual variation and the difference is not very great in any event. ‘They feed freely and erow rapidly; egg laying may begin seven or eight days after hatching at which time they have not yet reached their full size. As to the length of life of Philodina under favorable conditions, unfortunately nothing definite can be said except that it 1s con- siderably longer than has been supposed in the case of other roti- fers. Hydatina is said to live about two weeks; in the course of these experiments rotifers isolated at the time of hatching have been kept alive for more than six weeks, and under more favorable conditions they would probably have lived still longer. IV THE BEHAVIOR OF PHILODINA I Behavior under Ordinary Conditions The behavior of Philodina is more complicated than that of most rotifers owing to the two different modes of locomotion. Ordinarily the animals are found in one of the following four states, (1) swimming, (2) creeping, (3) attached by the foot and feeding or (4) contracted. The causes which determine the succession Effects of Desiccation on the Rotifer 221 of these states in a given individual have not yet been worked out in detail although the problem presents a number of points of interest. A swimming individual behaves in much the same man- ner as the forms described by Jennings (04); orientation in response to any stimulus takes place by random movements resembling those of an infusorian. In a creeping individual, on the other hand, orientation is much more direct; the process resembling that which occurs in a planarian. Any strong sitmulus causes the animal to stop and make a few testing movements with its proboscis; orientation either towards or away from the stimulus then occurs directly without the intervention of random move- ments. If the stimulus be very strong, the animal contracts at once, this reaction being caused by a variety of stimuli such as heat, injurious chemicals, hypertonic solutions, mechanical shocks and the onset of desiccation. It is interesting to note that in this one animal we have at one time the method of reaction of an infu- sorian and at another that of a typical metazoan. As has already been mentioned, the temperature and the purity of the water seem to be two of the factors which determine whether creeping or swim- ming shall prevail, although probably they are not the only ones concerned. Feeding movements may occur under almost any conditions; they are especially marked after a period of desiccation. The reaction of Philodina to light is of some interest. Normally itis almost indifferent. When a number of individuals are pres- ent in a dish they tend to become scattered to all parts of it regard- less of the direction of the source of light. If now they be dis- turbed, either by jarring the dish, drawing the water through a pipette a few times, or adding a few drops of fresh water, they immediately move towards the side of the dish away from the light, either creeping or swimming according to circumstances. ‘This reaction is very striking and never fails to occur. In the course of these experiments it was put to a practical use whenever it was desired to obtain rotifers free from sand. A small amount of the sand from the large culture was placed in a small dish and the water agitated with a pipette, being given a spiral motion so as to carry the sand to the center of the dish. In a few moments most of the rotifers had invariably collected in the clean water on the 222 Merkel Henry “facobs side of the dish away from the light where they could be removed by means of a fine pointed pipette. If left undisturbed for thirty or forty minutes they again spread out into all parts of the dish. 2 Behavior at the Onset of Destccation At the onset of desiccation, no matter under what circumstances it occurs, the animals all show a general restlessness. If they have been quietly feeding, for example, they withdraw the corona and begin to make active creeping movements. ‘These movements are more or less at random and thus differ from the responses made when creeping, to more definitely localized stimuli. ‘The negative phototaxis, which is shown so markedly under certain conditions, is always abandoned when the water begins to dis- appear. A group of rotifers which has collected on the side of the dish away from the light is quickly broken up, each individual starting to creep in the direction in which it happens to be turned at the time and continuing in this direction until it comes either to the edge of the drop or to a place where the surface film lies so close to the slide that it cannot creep under it. It then stops, contracts, makes one or two testing movements with the proboscis, and starts off again in a new direction until brought to a stop in a similar way. Contact with solid objects such as grains of sand does not cause the creeping movements to cease as Davis and others have assumed. So far as can be observed the behavior of rotifers dried on a clean slide and those dried with a quantity of sand is essentially the same. In both cases they continue to creep as long as it is possible to do so and when such movements become impossible they contract into a more or less spherical mass and dry wherever they happen to be. As has been mentioned the movements are purely at random. Frequently a number of rotifers may be seen in the field of the microscope at once, all creeping in different directions. When two rotifers meet they pay no attention to each other but continue on their way just as before. There is no instinct that leads them to gather together in groups for mutual protection as Hudson and others have asserted. Where such groups are found the cause is merely that they have all re- Effects of Destccation on the Rotifer 223 mained in the evaporating drop as long as possible thus being brought together as the latter decreased in size. When drying occurs under a cover glass or with sand, the contracting surface film of the drop may frequently be seen to sweep a number of rotifers together and by further contraction press them into a com- pact mass; the part the rotifers themselves play in the process, however, is purely a passive one. On the whole, the method of response to the first stages of des- iccation shown by Philodina is an advantageous one. It is true that the random movements sometimes lead single individuals into places where they are caught by sudden evaporation of the water and dried without any protection at all. On the other hand, the majority of the rotifers, by continuing to move as long as pos- sible, tend to find their way to the places where the water lingers longest and where drying, when it does occur, is most gradual. If they stopped at the first grain of sand they encountered they would frequently be dried under far less advantageous circum- stances. V VISIBLE CHANGES ATTENDING THE PROCESS OF DESICCATION When the water has so far evaporated that creeping is no longer possible, the rotifer contracts in the manner already described, drawing in both the head and the foot and then puckering the two ends of the body as though they were drawn together by a purse string. [his part of the contraction is accomplished by muscular action and occurs before the water surrounding the animal has evaporated. Before the final contraction occurs the head may be rapidly extended two or three times as if to test the external con- ditions. When the process of drying is very slow contraction may continue still further, the animal becoming noticeably smaller even before all of the water has disappeared. Under these conditions irregular wrinkles do not appear although the puckering at the two ends is discernible. When dried more rapidly, even when in contact with a large quantity of sand the animal assumes a more or less irregular form, both longitudinal and transverse wrinkles appearing in the cuticle and the internal organs shrinking away 224 Merkel Henry “facobs from the latter. Although the animal is not necessarily injured by this irregular shrinking it is more likely to be than in the case where it contracts regularly. Loss of water from the body is very rapid after the last trace of the surrounding film has disappeared. Wrinkles appear almost immediately in the case of individuals dried on a clean slide. In one minute the body may have become fairly irregular in outline and in two minutes assumed the characteristic appearance of desic- cation. Although further loss of water and shrinkage occur the change in size and appearance after two or three minutes is not very noticeable since the cuticle has assumed its final wrinkled form and further contraction is confined to the internal organs. In rotifers dried more slowly the contraction is more evident since the cuticle and hypodermis are pressed together against the inter- nal organs and follow them in their shrinkage. That muscular action is responsible for at least a part of the contraction is shown by the fact that dead rotifers when allowed to dry assume very irregular forms, the closing of the ends of the body above described not occurring. ‘The amount of shrinkage that occurs under favorable conditions is very considerable. It may easily be followed by keeping single individuals under the microscope during the drying process and making camera drawings of them from time to time. In those that are dried slowly and thus contract regularly it is easy to com- pare approximately the volumes at the different times. Such an examination shows that in the case of slow desiccation with sand, which is one of the most favorable methods, the animal may shrink to one-third or one-fourth of its original volume. When water is added to rotifers that have been dried for a time a very rapid swelling occurs. In one minute the animal has dou- bled in volume and in five or often less has reached its normal size. From this time until movements appear the visible changes are slight and consist in changes of form apparently due to the slow relaxation of the muscles that have drawn in the ends of the body. When those at the anterior end relax the animal becomes pear shaped; the foot may be extended either before or after this occurs. In rotifers which have been dried rapidly and which are therefore Effects of Desiccation on the Rotifer 225 much wrinkled, the first change that occurs is a filling out of the wrinkles, the amount of swelling apparently being slight since the greater part of the shrinkage has occurred in the internal organs, the cuticle having stuck fast to the slide or to a grain of sand. That a considerable swelling occurs, however, is easily observed when a group of rotifers has dried together, the individuals being pushed apart as the water is absorbed. ‘The usual time required for all of the wrinkles to disappear is about five minutes, although it may frequently be considerably less. The point at which the water enters may be determined by staining rotifers zmtra vitam with neutral red before drying them and then adding water which has been made weakly alkaline by the addition of a small amount of sodium bicarbonate. As the latter touches the organs which have been stained red it changes them to a bright yellow color and thus furnishes a very exact means of observing the method of penetration. Such an experi- ment shows that the water enters largely at the two ends of the animal and to a much less extent through the cuticle surrounding the body. Its entrance is practically instantaneous. The time required for movements to appear varies greatly with the circumstances attending the process of desiccation. ‘The times required under a number of different conditions are given in another section of this paper. Under the most favorable con- ditions they may appear in five minutes, as was the case with a number of individuals which had been dried slowly on filter paper; usually seven to ten minutes are required and sometimes an hour or more. ‘The first movements to appear are muscular contrac- tions within the body, although sometimes the flame cells of the nephridia begin to beat before muscular movements can be ob- served. In any case the beginning of the activity of the nephridia is one of the first signs of the return of the normal life processes. Care must be taken not to confuse with true muscular movements the jerky movements exhibited immediately after water has been applied which are due only to changes in tension set up by the imbibition of the water and are not indicative of life. In all cases where the time required for movements to appear is mentioned it is understood that it is the distinctively vital movements that are meant. Soon after the first movements have appeared the foot 226 Merkel Henry “facobs is extended, often with great suddenness. ‘This is apparently accomplished by the contraction of the circular muscles of the body which thus set up a pressure which causes it to be protruded. Even in individuals which have been killed by drying the foot becomes more or less extended but this is probably due to the absorption of water by osmosis. After the first extension of the foot, movements separated by periods of rest may recur at intervals for some time before the animal fully recovers and creeps away. Complete recovery some- times occurs in as short a time as ten minutes, but under certain conditions it may require several hours or even a day. Most workers have introduced a source of error into their observations by failing to keep their rotifers for a long enough time. In the present series of experiments, the animals were always kept at least twenty-four hours before being pronounced dead. After the rotifers have recovered they may show the effects of the desiccation in various ways. Sometimes the body is more or less crooked and distorted, at other times the cilia appear injured and do not beat normally. In other cases the animals remain con- tracted and make no movements although they are not dead as is shown by the fact that they undergo no disintegration. ‘That some of the life processes such as oxidations still continue may be shown by placing upon them a drop of a solution of neutral red which has been rendered weakly alkaline by a very small quantity of sodium bicarbonate. Living rotifers which are excreting car- bon dioxide change the color of the solution and are rapidly stained pink or red; dead ones do not show this reaction. A complete transition is therefore shown from individuals which are killed through those in which movements never appear but some of the life processes continue for a short time and those which have been more or less injured to those which completely recover and show no visible signs of injury. The facts just mentioned may be considered as an answer to the views of those who doubt the whole phenomenon of recovery after desiccation, among whom Zacharias (’86), Fredericq (’89) and Faggioli (91) may be named as some of the more recent workers. The evidence advanced by these men is purely negative in char- Effects of Desiccation on the Rotifer 227 acter. In certain cases they failed to observe the recovery of dried rotifers and they, therefore, conclude that such recovery is impossible. On the other hand, many observers have con- tributed an abundance of positive evidence, of which that just given is typical. It is possible for anyone, by using the proper pre- cautions, to dry rotifers and see for himself the gradual return of movements up to the time of complete recovery. It is certain, therefore, that the return of all of the normal vital activities in animals which by drying have been rendered motionless and apparently lifeless is not one of the myths of zodlogy as many have supposed. ‘This does not imply that the rotifers which have been subjected to the conditions of dryness have actually undergone a true desiccation. From the evidences so far presented it is possible that they have not. ‘This point will be discussed more in detail in the next section. It is necessary to observe that all rotifers are not equally resist- ant to a process of drying. ‘The only well authenticated cases of recovery after desiccation are found in the family Philodinidz (Philodina, Rotifer and Adineta) in species which normally live in places where they are exposed to drying at frequent intervals. Perhaps other rotifers possess this power, but that such is the case has not been proven. A number of experiments made to deter- mine this point with Megalotrocha and several of the Loricata gave purely negative results, no cases of recovery being observed. The observations of Lance (94) are of interest in this connection. He found in the tardigrades that only those species could resist desiccation which live in places where they are exposed to it under natural conditions. VI DEGREE OF DESICCATION ATTAINED BY PHILODINA One of the most disputed points in connection with the entire subject of suspended animation in rotifers and other aquatic ani- mals is the degree of desiccation they may attain without injury. Some workers hold that there is no limit to the amount of drying that may occur, while others just as strongly contend anything resem- bling a true desiccation is necessarily fatal. They believe that in 228 3 Merkel Henry Facobs all cases in which recovery occurs the loss of the body fluids has in some way been prevented. Among those who have held this latter view, perhaps the most important name is that of H. Davis (?73) who was the first to give a plausible explanation of the means by which desiccation could be prevented. Since his views have found a wide acceptance and are frequently quoted at the present day, they will be considered at some length and an attempt made to determine their truth or falsity. tr The Views of Davis and Others Davis as well as many others before him, had noticed that roti- fers dried on a clean glass slide are killed by the process while if a little sand or moss is present recovery almost always occurs. He supposed that in the latter case the rotifers had time to protect themselves by secreting a gelatinous waterproof cyst which effec- tually prevented evaporation while in the former case, having no protection, they were dried and consequently killed. According to his view a true desiccation of the animal is always followed by death and in all cases of recovery desiccation has been prevented. He showed that grapes may be effectually protected against drying by means of a thin coating of gelatine and assumed that in the case of rotifers exposed to desiccation we are dealing with a phe- ‘nomenon somewhat similar. Others had claimed that the sand has a direct protective effect. Ehrenberg, for example, had compared it to the woolen mantle of an inhabitant of the desert and others had considered that it always holds in its interstices sufficient water to prevent complete desiccation. But Davis and Hudson rightly objected that sand at 100° C., at which temperature dried rotifers may survive for a short time, would be but poor protection against loss of water in the case of such a soft bodied animal, and Davis’ explanation that the rotifers are covered by an actual waterproof capsule has the merit of being the first one to take this fact into consideration. Nevertheless, there are many reasons why this view cannot be accepted. In the first place the evidence for the existence of such a gelatinous secretion is extremely slight. Davis did not see it Effects of Destccation on the Rotifer 229 clearly himself and merely states in support of this view that several of his correspondents after reviving dried rotifers observed a quan- of sticky material adhering to them. ‘This is not surprising since the cement glands at the tip of the foot are continually secreting a sticky substance by means of which the animal fastens itself to solid objects and which under certain conditions may be poured out in considerable quantities. The mere presence of a small amount of this substance seems like rather slender evidence on which to base such far-reaching conclusions. Hudson (’8g) is more definite in his statements. In describing the appearance of a group of rotifers dried on a piece of paper he says, “Each Philodine is the center of a patch of glutinous secre- tion which meets the similar patches surrounding its neighbors in a succession of straight lines; so that the whole group has quite a tesselated appearance. Here and there where fibers pass over or through a group long tongues of the secretion stretch from the animals to the fibers.” The apperance observed by Hudson is a very common one in groups of dried rotifers; the inter- pretation he makes of it, however, is entirely wrong. What he thought to be the “glutinous secretion” is nothing but the cuticle of the animal from which the internal organs have shrunken more or less in the process of drying and which so closely resembles a secretion of some sort as to be easily mistaken for it on superficial observation. In the present experiments many careful observa- tions were made on rotifers during the processes of drying and subsequent revival with the result that absolutely no evidence could be obtained that any such substance is secreted at the time of dry- ing or dissolved when water is added. By adding a little methyl- ene blue to the water before the rotifers dry, it is easy to follow the process in detail. ‘This stain colors no part of the living animal but that portion of the cuticle which is outermost when it is con- tracted. Ifthe animals are not allowed to remain in it too long there are no injurious effects and normal recovery occurs. Roti- fers thus stained and then placed in clean water before being dried show very clearly all of the changes that occur during the process. If any secretion were poured out in the manner described by Davis and Hudson it should be visible outside of the stained cuticle. 2.30 Merkel Henry ‘facobs Such, however, is not the case. The cuticle itself at the time of drying becomes the “glutinotis patch” and on the addition of water resumes its original appearance. If we assume that the peculiar staining reaction of this portion of the cuticle is itself due to a protective secretion of some sort we still see that the explanation of Davis cannot hold. ‘The staining reaction is the same in all animals, creeping, swimming, and con- tracted, and appears very quickly. Evenin rotifers killed instantly by a drop of boiling water the blue color appears just the same as in living ones. ‘The secretion then, if such it be, is present in all individuals at all times and is not produced only on special occasions as Davis asserts and as it should be if his views were correct. Another objection to Davis’ theory lies in the behavior of the rotifers at the time of drying. Davis held that the failure of roti- fers dried on a clean slide to protect themselves was because of their behavior under such unusual conditions. ‘To quote from his paper, “The rotifers in crawling excitedly over the slide as they generally do trying to find more water or protection in their usual refuge—sand and dirt—part with much of their adhesive covering and the evaporation of the small quantity of water is so rapid that they have no time to settle down quietly as usual while more covering is secreted; they roam about almost to the last minute when they are overtaken by drought and shrink hastily into a ball to dry and perish.” Davis evidently had observed the behavior of rotifers dried on a slide without sand; he just as evi- dently had not observed their behavior when dried with sand. If he had, he would have observed no differences of any importance. Whether a rotifer is dried on a plain slide, under a cover glass, or with sand its behavior is essentially the same. It keeps on creep- ing until the water has so far evaporated that it can creep no far- ther and then it contracts and dries more or less rapidly accord- ing to circumstances. When the water begins to disappear it does not stop when it comes to a quantity of sand and quietly encyst itself as he supposed. In almost every case it keeps on creeping the same as before and often dries up at some distance from the sand. That most of the rotifers are eventually found in Effects of Destccation on the Rotifer 231 contact with the sand is due entirely to the fact that as the water disappears their movements become more and more restricted until they finally come to an end in its vicinity where the water lingers longest. It is true that the presence of sand is beneficial in the drying process but, as experiments to be mentioned later show, this is due entirely to the greater slowness of the drying under these circumstances and not to any preparations the animal makes to resist desiccation. ‘That the presence of sand is responsible for no essential difference in behavior is shown by the fact that roti- fers dried with a small quantity of sand show practically as great a mortality as those dried with none at all. If the mere presence of the sand enabled the rotifers to encyst themselves normally, we should not expect this to be so. If, on the other hand, the effect is merely one of the rapidity of evaporation, the fact finds a ready explanation. It is seen, therefore, that the two points of originality in Davis’ view, namely, the presence of protective secretion and a difference in the behavior of the rotifers when dried under different conditions are both based upon too slight evidence and are not confirmed by more careful observations. So far as his arguments go, there is no reason to believe that rotifers exposed to conditions of dry- ness can escape a true desiccation. On the other hand, in addi- tion to this negative evidence there is much positive evidence in favor of the view that an actual drying does occur. 2 Evidence for the View that True Destccation Occurs In the first place, it must be noted that during the process of desiccation there are very marked changes in the size and form of the rotifers. Shrinkage always occurs and there may be con- siderable distortion of the body if the drying is rapid enough. According to Davis’ view we should expect such shrinkage only in rotifers dried on a clean slide which are killed by the process. ‘The fact is, however, that it occurs in all cases. When rotifers are very slowly dried, with or without sand, they retain their original form very perfectly and fairly accurate estimates may be made of their loss in volume during the drying process. From careful 232 Merkel Henry “facobs measurements made on several different individuals before and after the addition of water it was determined that rotifers which had contracted to one-third or one-fourth of their original volume and remained in this condition for several weeks or a month were perfectly capable of being revived. A number of careful measure- ments showed that under all conditions of drying—on a clean slide, with sand, on filter paper, etc.—a very considerable shrinkage occurs, the exact amount being dependent on the conditions of the desiccation. How can such results, which may be obtained by anyone with very little trouble, be reconciled with Davis’ theory that loss of water is prevented by an impermeable secretion? If the water is retained within the body how can such loss of volume occur? It might be suggested, although such a supposition sounds rather improbable, that loss of water occurs up to a certain, point but that by the time this point is reached formation of some sort of a protective covering has gone far enough to prevent its further escape. That a protective covering impermeable to water is not present is shown by the following experiment. Several rotifers were dried for a month under the most favorable conditions pos- sible; at the end of that time they were examined and seen to be considerably shrunken. A small piece of wet filter paper was then placed in the dish in which they were contained in such a way as not to touch them, and the dish was covered with a piece of glass. Camera drawings were made of the rotifers at intervals of one minute to determine the effect of the moist atmosphere. A distinct swelling was observed in one minute and in five minutes the volume had increased perhaps 50 per cent. ‘The dish was then uncovered and shrinkage immediately occurred. When water was added the rotifers recovered normally. ‘This. experi- ment, therefore, shows conclusively that in rotifers dried in such a way that recovery is possible after a desiccation lasting for a month the covering of the body is freely permeable to moisture. It is certain, therefore, that no such protection as that demanded by Davis’ view is present. It might perhaps be objected that in the above experiment the swelling observed took place mainly in the superficial part of the Effects of Desiccation on the Rotifer 239 body—perhaps even in a special secretion surrounding it. ‘The following experiment shows clearly that this was not the case. A number of rotifers were stained intra vitam with neutral red and then carefully dried. Several were tested to be sure that revival occurred normally on the addition of water. Others were then subjected in the dish in which they were dried, to mozst ammonia fumes, being kept under observation all the while. Almost imme- diately the internal organs, which had stained deeply in the neutral red, began to change color and in less than a minute had become yellow, showing thereby the penetration of the ammonia. This experiment, like the preceding, shows that the cuticle of normally dried rotifers is not impermeable to gases and to water vapor and that it therefore cannot protect the animals from desiccation. It is quite possible and even probable that the cuticle, especially the thicker part about the middle of the animal which covers it when it is contracted, is useful in retarding evaporation. Experiments to be mentioned below show that too rapid evaporation is injurious. But although it may retard evaporation it cannot prevent it and it seems certain, therefore, that so far as any external covering is concerned there is no bar to complete desiccation of the animal. That rotifers which have been exposed to conditions favoring desiccation contain very little water is shown by simple physical methods. Spallanzani and Doyéere both noticed that such rotifers are so brittle that they break into pieces when pressed with the point of a needle. The same result was obtained in the course of these experiments and it was also observed that no water could be obtained by exerting pressure on the cover glass covering the rotifers, provided that the latter were examined immediately after removal from the desiccator. When kept in a damp atmosphere for a short time they absorbed sufficient water to be detected by this method of treatment, and this probably accounts for the results obtained by Davis, who claims to have been able to squeeze water from rotifers which had been dried for three weeks in a vacuum, since he admits that the water was obtained only after repeated pressure had been applied. ‘The fact that repeated pressure was necessary at all shows that the amount of water present could have been but slight, and it leaves room for the objection that there was 234 Merkel Henry ‘facobs time for moisture to be absorbed from the atmosphere. ‘That only a small amount of water is sufficient to give the appearance he describes may be proved by pressing under a cover glass small shreds of filter paper or pieces of apparently dry plant tissues. There is also a possibility that what Davis imagined to be water was only a thin film of air which under certain conditions arising in microscopic work closely resembles water. The fact that desiccation is often very complete may be shown by chemical means as in the following experiment. A number of rotifers, stained zntra vitam with neutral red, were allowed to dry under the most favorable conditions possible and were then placed in a desiccator over night. One lot was moistened and found to be normal in every respect. A second lot was exposed to ammonia fumes rendered as dry as possible by means of calcium chloride and a third lot exposed to the same fumes after having been allowed to remain in a moist atmosphere for five minutes. The third lot changed color from red to yellow almost instantly; the second lot retained their red color. ‘That the effect in the latter case was not due to the inability of the ammonia to penetrate the cuticle was shown by the fact that a number of the rotifers which had been purposely crushed also retained their color. The failure of the characteristic reaction between ammonia and neutral red in the case of rotifers dried in a desiccator must be considered to indicate that they retain very little water. There is also much indirect evidence that the desiccation is a real one. Some of this evidence will be mentioned later but it may not be out of place to refer at this point to the resistance shown by dried rotifers to high temperatures. Gavarret and others have found that rotifers in water are killed by a temperature of 51° C., giving all the evidences of heat coagulation. When partially dried and in a moist atmosphere they are killed at 81° C. and after thorough drying they may resist temperatures of 100°-110° C. Doyere even found a rotifer which after being dried in the sun for several weeks was not killed by a very short exposure to 153° C. How can such results be explained on Davis’ theory that the body fluids are retained within a watertight cyst? Does the protoplasm at the time of the formation of this cyst undergo some mysterious Effects of Desiccation on the Rotifer 235 change which causes it to resist the coagulating effect of high tem- peratures while still retaining its water? Such a supposition is on its very face highly improbable. On the other hand if we believe that there is an actual loss of water these facts fall in line with the observations of Lewith and others that in egg albumin with loss of water the coagulation temperature rises from 56° to 145° C. There is every reason to believe, therefore, that in Philodina we have to do with a true desiccation of the body. ‘This is shown by the chemical and physical tests mentioned above, by the great amount of shrinkage that occurs at the time of drying, by the fact that the cuticle is freely permeable to gases and to water vapor, and by the additional fact that coagulation is not caused by rela- tively high temperatures. Furthermore, it has been shown that the only plausible explanation that has ever been given of a method by which loss of water could be prevented is based on insufficient observation and is not borne out by the facts. We must not as- sume, however, as many have done, that the desiccation is ever an absolute one. Even with the most perfect desiccating devices known it is impossible to remove the last traces of water at ordi- nary temperatures. The chemist recognizes the fact that in organic analysis it is necessary to heat the substance which is being analyzed almost to the charring point to remove all of the water it contains. With rotifers this is impossible since long be- fore this point is reached certain irreversible changes have occurred which cause the death of the animal. It is certain that no rotifer has ever lived after an absolute desiccation. It is useless, there- fore, to speculate whether or not life would be possible after com- plete removal of water, since such a condition cannot be attained by the means in our possession without destroying the very struc- tures on which life depends. But the fact that an animal with no means of protecting itself against loss of water save the hygro- scopicity of its tissues may remain capable of resuming its normal vital activities after a sojourn of weeks or even months in a vacuum or a desiccator is in itself a striking one and one which must neces- sarily enlarge our conceptions of the properties of living matter. < 2.36 Merkel Henry “facobs } VII EFFECT OF DESICCATION ON THE LIFE PROCESSES OF PHILODINA It has been seen that recovery is possible in the case of rotifers which have attained a considerable degree of desiccation. It remains to consider in more detail the effect of this desiccation on the life processes of the animal. Are the latter brought to a complete standstill or are they merely retarded? Is the dried rotifer dead or alive—or neither? Does the process of desicca- tion have any injurious effects and if so what is the nature of these effects? Questions such as these have occupied the attention of physiologists for many-years and in the following section an attempt will be made to at least partially answer them. ‘The last one— having to do with the injurious effects of desiccation—will be con- sidered first since it appears to throw some light on the others. rt Inyurious Effects of Desiccation It has been the testimony of practically all observers that desic- cation is always more or less injurious. ‘This is shown by the fact that when large numbers of individuals are used some deaths always occur, even under the most favorable conditions of drying, and by the further fact that desiccation may not be repeated indef- nitely. Spallanzani was the first to observe the latter point. He found that in certain rotifers experimented on by him all were killed after the sixteenth drying. In the course of this work Spal- lanzani’s experiment was repeated with a small number of rotifers, the latter being dried once a day with a small quantity of sand. The following table shows the number of dead and alive on each successive day: TABLE I Date Alive Dead i Diecoun)\) gH Gnoda ane poduoodbnoGdancdccasdome dom ducdcSeaaSagunoa good II ° MJADLATY | Ps elsteteresoyele rete (olstoysle¥uieiave)s ler -5-9 4 ae. FeBRUARY UANUARY MARCH EGGS EGG-CONTAINING AQOULTS 2.1... LL BrGaer Effects of Desiccation on the Rotifer 255 observation. The three maxima follow each other in regular order, as should be expected, the maximum for the development of the eggs within the body coming first, then that for the deposited eggs, and finally that for the newly hatched young. The two most noticeable irregularities in the curves occur in the January series. The first is in curve I where the maximum comes after the max- imum of curve 2 instead of before it. This is probably due to the fact that no records were kept between January 2 and January 8. It they had been, a maximum in all probability would have been found during this period. ‘The second irregularity, that in the curve representing the number of eggs, on January 25, was due to finding a cluster of five eggs, the eggs usually being laid in groups. A careful search failed to disclose any more eggs and hence this apparently high percentage did not represent a general condition in the culture, being merely one of the errors of chance that con- not be eliminated from a method of this sort. However, taken as a whole, the curves show in a striking manner the relation between desiccation and egg laying. It might perhaps be objected that inasmuch as the periods of desiccation came at approximately equal intervals, their apparent influence on egg production might be due merely to a chance coincidence with a regular periodicity in the production of eggs which is independent of external conditions. That this objection has no force is shown by the fact that in other cultures dried at very irregular intervals exactly the same results were obtained and furthermore that in cultures allowed to continue for a long time without drying no such periodicity -was observed. In one such case the culture had not been permitted to dry for several months. During the latter part of this period the percentage of eggs had fallen almost to zero, it being a rare occurrence to find even a single egg. The culture was then allowed to dry for two weeks. Ten days afterward the percentage of eggs had increased to over fifty where it remained for a few days and then again diminished until a week or two later it was at zero. A number of experiments were made on small numbers of indi- viduals. By keeping them in a fairly small drop of water ina covered dish an accurate count could be made of the number of 256 Merkel Henry “facobs eggs produced. In one such case thirty rotifers which had laid no eggs for some time previously in less than a week produced fifty- five eggs. Similar results were obtained in other experiments. The length of time necessary for the appearance of eggs was about the same as that required in the large cultures. A factor which must be considered as a possible cause of the phenomenon just described is the food supply. Associated with Philodina there are always found large numbers of the unicellular plant, Spherella lacustris. ‘The rotifers feed to some extent on the partly grown cells but much more largely on the small micro- zooids which are produced in large numbers after each period of desiccation. It might be thought, therefore, that the effect observed was due not to the direct influence of the desiccation but to its indirect influence in causing an increased food sup- ply. To decide this point, experiments were tried in which the rotifers after being dried were placed in water free from micro- zooids. In all such cases, although there was no increase in the food supply, egg production was brought about just as before. The following experiment is a typical one selected from a number showing similar results. Six rotifers were taken from a small culture in which no eggs had been laid for several weeks, and dried on a clean piece of filter paper for twelve days. On March 7 they were placed in a dish with water and as soon as they had revived the filter paper was removed. No food of any sort was present. In one of the rotifers the ovary was slightly enlarged, in the others it showed no signs of developing eggs. On March 8 the alimentary tract of all had acquired a glandular appearance but contained no food. Traces of egg formation were observed. On March 9g three con- tained fairly well developed eggs. On March to all six contained eggs and two individuals contained two eggs—a rather unusual occurrence. On the rith one of the rotifers had died but the other five all contained eggs. On the 12th seven eggs had been laid, on the 14th eleven and on the 16th twenty-one, of which four had hatched. By the 17th ten of the twenty-one eggs laid had hatched and by the 21st all had hatched, no new ones being laid after March 16. On the 21st only one of the five rotifers showed any signs of —s. -.* - Effects of Destccation on the Rotifer 257) an egg in the ovary and this one did not develop. No further egg laying occurred. This and similar experiments show that the production of eggs is not due to the better food supply after a period of desiccation; since egg production takes place in the absence of food. Further- more, in the cases where a few individuals were kept under direct observation the formation of eggs was sometimes observed to begin before the microzodids had been produced. That the food supply is not without effect, however, is shown by observations such as the following. Seven rather small rotifers showing no trace of eggs were dried on filter paper for three weeks and then placed in a dish of clean water without any food whatever. The development of eggs started just as before and in three days all except one con- tained partially grown eggs. At this point, however, develop- ment ceased, the eggs becoming no larger. Six days later micro- zooids were added from another culture. ‘The rotifers fed rather freely on them and in three days all contained large eggs. An egg laying period then ensued which did not differ from those already described. ‘This experiment shows that although an abun- dant food supply is not the factor that starts egg production, a suf- ficient supply of food must be present in order that the latter shall continue. ‘This is especially the case in smaller rotifers in which the reserve supply stored in the tissues of the body is not great. Another factor that must be considered 1s the amount of oxygen present in the water. It 1s conceivable that the failure of old cul- tures to produce eggs is due to an exhaustion of the oxygen supply of the water and that the greater activity after a period of drying is due to a renewal of this supply. ‘To this objection it may be answered that cultures in which Spherella is present in abundance and in which, therefore, considerable oxygen must be set free in the presence of the light, behave exactly in the same way as cul- tures in which Spherella is altogether absent. Furthermore, the addition of fresh water when the culture is almost dry or even after a very short period of desiccation produces no noticeable effect. This last mentioned fact speaks against the view that a diminished production of eggs in old cultures occurs as the result of the accumulation in too large quantities of injurious waste 258 Merkel Henry ‘acobs products. No matter how large the volume of water added no increase in egg production occurs without a fairly long desiccation. In cases where rotifers were transferred to other dishes and placed in fresh water there seemed to be no noticeable increase in repro- ductive activity. Although desiccation is able to cause the formation of eggs at times when they would not otherwise appear it is not always neces- sary for their production. In an experiment made to determine the length of life of Philodina, several individuals, raised from the egg without any desiccation, when about a week old began to lay eggs and continued to do so for several weeks. Furthermore, it must be noted that under more natural conditions than can be obtained in the laboratory, eggs seem to be produced more freely than in the cases mentioned. ‘Their production does not seem to be so clearly related to a period of desiccation. This may be partly due to the greater variety in the conditions met with out of doors in a large stone urn and partly to the more abundant food supply which would permit the effects of a single drying to last for a longer time. However this may be, and whether or not desiccation is the only factor concerned in the production of eggs the experiments described leave little room for doubt that it is a factor and under certain conditions a most important one. Un- der these conditions the process of desiccation serves as a stim- ulus to cause the production of eggs at times when otherwise they would not be produced. ‘The drying in some way starts into activ- ity the previously resting germ cells and inaugurates a process of growth and division which may continue for a considerable time after the initial stimulus has ceased to act. ‘This fact is of some interest when considered in connection with certain other well known phenomena relating to the development of both plants and animals. It is well known that most cells cannot continue to grow and divide indefinitely, without receiving certain stimuli from without. The behavior of some of the Protozoa furnishes a striking exam- ple of this point. Several workers have shown that in Paramee- cium after a number of generations division becomes slower and finally ceases unless an appropriate stimulus be supplied. In Effects of Desiccation on the Rotifer 259 nature, this takes the form of a conjugation with another individ- ual, but as Maupas, Calkins, and others have shown, other stimuli may serve the same purpose. After division has ceased it may again be started by a change in food or by various chemicals. Perhaps the list of stimuli that produce this effect is larger than we now suspect. Many examples might be given of the effect of external stimuli in causing growth to proceed beyond its usual limits. The production of plant galls by the stings of insects is one of the best known cases. Possibly the formation of certain pathological growths in animals is due to similar causes. An- other example is furnished by the regenerative processes that occur in both plants and animals after injuries, in which resting cells are stimulated into activity. In the case of the germ cells, which are potentially immortal the effect of outside stimuli is also clearly seen. In the life of nearly every egg cell, after a period of active growth and division comes a stage in which further development is impossible without some outside stimulus. Usually this stimulus is supplied by the entrance of the sperm. ‘The process of fertilization has for its purpose two objects, the one being the introduction of new hered- itary qualities and the other the stimulation of the resting proto- plasm of the egg to develop. ‘That these two processes are dis- tinct is shown by the fact that under certain conditions develop- ment may be made to occur in the absence of the sperm. Cases of artificial parthenogenesis have been reported in many groups of animals in which fertilization is the rule in nature. ‘The echino- derms, molluscs, annelids, and insects all furnish examples, and even in some of the vertebrates the early segmentation stages have been obtained by means of appropriate stimuli. Artificial parthe- nogenesis may be brought about ina variety of ways. Heat, chem- icals, and hypertonic salt solutions have all been used with success. Furthermore, as Loeb has shown, a combination of two of these methods may be more effective than one alone. Experiment will doubtless show new methods and new combinations of old methods to secure the same result. Is it not possible that in the case of Philodina the process of dry- ing may furnish a stimulus comparable to those just mentioned, » 260 Merkel Henry “facobs the desiccation serving to start the development of the resting germ cells? Analogous cases are known in plants. Many seeds are known to be incapable of germination until they have been dried for a certain length of time. Drying is said to start into activity the resting buds of trees. In Spherella lacustris, which is always found associated with Philodina, microzodids are always pro- duced in large numbers after a period of desiccation. Further- more, in Loeb’s experiments on artificial parthenogenesis, the extraction of water from the eggs by hypertonic solutions fur- nished a very effectual means of starting developmental processes. Loeb has shown that the important factor in this case is the actual loss of water rather than its reabsorption when the eggs are again placed in normal sea-water. There seems to be no doubt, there- fore, that under certain conditions loss of water may be an active agent in starting processes of growth and development in resting cells. | It is interesting to note, although the fact may have no signif- cance, that the group of rotifers to which Philodina belongs, which are the only ones of the rotifers normally subjected to frequent periods of desiccation, at the same time are the only ones in which sexual reproduction has been entirely lost. “The thought suggests itself that in the presence of such great changes in the external conditions the stimulus furnished by the union of the sex cells of different individuals is not so necessary as in most other animals. On the other hand, it must be noted that in the aphids parthenogenesis is said to continue as long as the external conditions remain uniform, males being produced only when there is a change in these conditions. Be this as it may, the fact remains that desiccation in certain rotifers is able to bring on a period of ‘reproduction. Whether or not this is the only factor involved it undoubtedly is an important one and one which may be found to have a wider significance than we now suspect. VIII SUMMARY 1 Rotifers of all ages may recover all of their normal activities after an extended period of desiccation. Effects of Desiccation on the Rotifer 261 2 The tissues of the body become truly desiccated, as is shown by the amount of shrinkage that occurs, by physical and chemical tests for water, and by other more indirect methods. ‘The desic- cation, however, must not be considered an absolute one, it being impossible to remove the last traces of water at ordinary tempera- tures. 3. The cuticle of Philodina under all conditions of desiccation is freely permeable to water vapor and gases. No evidence of a waterproof cyst can be found. 4. The presence of sand, etc., is not necessary that recovery shall occur as many have claimed. By proper conditions of dry- ing almost all of the individuals recover after a desiccation of at least four days in the absence of all solid matter whatever. 5 The process of drying is always more or less injurious under any conditions, as 1s shown by the fact that some deaths almost always occur and that it may not be repeated indefinitely. 6 The chances of recovery of a given individual depend on its previous condition as well as on the circumstances under which drying occurs. 7 The rapidity of the drying process has a very important effect on the mortality, rapid drying being more injurious than slow drying. It is apparently heed by what means the process of drying is slowed. 8 The injurious effect of rapid drying seems to be due to mechanical injuries to the vital organs and the cells composing them caused by the rapid extraction of water. g The final intensity of the desiccation makes little differ- ence in the mortality provided it be attained gradually enough. 10 Other things being equal, a fairly high temperature at the time of drying is favorable, a low one unfavorable. 11 Alternations of moisture and dryness are very injurious. 12 Rotifers kept at a high temperature show a greater mor- tality than those kept at a low one. 13 Rotifers kept in a moist atmosphere show a greater mor- tality than those kept in a dry one. 14 Metabolic changes in the tissues probably continue in dried rotifers though very much retarded. The functions of the various organs as such are suspended. % 262 Merkel Henry “facobs 15 he suspension of these functions occurs very suddenly, possibly owing to nervous control. 16 Desiccation usually brings on a period of reproductive activity. BIBLIOGRAPHY BiainviL_eE, M. H. de ’26—Sur Quelques Petits Animaux qui, aprés avoir perdu le movement par la dessiccation, le reprennent comme aupa- ravant quand on vient a les mettre dans |’eau. Ann. des Sci. Nat., ix, 104-110, 1826. Broca, P. ’60o—Rapport sur la Question soumise a la Société de Biologie par Pouchet. MM. Pennetier, Tinel, et Doyére au Sujet de la Reviviscence des Animaux Deséchés. Lu a la Société de Biologie le 17 et le 24 Mars 1860 par M. Paul Broca au Nom d’un Commission composee de MM. Balbiani, Berthelot, Brown-Séquard, Dareste, Guillemin, Ch. Robin et Broca, Rapporteur. Mem. de la Soc. Biol. (3), ii, 1-139. Davis, H. ’73—A New Callidina: with the Result of Experiments on the Desicca- tion of Rotifers. Mon. Micr. Journ., ix, 201-209. DoyerE, M. P. L. N. ’42—Memoire sur les Tardigrades. Ann. des Soc. Nat. (2), XVill, I-32. EHRENBERQ, C. G. ’38—Die Infusionsthierchen. Leipzig, p. 492, 1838. FaccIo!, F. ’91—De la prétendu revivscence des Rotiféres. Archiv. Ital. de Biol., 360-374, 1891. FreDERICQ, L. ’89—La lutte pour |’existence chez les animaux marins Paris, 1889. FROMENTEL, E. de *77—Recherches sur la revivication des rotiféres, les anguil- lules et les tardigrades. C. R. Assoc. Franc. l’avanc. des sci. vi, 641-657. GavarrET, J. ’59—Quelques, Expériences sur les Rotiferes, les Tardigrades, et les Anguillules des Mousses des Toits. Ann. des Sci. Nat. (Zool.), (4), X1, 315-330. Hazen, T. E. ’99—The Life History of Sphzrella lacustris. Mem. of the Torr. Bot. Club, vi, 3, 1899. Hupson, C. T. ’73—Remarks on Mr. Henry Davis’ Paper “On the Desiccation of Rotifers.” Mon. Micr. Journ., ix, 274. The Desiccation of Rotifers. Journ. Roy. Micr. Soc. (2), vi, 79. Hupson, C. T. anp Goss, O. H.’ 89—The Rotifera or Wheel Animalcules. Lon- don, 1889. Effects of Desiccation on the Rotifer 263 Jenninecs, H. S. ’04—Contributions to the Study of the Behavior of Lower organ- isms. Carnegie Inst. Publications, 1904. Kocus, W. ’90—Kann die Kontinuitat der Lebensvorgange zeitweilig vollig unter- brochen werden? Biol. Centralb., x, 673-686, 1890. EcxstEIn, K. (’83)—Die Rotatorien der Umgegend von Giessen. Zeitschr. f. Wiss. Zool., xxxix, 343-344. 1883. LEEUWENHOEK, A. von, 1719 —Continuatio Arcanorum Nature. Epist. 144 ad Henr. Bleysvicium Lugd. Batav., 1719. Letpy, J. ’74—Remarks on the Revivication of Rotifer vulgaris. Proc. Acad. of Nat. Sci., Phila., p. 88, 1874. NeepuaM, T., 1743—A Letter Concerning Chalky Tubulous Concretions, with some Microscopical Observations on the Farina of the Red Lily, and on Worms Discovered in Smutty Corn. Phil. Transact., xlii, 634-641, 1743. Poucuet, F. A. ’59—Expériences sur la resistance vitale des animalcules pseudo- ressuscitants. Comp. Rend., xlix, $86. *59—(2) Nouvelles Expériences sur les animaux pseudo-resuscitants. Comp. Rend., xlix, 492. (3) Recherches et expériences sur les animaux resuscitants. Paris: J. B. Balliere et fils, 92 pp., 1859. PreYER, W. ’91—Ueber die Anabiose. Biol. Centralb., xi, 1-5. 1891. Rirzema Bos, J. ’88—Untersuchungen tiber Tylenchus devastatrix. Kiihn. Biol. Centralb., vii, 650-659. 1888. Sxack, H. J. ’73—The Desiccation of Rotifers. Mon. Micr. Journ., ix, 241, 1873. SPALLANZANI, L., 1787—Oeuvres: Opuscules de physique, animale et végétale, etc. trans. Jean Senebier. Tome, ii, 203-285. Zacuarias, O. ’*86—Konnen die Rotatorien und Tardigraden nach vollstandiger Austrocknung wieder aufleben oder nicht? Biol. Centralb., Vi, 230, 1886. A. hdl | : | Waa hai, a we bs PROTOZOAN STUDIES BY J. F. McCLENDON Witu Two Piates While studying the Protozoa under Dr. H. 8. Jennings in 1904— 1905 I was impressed by the number of the theories relating to the physiology of these minute organisms and began to devise experiments with a view to testing some of the theories. With- out attempting to draw very general conclusions from these experi- ments it is hoped that they will at least suggest further problems and make clearer the fact that Protozoa are very complex organ- isms. Not until quantitative studies of several forms are made will the physiology of the Protozoa be understood as clearly as that of the higher vertebrates. We cannot until then be sure of the significance of reactions such as are described in this paper and for this reason they are not fully discussed. I. REACTIONS OF AMCEBA PROTEUS TO MINUTELY LOCALIZED STIMULI? The reactions of Amceba to various stimuli have been described by various writers, but with the aid of the apparatus shown in Figs. 1 and 2 a more precise localization was possible.? As has been the case with other studies on Protozoa, so here, a detailed study reveals complexities comparable with those found in higher organisms. Mechanical Stimulation Amoebz were stimulated with an extremely fine glass needle. The time was counted with a metronome and distances measured 1 These experiments were made at Randolph-Macon College, Ashland, Va., in 1907. 2 To any one wishing to have apparatus like this made I would be glad to furnish descriptions or other data. Tue JouRNAL or EXPERIMENTAL ZOOLOGY, VOL. VI, NO. 2. 266 F. F. McClendon with an ocular micrometer in one eye piece of a Zeiss binocular microscope. ‘The results of a large number of experiments show that the Amceba does not respond to mechanical stimuli of very small area unless they be repeated at short intervals of time (one to two seconds), and that this interval is in inverse ratio to the area stimulated. Even when a glass needle was thrust through the Ameeba so that the end protruding from the other side was seen, no response was obtained, but the Amoeba moved along as usual, the needle cutting a path through the protoplasm until the Ameeba had passed beyond it. When an Amceba was cut in two gradually by an extremely fine glass needle pressed upon it horizontally, the cutting produced no reaction that could be detected either in the piece with, or the piece without a nucleus. Some time elapsed before the non-nucleated piece behaved dif- ferently from the nucleated. Chemical Stimuli It was found impossible to confine fluids poured out of capil- lary tubes to very small areas, so I resorted to the following in- direct method: A fine copper wire was ground to a needle point and further sharpened by erosin in acid. This copper needle was stuck into the ectosarc of the Amceba. The mechanical effects should be no greater than those of the glass needle (i.e., unnoticeable) but the metallic copper, and colloidal particles flying off from it should chemically affect the adjacent protoplasm. Marked local changes occurred, and if the needle remained in the protoplasm long enough the adjacent area was killed. It appeared to me from a number of observations that this stimulus produced responses very quickly in remote parts of the Amceba. This was very difficult to test for the following reasons: (1) ‘The Amoeba may be considered as constantly receiving stimuli from one or more directions. It is probable that some of these stimuli come from within and are very variable. An additional stimulus must be very strong to produce a reaction that can be distinguished from others. (2) The Amoeba may be considered as a closed bag of ectosarc containing endosarc in the “‘sol” stage, and a con- traction of the ectosare at one place might produce hydrostatically Protozoan Studtes 267 an extension at a distant point. Dellinger (06) supposed that there are strands of denser protoplasm running through the endo- sarc, and gives as evidence the observation that an elongated ingested diatom will move freely along with granules in the endo- sarc when it lies lengthwise to the direction of flow, but will stop when turned sidewise, as though the meshes between the strands were greater than the breadth but less than the length of the diatom. ‘The same effect might be produced, however, by the resistance of the ectosarc when indented by the diatom. The diatom with its silicious shell would probably be heavier than the endosarc and press against the ‘“‘ventral”’ wall of ectosarc. If it were lighter than the endosarc it would press against the “dorsal” body wall of the Amceba. In either case the resistance against being swept along by the current of endosare would be less when it lay lengthwise than when it lay, crosswise to the direction of flow. | To demonstrate this I measured the force required to pull an ordinary glass slide over the surface of a soft gelatine plate under water. It required 2548 dynes when placed crosswise but only 2078 dynes when placed lengthwise. ‘The same would hold for the diatom if it were heavier than or lighter than the endosarc. The current of endosare would act on a larger surface when the diatom were placed crosswise but whether this would be sufficient to overcome the increased resistance it is impossible to determine without knowing the viscosity of endosare and ectosare. But without morphological evidence to the contrary we may safely assume the endosarc to be without a fixed structure. To eliminate the hydrostatic effect I took advantage of the reaction of the Amceba which removes a strongly stimulated area from the source of the stimulus. Ordinarily this is done by local contraction of the ectosarc in the region stimulated. However, if this area is in the middle of a flat side, such a contraction is of no avail and | have observed none. Furthermore, if the opposite point of the Ameeba is in contact with the substratum, its con- traction would not aid in the removal of the stimulated point. By studying Amoebz both from the side after the method of Del- linger and from above, I learned to distinguish from above those 268 F. F. McClendon portions that were attached to the substratum. I found froma large number of observations that an Ameeba stimulated in the center of a large flat area over an attachment to the substratum, by introducing into the ectosarc a copper needle, showed a tem- porary stoppage of the extension of pseudopodia in the most remote parts. ‘The interval of time was in half the cases less than that calculated for the movement of hydrogen ions in aqueous solution (.03 mm. per second). ‘The reaction time in Amceba is considerable (though apparently very variable) and allowing for it, it is probable that in all cases the stimulus traveled at a speed greater than .o3 mm. per second. Lest there be an electrostatic action at a distance I “grounded” the copper needle and repeated the experiment many times but with the same results. In order to facilitate observation I selected large Amcebe moving along without dorsal or lateral pseudopodia. These gave the same results. The Food Taking of the Ameba From the above results, and observations on food taking I pro- pose the following hypothesis to account for the latter process: Chemical and physical influences of the medium cause a har- dening and shrinkage (by loss of water) of the ectosarc (Rhum- blers “‘Geletinisirungsdruck’’). Chemical processes within prevent this hardening from extending to the endosarc, and dissolve portions of the ectosarc that are displaced inward. “The medium affects different portions of the surface to different degrees, causing regional differences in degree of hardening and shrinking, thus producing amoeboid movements. A food body being protoplasmic and therefore similar to the substance of the Amceba might, in lying near an Ameeba, protect it from these outside influences. The protected region would become more fluid, and shrinkage of other regions of the surface would press it out toward the food until it touched it. The food would be pushed along and sometimes rolled over and would rub on the surface of the pseudopod pro- ducing mechanical stimuli of sufficient frequency to cause a local shrinkage of the ectosarc. This stimulus would spread through the protoplasm but being very weak and rapidly growing weaker Protozoan Studies 269 would cause the contraction of only a small area. Beyond the contracted area the protoplasm would continue moving toward the food and surround it from the sides. Probably many other factors enter into and complicate the process and sometimes make it resemble the food taking of higher animals. Ze THE "EFFECTS OF CENTRIFUGAL FORCE ON PARAMCECIUM®? Methods For short periods of time the hematocrit attachment of a hand centrifuge was used. For longer periods I made an electric cen- trifuge. I made several centrifuges that could be run by a one- fortieth horse power or a one-twentieth horse power hot air motor. The University of Missouri furnished me with a Bausch and Lomb electric centrifuge with a special revolving arm of 158 mm. radius, carrying two one-half drachm vials. I enclosed this in a close fitting chamber which increased the speed by preventing radial air currents. With shunt winding 4000 revolutions per minute were obtained. ‘The speed was regluated with a circular rheostat having 32 stops. In most cases I used gum arabic (or other gums) dialysed through filter paper until it was neutral to litmus, to buoy up the Parameecia in the centrifuge, and I repeated these experiments without gum up to as high a speed as the Parameecia could survive. For a convenient index of the centrifugal force the formula n?r was used —where n is the number of revolutions per minute and r the radius in millimeters. In earlier experi- ments the revolutions could not be counted with a speed-counter and had to be calculated from the gear, and the results were prob- ably too high owing tos lipping of bands. The word outward is used to denote direction from the axis of the centrifuge and znward toward the axis. The recorded experiments are on Parameecium caudatum but Parameecium aurelia gave simular results. For permanent preparations | found the best method to be fix- ation for one minute in 1 per cent chromic acid and staining from three to five minutes in Biondi’s methyl green, orange G and acid 3 These experiments were carried on during the session of 1906-1907 at Randolph-Macon College, Ashland, Va., and continued during the winter of 1907-1908 at the University of Missouri. 270 F. F. McClendon fuchsin mixture with a little less fuchsin and of about one-fourth saturated strength. To facilitate changing rapidly from one fluid to another the hematocrit was used to precipitate the Para- meecia. ‘The chromatin was stained green, plasmosomes orange, cell granules red or orange, trichocysts red and cilia and discharged trichocysts sometimes green. Every part of a whole mount could be studied with the 2 mm. Zeiss apochromatic objective so but few sections were cut. Ex periments After centrifuging 15 minutes with n’r = 13,950 X 10° the heavier substances of the food vacuoles and phosphate crystals if present lie in the extreme outer end of the Paramcecium and some may even be forced through the ectoplasm. Next to these come the micronucleus and then the macronucleus. Fig. 3 shows a specimen subjected to this force five minutes. The chromatin has been precipitated so violently as to stretch the nuclear wall, but otherwise the macronucleus has not been displaced. It appear as though the macronucleus were attached in some way, but the appearance might be due to the nuclear sap being less dense than the endoplasm or to the viscosity of the endoplasm preventing the rapid precipitation of the whole macronucleus. ‘The anterior end of the Paramcecium was in this case turned outward. Dr. Lyon (05) showed that this is usually the initial orientation, but the geotropic reaction may be strong enough to turn the anterior end in the opposite direction, as is shown in Fig. 5. Fig, 4 is drawn from a specimen subjected for half an hour to less force (n?r = 6200 X 10°). The micronucleus is almost in the extreme outer end of the animal. ‘The precipitation of the chromatin has greatly stretched the wall of the macronucleus and the wall has burst at its inner end. In a lot that were centrifuged longer one Paramcecium was found to be without macronucleus or micronucleus or even scattered chromatin material. I have this specimen stained and mounted and have examined it repeatedly | without finding a trace of chromatin. Whether the wall of the macronucleus burst as in the preceding case and the nuclei disinte- grated, or whether the nuclei were.forced through the wall of the Protozoan Studies a Paramcecium and lost, or whether the Paramcecium is the result of a division in the centrifuge in which one daughter was pre- vented from receiving nuclear material by the centrifugal force, it is impossible to say. I have found what appeared to be division stages in specimens just taken from the centrifuge. However, if the centrifugal force had been great the process of division in these specimens was usually not continued, but the elongate and sometimes partially constricted animal would swim about for days without undergoing much change before death. Fig. 5 is of a specimen centrifuged 24 hours, n?r = 742 x 10°. ‘The poste- rior end was turned outward, and both nuclei have crowded into that end. All the Paramcecia subjected to this experiment were very small at its close, probably due to loss of water under the increased pressure. Judging from many individuals, the time that the macronucleus remained displaced after removal from the centrifuge was about the same as the time it had probably remained displaced in the centrifuge. Many exceptions to this rule, and the fact that some individuals changed their orientation while in the centrifuge, makes an exact statement impossible. E. P. Lyon (’05) showed that Parameecia centrifuged for some time are not negatively geotropic. Ifthe geotropic reaction be due to the pressure in one direction of substances in the Paramcecium of specific gravity different from the surrounding protoplasm, (statolith theory) an abnormal displacement of such substances might upset the mechanism of geotropism. I found in con- firmation of Lyon’s results that Paramcecia recently taken from the centrifuge are not negatively geotropic but apparently swim as often in one direction as in another and gradually reach the bottom by their own weight. Great care must be used in this experiment to rid the Parameecia of gum if they have been put in it, as Ostwald (’06) has shown that altered viscosity of the medium may change the sign of the reaction. Jensen (’93) has shown that Parameecia are positively geotropic on hot days, and the friction of the air may raise the temperature of fluids inthe centrifuge. ‘The effects of a rise of temperature and increase in viscosity due to a little gum in the medium would tend to neutralize 272 F. F. McClendon one another. It has been shown above that the pressure in the centrifuge reduces the size and probably increases the density of the Parameecia. This might cause them to go to the bottom even though they preserved a negatively geotropic orientation. I found the time elapsing before the return of the negatively geo- tropic reaction to roughly correspond to the time required for the return of the nuclei to their normal position. ‘This might indicate that the nuclei in normal position acted as statoliths. ‘The fact that the Parameecia are constantly revolving on their long axes does not prevent the application of the statolith theory, because Parameecia moving horizontally do not react to gravity (Jennings ’04); it is only when they start to swim downward that they react. I kept Parameecia in the centrifuge for various periods of time up to one week to test whether distance from the nucleus would effect the structure of the ectoplasm, and obtained only negative results. Probably the circulation of the endoplasm is sufficient to equalize the distribution of substances diffusing out of the nucleus. In case gum solution was used control experiments of Parameecia in gum but not centrifuged were made. The gum was eaten and gave the Parameecia a slightly swollen, vacuolated appearance. Some of the experiments are tabulated below: wr TIME REsuLT 164 X 10) 7 hrs. | Many nuclei displaced, they had returned in a few hours. 384 X 10 | 15 min. | Many nuclei displaced, they had returned in 15 minutes. 585 xX 10| 1 day | Many nuclei displaced, they had returned in 1 day. 585 X 10] 1% days | Many nuclei displaced, they had returned in 30 hours. 585 X 10] 34 days | Many nuclei displaced, they had returned in 24 days. 585 X 10| 4 days | Similar results, the displacement was in some cases transmitted to products of fission. 585 X 10] 5 days | Similar results. 585 X 10} 6 days | Similar results. 764 X 10} 6 days | Similar results, many misshapen.* 1318 X 10|} 1 day | Similar results, many misshapen.* 1316 X 10| x day | Similar results, some dumb-bell shaped. In those marked with * no gum was used. A marked effect on the rate of division was noticed in individ- uals taken from the centrifuge. It has been shown by Calkins Protozoan Studies 273 and his students that the rate of division may be increased by various changes in the chemical composition of the medium. ‘To eliminate error from this source I was careful to have the medium of the centrifuged individuals and the control exactly the same. In the table below, the divisions of three individuals centrifuged 24 hours (7?r = 585 X 10°) are compared with the divisions of three control individuals. Both daughters of the first division of No. 1 were kept, and one of them designated as ta. In all other cases one daughter of each division was thrown away. Days No. a Torar 1] 2} 3] 4] 5| 6 7) 8) girojr1/12)13)14\15/16)17|18|19|20)21 2224 | ee Ee ee ee ee ee ee ee ee Centrifuged 1 Bec qorta stata sscnaneepevets cievers vataart¢ di isladddaddddddddadadddo 4 Dt aaccsiteeleveremniers eidistnie a: one eas 9 a] oc dado god dd agdooo De rece Mayet wea a atstats | 9 0 I] ce] 1 0 OF OOO 1 A AAO DFO dad oo 3 A\SoRron coca Rcd bee Cotto mat Oo} 1] 1] 1] Of | OF OO I] Of C1 OO | IO} 1 Ol OO OTF 8 Control et | | | Defi ue tehee Cons tas eoeraisecme ci o| of 1] ¢ oo 0 OO Jado 00 0 0qdq00 3 DRS RN ast Petar eM CL eRe xe o| o| of 1 o| o} Oo} 1| of & Oo! Of Of} Oldejad 2 BA ecm eta borate erin note eay ees loo 1 o| | o of o of of Odelad)...|..)..]..)..)..].-.[..]..] I | | In the three series from the centrifuge (not counting Ia) there were 15 divisions in 23 days whereas in the check there were only 6 divisions in that time and two of the series had run out. In the table below three individuals centrifuged 32 hours at the same rate are compared with three in the check: Days No. | Tora 1] 2} 3/ 4| 5] 6| 7| 8] gi1ojr1 rele learete later elie ae aa ee Nm Centrifuged | | | Tisfete nistege tere cistosskctepora.ascvelere mbaaxsys 1| 2} 3] ©} 1| oldelad Waele} | leiats 7 DeVoe co chee sit ahora a\eyslVoepace a vsiOs a 1) a] of of 1/ of of | of a] 2) Oo 1] 1) oc fe) Biase lelalon=hciel sh cieoi=ackefey sie sfelesere}eletstons 2) Ij 1) I] Ij oO} c] I] co] c} I I] oo fod 3a) ala 13 Control | | lepoonsanhoodpeeobauuse bondodc 1] 1] 2] 1] ojdejad | | 5 hs ad cae ORR O EMA CERCA E ae 1} 1| 1] o| o| oldelad te | 3 g\ncod er lotan aa Quomdesedbc 9 1° 1\dejad | 2 Neh | Ire! ac 2.74 F. F. McClendon In the above experiment there were in 20 days, 30 divisions in the three series from the centrifuge and only Io in the control. Dur- ing the five days in which the control lived, there were 16 divisions in the centrifuged series. Many experiments with Parameecia centrifuged from one hour to six days, both with and without gum, gave similar results, only the effect was not so marked in the longer periods. All the above lots were from the same culture. It was stated above that many Parameecia are misshapen in the centrifuge. The end into which the nuclei are precipitated is bulged out by them, and in a few cases the other end 1s also bulged out (by the accumulation of substances of low specific gravity ?) giving the animal the form of a dumb-bell. I fed Parameecia on egg yolk, and globules of a dark brown fatty substance were formed in the endoplasm in such numbers as to make it appear black. These black Paramcecia were subjected to as great a centrifugal force as they could stand, in some cases for two or three days. The fatty substance being of low specific gravity, accumulated in the end opposite the one into which the nuclei were precipitated. The result was a pear shaped body, the large end being black and the small end of the normal Parameecium color. So great was the difference in the specific gravity of the two ends that the animal could only with difficulty assume a horizontal position or turn the small end uppermost. No marked increase in rate of growth or reproduction that could be ascribed to the stimulating effect of the lecithin in the yolk was observed. It was noted above that some of the Parameecia when taken from the centrifuge appeared to be undergoing division. More often the body was abnormally elongated, with or without a con- striction in the middle. In these cases the end not containing the “nuclei shriveled after a day or so, and the animal died within a week. It may be that the centrifugal force keeping the nucleus in one end prevents its division though the animal is large enough to form two, and the portion around the nucleus attempts to form itself into an individual, thus causing the constriction in the middle. If such a division is actually completed I have never observed it. When Parameecia are centrifuged without gum in the medium, Protozoan Studies 275 the bacteria upon which they feed are all precipitated to the outer end of the receptacle. Although the centrifugal force may be such that the Parameecia can swim against it, the lack of bacteria elsewhere may cause them to remain in the outer end of the recep- tacle in the area made acid by the bacteria. Here some of them are pressed out of shape and may be reduced to thin lamelle. I have seen such forms after being taken from the centrifuge swim about for days in this flattened condition, and sometimes finally regain their normal proportion. More often the Paramoecium turns a number of times while being pressed in the outer end of the receptacle and is reduced to an irregular mass. Such masses may grow to large size and develop several buccal grooves. They may also divide and some of the products of division be irregular, while others form normal Parameecia. A curious case is shown in Fig. 6, though as this specimen was fed on egg its abnormality is probably as much due to the bulging out of one end by the fatty substance as to the pressure of the wall of the receptacle. Its condition when removed from the centrifuge is shown at a. The large end is three-lobed and it appears as though it were about to divide itself longitudinally into three individuals. The next day there are four lobes on the large end and two on the small end (0). On the third day two of the constrictions have disappeared and the animal is almost divided longitudinally into two. In this con- dition it died. One of the irregular individuals taken from the centrifuge and isolated in small watch crystals, divided into’two daughters that appeared normal in every respect save that each bore a long horn on the oral side. Dr. H. S. Jennings found a Paramcecium sim- ilar to one of these, in an old culture, and found that for a number of generations (1.e., as long as the series lived) the horn was passed to one of the products of each division and the other was normal. I thought it would be interesting to compare the transmission of these horns produced mechanically, with his observations. ‘The diagram in Fig. 7 presents the results: ‘The irregular individual taken from the centrifuge is represented by a; it divided to form the first two daughters, each with a horn. Some daughters are represented by small sketches, the others by the letters 7 meaning 2.76 F. F. McClendon normal and h, bearing a horn. One of the first two daughters has the horn nearer the anterior end the other nearer the posterior end. After each division the horn is in a different position, and we can predict the position of the horn in each generation by drawing an imaginary line bisecting the animal in the preceding generation transversely. Many of the “normal” daughters were kept many generations without showing any abnormality in their offspring. Although one series died out in the sixth and the other in the eighth generation there is no reason to believe that the horns would have been lost had the series lived. Sometimes the horn grew and at other times decreased in size but in the later genera- tions it was as large as in the earlier. It is easy to see why such deformities are seldom found in nature, for in case one is pro- duced, even if the deformed individual has an equal chance with normal ones of living and reproducing, after seven generations less than one per cent of its offspring will show the abnormality. The main difference between the results of the reproductive process here and in the Metazoa is the transmission of acquired characters to a small per cent of the progeny in case such charac- ters do not cause the death of their possessors. “To speak of a germ plasm in Paramcecium without morphological evidence might seem unwarranted. 3. ON ABNORMALITIES PRODUCED BY ENCYSTMENT AND OTHER CAUSES IN PARAMCECIUM AURELIA.* « The encystment of Paramececium putridum was described by Lindner (’99) that of P. busaria by Prowazek (’99) and that of P. caudatum by Simpson (or). I have repeatedly observed Para- moecium aurelia forming a thin membranous cyst in which it might be confined for a week but in which it was killed by drying. For this reason it might be wrong to compare these cysts with those observed by others in Paramcecium. ‘These cysts are most often seen in the interior or on the surface of bacterial zodglea and I thought at first that the Paramececia were simply entangled in the zodglea, but as other Paramcecia were seen at the same * These observations were made at the University of Missouri. Protozoan Studtes 277, time making their way with ease through the same zoéglea at least part of the wall of the cyst must be secreted by the Para- meecium. During the formation of the cyst the animal contin- ually rotates inside of it and the ciliary coat is never lost. The cyst gradually contracts until it is shorter than the occupant, which may have the anterior end folded over the middle of the body or the ectoplasm thrown into folds. After the animal has been in the cyst for some time the folding of the ectoplasm may assume the character of an invagination of the anterior (rarely posterior) end. ‘The invaginated ectoplasm seems to be in large measure absorbed, for after several days the occupant of such a cyst looks like merely the posterior end of a Paramcecium. If the cyst is opened or if one waits until the occupant comes out of its own accord, the latter will swim about and ingest its food as though an anterior end were superfluous. I found numbers of Parameecia without anterior ends, and a few without posterior ends, in the old culture in which the encystment was found. One of these is represented in Fig. 8, a. It was isolated in a watch crystal and remained in the form of the figure two days. On the third day it had changed to the form shown in Fig. 8, 6. This might be interpreted as a division in which the reduced vitality of the animal prevented the complete separation of the daughters, followed by a partial division of one of the daughters. The speci- men was lost so that its later history could not be followed. Some abnormalities seem to be the result of the plasticity of the adoral side at the time of conjugation. A pair of conjugants were isolated, and when they separated the adoral regions were drawn out into prominences (Fig. g). Such prominences are gradually absorbed although they may remain for several days. ‘These prominences do not seem to be of the same nature as the horns shown in Fig. 7 since so far as the investigations go the horns do not and these smaller prominences do disappear. If Parameecia are shaken up with broken glass all those that are not killed or cut In two regain their normal form in a short time (and even fragments if they live regenerate after a longer period) although they may be so torn and mashed out of shape as to bear little resemblance to the type. Probably the form of a Paramcecium 278 F. F. McClendon cannot be permanently changed without the formation in it of one or more new chemical substances. In other words we might hypothetically consider the horn alluded to as of the nature of a graft differing chemically from the Paramcecium to which it was attached. 4. ON VARIATIONS IN PARAMCECIUM CAUDATUM AND P. AURELIA® In making the foregoing studies I was interested in inquiring into the identity of the species used. In the experiments at Ashland, Va., only one form was used. It had but one micro- nucleus and agreed in other respects with descriptions of Para- moecium caudatum. At the University of Missouri two forms were found, one similar to that studied in Virginia and another having two micronuclei, and in this respect agreeing with descrip- tions of Paramcecium aurelia. Calkins (’06) found a case of P. caudatum acquiring two micronuclei and some of its offspring losing one micronucleus and becoming normal P. caudatum again. He further states that the number of micronuclei is the only invariable character for separating caudatum from aurelia and therefore they are probably the same species. Whether caudatum and aurelia form two species or not cannot be decided from the data at hand but I have evidence to show that these two forms are quite distinct. In none of the numerous cultures which I have kept for months and in one case a year, have | found indi- viduals with different numbers of micronuclei in the same culture. No characters of outward form were found that would serve to separate caudatum and aurelia. ‘The size character is best studied in graphs of the lengths of a large number of individuals of each culture (Fig. 10). In each curve in Fig. 10 the lengths of the indi- viduals measured in fractions of a millimeter are plotted as abscis- sas and the number of individuals in a class represented by arbi- trary units on the ordinates and marked at the top by a cross. The curve itself is merely to aid the eye in comparing measure- ments from one culture with those from another, as are also the 5 These observations were made at the University of Missouri. Protozoan Studies 279 continuations of the .1, .2, and .3 mm. ordinates. ‘The height of one curve is not to be compared with that of another as it depends on the number of individuals measured and the number of classes, but the spread and limits of the curves are to be compared. ‘The specimens were measured after killing in one per cent chromic acid, a mechanical stage was used to prevent any unconscious selection and the measurements for some curves were made with an ocular micrometer, and for other curves with a camera lucida. Fig. 10, a represents the lengths of 218 individuals from a culture of P. caudatum from Hinkson Creek, Columbia, Mo., kept in the laboratory one month. Individuals .2 mm. in length form the class of greatest frequency. The next curve (b) is of 234 indi- viduals from a pond near by. The class of greatest frequency is .1g mm. ‘The third curve (c) is of 219 individuals from Ash- land, Va. They had enough food for health but not enough for reproduction. ‘The class of greatest frequency is .182 mm. After feeding this culture 24 hours on hay infusion, 184 individuals were measured (d), and the class of greatest frequency calculated at .Ig1 mm., showing that the majority had increased in length. It will be noticed that the curve extends farther to the left. “This is due to the fact that many individuals had recently divided and were shorter than before feeding. ‘These and all other cultures of P. caudatum gave a nearly symmetrical curve. Such was not the case with P. aurelia (shown in the last two curves). ‘The curve e is plotted from the lengths of 297 individuals from an old cul- ture found at the University of Missouri. The class of greatest frequency is .133 mm. Hence most of them are shorter than the majority of P. caudatum, but it will be seen that the second mode of the curve is composed of individuals longer than the majority of P. caudatum. At first I thought some individuals of caudatum had gotten into the culture, but an examination of the micronuclei of a large number of individuals both large and small proved that not to be the case. “The next curve (d) is of 127 individuals from a culture from a drain ditch at Columbia, Mo. It shows two modes similar in position to those in the preceding, but the highest is of the largest individuals (.26 mm. in length). From these two cultures of P. aurelia I isolated individuals of different lengths 280 F. F. McClendon and started separate cultures from them, which were kept for months. In every case the progeny of one individual showeda curve of as little spread as those given of P. caudatum. Thus it was possible to obtain a culture of minute individuals or one of giants or one of medium size. Subjecting a culture to higher tem- perature or increased salinity decreased the size of the individuals while lower temperatures increased the size of the individuals. P. aurelia of this region must then be dimorphic or else it hap- pened that in starting both the above cultures (e and /) only large and small and no medium sized individuals were procured. Large samples of water containing decaying leaves, etc., from various places developed no cultures of Parameecia, so that more data for aurelia was not obtained. We may say then that aurelia differs from caudatum in the presence of two micronuclei and that some aurelia are smaller than the smallest caudatum. It has frequently been noted that conjugating individuals are smaller than non-conjugants. I think that the fact that the mouths of conjugants are closed is sufficient to cause the smaller size. By comparing curve c (of individuals fed less) with curve d (of individuals fed more) you may note that those fed less are smaller. However, no matter how little food is given Parameecia they can still take in water through the mouth and be swelled up with vacuoles, which is not the case with conjugants. Note that the variation of the well fed individuals is greater than that of the poorly fed. Dr. Pearl (07) found the same to be true of non- conjugants as contrasted with conjugants. In conclusion, a note on the effect of a certain food may be of interest. It is well known that hay infusion has to be often re- newed to keep a culture of Parameecia in good condition. If the culture is left in the same infusion the individuals become smaller, sluggish and finally die. However, if a considerable amount of cane sugar be added to the culture, the Parameecia are very little affected at first but after several weeks become more active and live for months. Whether this be due to alcohol that appears or to bacteria produced in the culture was not determined. Protozoan Studies 281 APPENDIX Dr. Jenning’s paper entitled “Heredity, Variation and Evolu- tion in Protozoa. I” (this journal, vol. v) appeared after this paper had been sent to press. I would emphasize a probable difference between the process of heredity in Protozoa and in Metazoa other than the difference in complexity. In Protozoa the “‘germ-plasm,”’ whether it be all or part of the individual, is probably equally as accessible to the environment as the “soma.” I use the words “germ plasm” and “soma” for brevity. BIBLIOGRAPHY Of Part 1 BERNSTEIN ’00—Chemotropische Bewegung &c. Arch. f.d. ges. Physiol. 80 Bd. De uinceER, O. P. ’06—Locomotion of Amceba and Allied Forms. Jour. Exper. ZoOl., vol. iii. Jennincs, H. S. ’04—Contributions to the Study of Lower Organsims. 6. The Movements and Reactions of Amoeba. Publication No. 16. Car- negie Inst. McCtenpvon, J. F. ’07—Experiments on the Eggs of Chztopterus, etc. Bio. Bull., vol. xii. MatHews, A. P. ’07—An Apparent Pharmacological Action-at-a-distance by Metals and Metaloids. Amer. Jour. Physiol., vol. xviii. MaxweELt, S. S. ’07—Is the Conduction of the Nerve Impulse a Chemical or a Physical Process? Jour. of Bio-chemistry, vol. iii. Montcomery, E. ’81—Zur Lehre von Muskelcontraction. Arch. f. d. ges. Physiol., 25 Bd. Ruumser, L. ’98—Physikalische Analyse von Lebenserscheinungen der Zelle. I. Arch. f. Entwicklungsmech. d. Organismen, vol. vii. ’o5—Zur Theorie der Oberflachenkrafte der Amoben. Zeit. wiss. Zool., 83 Bd. SUTHERLAND, W. ’06—A Molecular Theory of the Electric Properties of Nerve. Amer. Jour. Physiol., vol. xvii. Of Part 2 Carkins, G. N. anp Lies, C. ’02—Studies on the Life History of Protozoa, II. Arch. f. Protistenk., 1 Bd. 282 F. F. McClendon Carxins, G. N. ’04—Studies on the Life History of Protozoa. IV. Jour. Exper. Zool., vol. i. CzarEk ’02—Stoftwechselprozesse in der geotropisch gereizten Wurzelspitze, etc. Ber. deutsch. bot. Gesell., 20 Bd. Darwin, F. ’03—On Artificial Production of Rhythm in Plants. Ann. of Bot., 1903. Hatal ’03—Lecithin a Stimulant to Growth. Amer. Jour. Physiol., vol. x. Jensen, P. ’93—Ueber den Geotropismus niederer Organismen. Arch. f. d. ges. Physiol., 53 Bd. Jennincs, ’04—The Behavior of Paramcecium. Jour. Comp. Neur. and Psy., vol. xiv. KuawkInE, M. W. ’88—Le Principe de |’Hérédité et les Lois de la Mécanique en Application a la Morphologie de Cellules Solitaires. Arch. Zoo. Expér. et Gén. (2) Tome 6. Lyon, E. P. ’05—On the Theory of Geotropism in Parameecium. Am. Jour. Physiol., vol. xiv. ’o7—Results of Centrifugalizing Eggs, I and II. Arch. f. Entwicklungs- mech. d. Org. 23 Bd. Maupas, E., ’88—Recherches Expérimentales sur la Multiplication des Infusiors Ciliés. Arch. Zoo. Exp. et Gen. (2) Tome. 6. Mirtropuanow, P., ’°03—Nouvelles Recherches sur |’Appareil Nucleaire des Para- mécies. Arch. Zoo. Exp. et Gen. (4), Tome i. Morean, T. H., ’06—Influence of a Strong Centrifugal Force on the Frog’s Egg. Arch. f. Entwicklungsmech. d. Org., 22 Bd. Mortier, D. M. ’oo—The Effect of Centrifugal Force on the Cell. Ann. Bot., vol. xiii. OstwaLp, WoLFcGANG, ’03—Zur Theorie der Richtungsbewegungen schwimmen- der niederer Organismen I. Archiv. f. d. ges. Physiologie, 95 Bd. *o6—II. Archiv. f. d. ges. Physiologie, 111 Bd. Wipers, E. ’oo—Inutilité de la Lécithine comme Excitant de la Croissance, etc. La Cellule, T. xvii. Of Part 3 LinpnEr, G. ’99—Die Protozoenkeime in Regenwasser. Biol. Cent’bl., xix Bd. ProwazEK, S.’99—Kleine Protozoenbeobachtungen. Zool. Anz., xxii. Simpson, J. Y. ’°o1—Studies in Protozoa. I. Pro. Scott. Micro. Soc. iti. Protozoan Studies 283 Of Part 4 Grrassimow ’04—Physiology of the Cell. Hertwic, R. ’03—Wechselverhiltniss von Kern und Protoplasma. Miinchen. (Lehmann). PEarL, R. ’o7—Biometrical Study of Conjugation in Paramecium. Biome- trika., vol. v. Pirate I Fig.1 “Mechanical hand” for moving a fine pointed instrument in stimulating Amoeba. When in use it is clamped to the stage of a Zeiss binocular microscope (Greenough model) so that the whole appara- tus except the point of the instrument lies to one side of the field of the microscope. When necessary two can be clamped to the same microscope. A crude form of this apparatus, figured and described in an earlier paper (McClendon, ’o7), had the disadvantage of surrounding the field of the microscope on three sides and sometimes colliding with the watchglass or slide on which the objects were placed. By turning the milled head on the left an instrument held in the clamp is moved crosswise, by turning the milled head in the center the instrument is moved back and forth, and by turning the one on the right it is moved vertically, all three are held in the hand and turned by different fingers. Fig. 2 A simpler form of the same apparatus. The milled head moving the instrument vertically is high above the other two. PROTOZOAN STUDIES PLATE I J. F. McCrenpon Y- Lo KASS } pies = Roan = Ra . Ve. ONIOCZOZTZIZCIIZRA "Se SLEAZE bay IAN SAN TI.—OEE WO x sy = is Sey EZ Tue JouRNAL or ExPERIMENTAL ZO5LOGY, VOL. VI, NO. 2. Pirate II Fig. 3 Parameecium caudatum, (optical section) centrifuged 5 minutes; n2r = 13,950 X 10% Fig. 4 Ibid., centrifuged 4 hour; n?r = 6,200 X 10%. Fig. 5 Ibid., same scale, centrifuged 24 hours; n?r = 742 X 10% Fig. 6 a. P. caudatum fed on hens’ egg yolk and centrifuged two days, n?r = 764 X 108. 5, the same one day later. c, the same one day later. Fig. 7 Genealogical table of descendants of a, a P. caudatum mutilated in the centrifuge (6 days, n’r = 764 X 108). hk = horned and n = normal individual. Fig.8 a= P.caudatum from old culture. 6 = the same two days later. Fig. 9 A pair of ex-conjugants showing deformities produced by conjugation. Fig. 10 Series of curves showing variation in length in P. caudatum and P. aurelia. a, P. cauda- tum (218 individuals) from Hinkson Creek, Columbia, Mo. 6, P. caudatum (234 individuals) from a pond at Columbia, Mo. c, P. caudatum (219 individuals) from Ashland, Va., on maintenance. d, P. caudatum (184 individuals) from Ashland, Va., well fed 24 hours. e, P.aurelia (297 individuals) from old culture at the University of Missouri. f, P. aurelia (127 individuals) from a drain ditch at Columbia, Mo. PROTOZOAN STUDIES PLATE II J. F. McCienpon Tue JourNaL or ExPERIMENTAL ZOOLOGY, VOL. VI, NO. 2. THE DEVELOPMENT OF ARTIFICIALLY _PRODUCED CYCLOPEAN FISH—‘THE MAGNESIUM EMBRYO” BY CHARLES R. STOCKARD Witu One Pirate anp Sixty-THREE Text Ficures Hn ETO LUCE Oralmere tat te beey Toe gna Pete ls 2,3) oy esters ies sve Sievers srtiarse.diacvalevesci ad orb eso ocre ole, ole aeons 285 Iiraterralleenrecd meat tho emmete iesvale aici tie eiclele eiereielevaie nits ae oF ala satan taecd ile aw ale doine ant wleree Someones 287 Cyclocephali and Monstra Monophthalmica Asymmetrica.. ............0eeceeeceeencececees 292 a Living cyclopean monsters from the first indication of the defect to the time of hatching. .... 294 DapErcess wimpy cyclo peanenishy-ye aaycyers r= «,s-eteseiateciciaroreko o cles &ntatela)-teleinvs sihs eve a ence eRe 298 CaplavinesVWonstta: Monophthalmica Asymmetrical) 4..1/10-)el eee oes ence 304 MMorpnoloeyom Gy cloceplalialssrmjccesioreleteace- soe sieiotea Gis trareisies rerwaie © vieleious te nem amieea coe eee 305 a Earliest indication, exact position and condition of the eye.............0.cesceeceuseees 305 bapelncompletelcy clo pias doublereyesics\octvs.s s-simars fia ere x dans ae ogee Aazevod ocd aves aise savant ae 310 eeebchecusinpletcyclopean eyernd money brains) occ alae serie eee aie ale cetera eels 311 d Extreme cyclopia; from the abnormally small anterior cyclopean eye to entire absence of CHES rojas ayia seis ep etehern cs ciete e ast teae @ Shanty acsearle g's spt alavypatletmatol une abetwtnte yw mpere ravens Camas 315 Incomplete diprosopus, with three eyes and one additional lens.............0.eeeeeeee eee eeeees 322 Morphology of Monstra Monophthalmica Asymmetrica .......... 0. cee e cece eee eee e ences 325 Independent origin and self-differentiation of the crystalline lens..............0-0-eee eee cence 328 MD ScrIsslO Usa CKCOUCIUISLONG Hare trctele tel TOE dot Bivln. cs deo tec sore hbet oslavatshers Sites ele cladene ais Mee orotate 330 SERIE IN chs aidte doin Eb COED RO CEO IGE RODEO s Hae GAO ne Or ane d pe mbm TA Sob Ot Cunt ocoNaco 334 BILAL NENGILC CE arerk tant eters ayes yvrss oo. sre Xesee os olara shsfete: tyarovaishelcté ie Gob Wal erarafessNoprs ok sie. oor yoverne ener stele 336 INTRODUCTION Development is the resultant of the interaction between the inherent tendencies contained within the egg substance itself and the external conditions which surround and act upon this sub- stance. The usual interaction of these factors gives rise to normal animal forms. When, however, either factor is changed an unusual form results; in the one case there arises a germinal variant, and in the other an anomaly occurs as a response to the strange external environment. ‘The product of development, the formed animal, is then to a certain extent a creature of its environ- ment. On the other hand the importance of the internal factors must be recognized although modern experimental work often- THE JouRNAL or EXPERIMENTAL ZOOLOGY, VOL. VI, NO. 2. 286 Charles R. Stockard times points in a direction which would indicate that these factors may be largely modified in their influences by the external con- ditions. Most monstrosities or abnormalities in development are due to the action of external factors, either mechanical, as pressure, or chemical. Mammals, birds and reptiles, with their complex embryonic membranes, offer many opportunities for the produc- tion of secondary abnormalities arising from unfavorable mechan- ical or physical conditions. Fcetal amputations and scars, mem- brane fusions distorting facial development, and many other such deformaties are in most cases probably due to secondary influences on development. Besides these there are deformities of a dif- ferent nature, such as the excessive monsters, monstra in excessu, in which certain organs have over developed or produced super- numerary parts; and contrasted with them are defective monsters which fail to complete themselves and are therefore less than a normal individual. It is with this latter class, monstra in defectu, that the present study is concerned. ‘These defective individuals may be grouped into two sub-classes: first those in which certain organs fail to complete themselves, as in cleft palate, hare-lip, arrests in the development of the heart and other parts of the circulatory system. Second, individuals in which certain paired organs occur singly or without mates. “True Cyclocephali or cyclops monsters find their place in this last group. Cyclops monsters have long been known to occur in manand other mammals and are described in many of the earliest medical works. In these beings the one eye is in the middle line of the face and often shows external evidence of a double composition. The nose which normally arises above the eyes and grows down between them as the face develops is here mechanically prevented from descending by the presence of the median eye in its path. The foetus, therefore, has a proboscis-like nose above the eye. The brain and other parts of the body are sometimes deformed though they may be normal. Among the lower vertebrates true cyclops monsters have been recorded by Spemann (’04) as resulting from mechanical injuries to the eggs of the amphibian, Triton tzeniatus. These mon- Artificially Produced Cyclopean Fish 287 sters were double-headed with one or both heads showing the cyclopean defect and were not of the usual single cyclopean type found in man and other mammals. Two years ago (1907) I carried out experiments in which I was able to produce typical single cyclopean fish. ‘This was the first record of the occurrence of cyclopia among fishes. It is also the first case of consistently producing vertebrate monsters such as are known in nature by changing the chemical environment in which the eggs develop. These embryos are in main details similar to the mammalian cyclops, having a single median eye and anteriorly placed double nasal pits. The monsters were produced by allowing the eggs to develop in sea-water in which there was an excess of MgCl,. Cyclopia occurred in a large percentage (at times 50 per cent) of the embryos. The discovery was made so late in the spawning season that it was impossible to investigate the details of the cyclopean defect or rear the embryos to hatching in order to observe their ability to swim ortosee. ‘The method of produc- tion, however, offered such an exceptional opportunity to obtain abundant material for studying all stages of development and degrees of cyclopia that this more extended survey was under- taken. The following account includes a comparative study of cyclopean embryos from the earliest appearance of the optic vesicle to the perfectly formed free-swimming fish with a functional cyclopean eye. The experiments were conducted in the Marine Biological Laboratory at Woods Holl, Mass., during the past summer, while occupying one of the rooms of the Wistar Institute. MATERIAL AND METHOD. As in my previous experiments, the eggs used were those of the teleost fish, Fundulus heteroclitus. The method of producing the defect was much the same as that previously employed although expanded and modified in many ways. During the early part of the season it was difficult to find 288 Charles R. Stockard solutions of the proper strengths and the eggs were either killed or unaffected. After a few experiments, however, a strength of MgCl, in sea-water was found that gave a large percentage of cyclopia, in many cases again causing 50 per cent of the eggs to form such individuals. ‘This was a 42 m solution prepared as follows: 19 cc. of a molecular solution of MgCl, in distilled water was added to 41 cc. of sea-water. This is not then an actual 19 Mm MgCl, solution but it is 18 parts molecular MgCl,. Mak- ing the solution in this way adds to the sea-water, water lacking all of its constituents except the Mg and thus increases in a greater proportion the excess of Mg present. Cyclopia occurred in a series of similarly prepared solutions ranging as follows: 16 M, 17 M, 18 M, 19 M, 29 M, 21 M and 22 M MegCl,. A point of importance is that the proportion of cyclops embryos produced gradually rises in this series up to the 12 M solution and then falls off again. ‘To illustrate concretely, in Experiment VII the 4% M solution caused 12 per cent of the eggs to form cyclopean embryos, the 17 M gave 30 per cent, the 18 M 22 per cent, while 12 M gave 50 per cent with the cyclopean defect. Continuing the series, the 2° Mm falls off to 30 per cent and the 21 M gives 23 per cent, while in the 22 M no cyclopia occurred and the eggs were all killed. It must be born in mind that these percentages are for the eggs that formed embryos and not for the total number of eggs first put into the solution. The peculiar fact is, that in a series of MgCl, solutions we reach a place where a maximum number of cyclopean embryos occur and in strengths both weaker and stronger than this the number of cyclopean individuals is less. If the defect is due to osmotic pressure, we should not expect a greater pressure to bring about a more normal development. If the action is chemical, we do not usually reach a chemically effective dose and find that a greater dose 1s less effective. It might be argued that below the point of maximum occurrence of the cyclopean defect, the solutions are insufficient to effect any but the weaker embryos, so that a small number of cyclops appear; above this point the solutions are so strong that all except the hardiest embryos die in early stages and those sur- viving are so resistant that only a few give the cyclopean defect. Artificially Produced Cyclopean Fish 289 At the maximum point the normal or ordinary individuals, which predominate, would be affected, and here the greatest number of cyclopean embryos occur. As I previously mentioned, the MgCl, is found to be rather toxic to these eggs during the earlier stages of development. Many die at this time, but in the medium strength solutions 70 to 80 per cent live and form embryos and in the weaker solutions often more than go per cent live. After the early embryo is formed, however, the high death rate falls and a dead embryo is of rare occurrence in any of the solutions. Many embryos were kept in the solutions thirty days and some hatched in strengths as strong as 18 M. If, on the other hand, the eggs are removed from the solutions when sixty or seventy hours old, when the cyclopean condition is readily distinguishable, and placed in sea-water they grow much better and many hatch normally. Some of the cyclopean fish came out on the twelfth day after fertilization, though usually they were much slower in emerging. ‘The control embryos hatch in from eleven to twenty days, depending chiefly upon the temperature. Solutions of MgCl, in Distilled Water Distilled water solutions of MgCl, of several strengths; 1% m, 11 mM, 12 M, 13 M, 14 M and 15 M were not effective. The eggs either died during early stages or developed into embryos with two normal eyes. I had found (1906) that salts of lithium induce the same typical defects in Fundulus eggs in both sea-water and distilled water solutions. Such solutions have opposite conditions of pressure, being in one case hypertonic and in the other hypo- tonic and thus remove all question of osmotic effects as a cause. It was hoped that Mg might also act in the two solutions which would have made it certain that the direct action of the magnesium ion is responsible for the cyclopean condition of these embryos. The problem of cyclopean formation seems, however, to be more complex. It involves the action of magnesium in the presence of certain or all of the sea-water salts. 290 Charles R. Stockard Solutions of MgSO, and Mg(NO,), in Sea-water Sea-water solutions of MgSO, prepared in a similar manner to the MgCl, solutions above were employed. ‘The following Strengths $M, §} M, $3 M$} M,4o M, io M, 33 M, 33 M, $$ M, 24 M, 25 M, and 27 M were ineffective, the eggs in these solutions developing normally with very few deaths at any stage. Meg(NO,), solutions in sea-water caused typical cyclopia indis- tinguishable in all respects from that produced in MgCl,. The following strengths were used: 13 M, 14 M, 15 M, 18 M, 17 M, 18 M, 12 mM, and 22 mM. These Mg(NO,), solutions also killed many embryos during the early stages of development. Cyclopia occurred in from 4 per cent to 40 per cent of the eggs in 22 M, 19 M,18M,16M,andi3M. ‘These strengths are comparable to those most effective for MgCl,, both as to the amount of magnesium present and as to osmotic pressure. Mixtures of MgCl, + NaCl; MgSO, + NaCl; and Mg (NO,), + NaCl Mixtures of MgCl, and NaCl were added to sea-water as fol- lows: 12 cc. of a molecular solution of MgCl, was added to 12 cc. of NaCl, and 36 cc. of sea-water was then taken to make the entire quantity up to 60 cc. This solution will be spoken of as 1M + 1M, the first term referring to the MgCl, present and the second to the NaCl. On this basis the following mixtures were used: }M + 4M,4M +1M, 3M + 4M, 5 M + 4 M,1n which the MgCl, was varied and the NaCl kept constant, and 4 mM +} M,+M-+4M,.8,M + 4M,4M + 4M, in which the amount of NaCl was varied also. Such mixtures caused the development of cyclopia, the best results were obtained in } M + 1 M, where as times as many as 25 per cent occurred. The ;4; M + 1M gave in one case 30 per cent of cyclopia. The ~,mM + 4M gave 11 per cent. It will be seen that the amount of MgCl, present in these mixtures 1s less than that necessary to cause similar results when used alone. This is a peculiar fact and one for which I know of no explana- tion. Similar results (Stockard ’o07b) were found with mixtures Artificially Produced Cyclopean Fish 291 of salts in distilled water where the final pressure was less than that of sea-water, the normal medium of theeggs. It is also true that if such substances as the sugars be added to a salt solution, a smaller dose of the salt becomes effective in the presence of the sugar. Morgan (’06) first called attention to this peculiar fact in studying the effects of solutions upon developing frogs’ eggs. This would seem to indicate that the effects were due to osmotic pressure conditions and by slightly raising the pressure with another element the effective agent was assisted in its action, but my lithium experiments (1906 and 1907b) are against such a view. A number of mixtures of MgSO, and NaCl were tried, all giv- ing negative results. Mixtures of Mg(NO,), and NaCl as fol- lows were used: }M + 4M,54°M + 4M and 7, M + 2M. The first two caused eggs to develop cyclopia. ‘These are mix- tures closely similar to the effective MgCl, and NaCl solutions. We conclude that cyclopean monsters are produced in Fundulus eggs by the action of sea-water solutions of MgCl,, Mg(NO,), and mixtures of MgCl, and NaCl and Mg(NO,), and NaCl. No other solutions of the many | have tried during three summers gave similar effects. Other salt solutions and sugar solutions exerting practically the same osmotic pressure also fail to cause cyclopia. Another argument opposed to the view that osmotic pressure is the cause is the fact that Fundulus embryos are so resistant to changes in pressure. Since two Meg salts give similar results when used in sea-water solutions, it seems probable that the action of Mg, either directly or indirectly, is responsible for the result. Eggs have been subjected to this action before the first cleavage, during the two-cell stage and just before going into four cells, with similar results. No attempt was made to determine at how late a stage the cyclopean condition could still be caused, though it could doubtless be induced after the eggs had passed much beyond the four cellstage. The fact is that cyclopia may be caused in an egg which has started its development normally and which would have given a two-eyed embryo. The idea of a germinal origin of the defect in this case seems excluded. Cyclopia in this instance is the result of unusual external conditions. 292 Charles R. Stockard. CYCLOCEPHALI AND ‘‘MONSTRA MONOPHTHALMICA ASYMMETRICA ” The magnesium solutions induce the formation of two distinct types of eye monstrosities. ‘The first type is the typical cyclopean monsters, which exhibits a series of individuals showing various degrees of cyclopia. Beginning with a normal individual having eyes in their usual position, we find others in which the eyes are slightly inclined forward and somewhat closer together than usual; or the eyes are still more approximated and occupy an unusually anterior position. (See the diagram, Fig. 1). Next in the series are individuals with their eyes approximated but still distinctly separate, having two optic nerves and two eyeballs with their choroid coats in intimate approximation. We next find the true cyclopean eye which still shows a double nature having two optic nerves; the retina has a paired arrangement and either one or two lenses may occur, depending upon the degree of distinctness of the two components. ‘This eye generally occupies a ventro-median position and looks forward, inclining slightly downward. ‘The eye in others is completely single, showing no indication of a compound structure; it has one optic nerve, a single retinal arrangement, one lens and one pupil. ‘This is the perfection of cyclopia and many embryos possessing such an eye are apparently normal in other respects, except the mouth and nose. They have a typically bilateral brain and are perfectly capable of free-swimming movements. Passing beyond | this stage of cyclopia, we find embryos which have gone to the extreme and show only a defective antero-median eye. In some indi- viduals the eye is represented merely by a choroid vesicle. The step beyond this is the entire absence of the eye. Diagram Fig. 1 gives a schematic illustration of the various degrees in the cyclo- pean series thus outlined. The histological conditions shown by such a series will be considered beyond. It is important to understand that this series is made up of different individuals showing various degrees of cyclopia and that a cyclopean monster does not pass through these steps in its development. ‘The cyclopean defect is foreshadowed in its final condition when the optic vesicle first separates itself from the brain. Artificially Produced Cyclopean Fish 293 Fig.1 Diagrams of the various conditions of the cyclopean defect as shown by the “Magnesium embryos,” from the normal A to complete absence of the eye G. Fig.2 Diagram of the monstrum monophthalmicum asymmetricum series, from one defective eye B to complete absence of one eye E. 204 Charles R. Stockard The second type of optic defect caused by magnesium is a new monstrosity and may be termed Monstrum monophthalmicum asymmetricum, the monster with one asymmetrical eye. It has only one perfect eye which represents one of the normal pair and occupies the usual lateral position. ‘This eye is in all cases perfect while its mate may be indicated by either a small eye, by a mere cellular mass representing an optic cup, or all indications of the second optic cup may be wanting. (See Fig. 2.) ‘This peculiar one-eyed condition exists in many of the embryos in the magnesium solutions. Had such a defect resulted from a mechanical opera- tion, it would probably have been interpreted to mean that one eye anlage was injured and the other not. With the solutions, however, we get a clear case of the gradual dropping out of one eye by comparing different individuals, and here as in cyclopia the defect is present from the earliest appearance of the eye, and is not due to a gradual degeneration, or arrest during develop- ment. A study of sections of these embryos makes the conditions clearer. a The Living Cyclopean Embryos from the First Indication of the Defect to the Time of Hatching The optic vesicles appear in most eggs when about thirty hours old; at this time the blastopore is just closing and the embryo is well mapped out on the embryonic shield. Many attempts were made to select cyclopean individuals at this stage but it could not be done with a great degree of certainty, since some embryos are always slow in giving off the optic vesicles and these at times appear to have only one, but when examined some hours later are found to be normal. A number of eggs were selected, how- ever, at thirty hours old which proved to be cyclopean on later examination. At about forty hours the defect is plainly detectable so that one may arrange the eggs very accurately into two groups, the cyclo- pean individuals and the normal. After such a separation, none of the normal embryos ever exhibited the cyclopean defect in later stages, although kept in Mg solutions. A number of such tests as this in connection with the study of sections convinced me that Artificially Produced Cyclopean Fish 295 the cyclopean condition existed as such from the first appearance of the optic vesicles, and no subsequent fusion of the two optic vesicles or cups took place after that time. A forty-two hour embryo is shown in Fig. 3. It is seen to be well formed and the optic vesicles are clearly outlined on either side of the head. Fig. 4 illustrates a cyclopean individual of the same age. The single optic vesicle occupying a ventro-median position is shown through the transparent embryo. This young individual with its newly formed optic vesicle shows a typical cyclo- pean condition, and no indication is seen of two separate elements that would later fuse. Other embryos at this age have abnormally twisted cephalic regions and show no indication of eyes, although the cyclopean eye might easily be concealed by the bent brain (Fig. 5). Such embryos at later periods are found to be cyclopean and to have narrow tubular brains showing more or less abnormal bendings. When the embryos are about three days old, the brain has expanded and presents a distinctly bilateral appearance; the optic cups are well developed and the lenses are partially formed (Fig. 6). A cyclops monster at this time has a well formed body and the brain 1s often normal, though in Fig. 7 it is inclined toward the narrow tubular condition and is anteriorly twisted. ‘The ven- tromedian eye is clearly seen through the brain and the outline of its lens is distinct. A somewhat younger, sixty-five hour, embryo is shown in Fig. 8 with a superficially perfect brain and two optic cups intimately approximated. The telencephalon is seen to protrude beyond the eyes, as is the case in the normal individual (Fig. 6). Three four-day embryos are shown by Figs. 9, 10 and 11. The brain and spinal cord at this time are clearly mapped out by a coarse pigmentation, the two hemisphere-like portions (corpora bigemina) of the mid-brain are distinctly formed and the eyes are large with the lens clearly outlined within the cup. A cyclopean monster with a perfectly formed large ventro-median eye is illustrated by Fig. 10. Comparing its brain and other parts with the normal (Fig. 9), one fails to find any important deviations. The abnormal condition of the narrow tubular brained cyclops, 296 Fig. 3 Big. 4 resented in IDs fs Fig. 6 Fig. 7 Fig. 8 Fig. 9 Fig. 10 Fig. 11 Charles R. Stockard Camera .lucida sketches of living embryos from MgCl solutions A normal embryo of forty-two hours, the two optic vesicles present. A typical cyclopean individual of forty-two hours. The singlemedian eye (0. V.) is rep- circular outline. An embryo of same age, twisted brain, no optic vesicle shown. Normal seventy-two hour embryo. Cyclops of same age. The eye, op.c, in ventro-median position. Sixty-five hour embryo, two ventrally approximated optic cups. Normal four day embryo—bilateral brain outlined. Four day cyclops, large ventro-median eye and typically bilateral brain, op.c, the eye. Four day cyclops with narrow tubular central nervous system. Figs. 12 and 13 Five day cyclops, narrow tubular brain with waist-like constrictions dividing them into fore, mid and hind-brain regions. Ventro-median eye. Fig. 14 Five day cyclops with ventro-median eye and dorsally humped brain. 297 Artificially Produced Cyclopean Fish 298 Charles R. Stockard Fig. 11, is evident. Fig. 12 shows a common type of cyclopia with the three primary brain regions separated by waist-like con- strictions. ‘Iwo other variations of the narrow tubular condition are found in Figs. 13 and 14. The embryos are five days old and no changes of importance occur from this time until the hatching period is reached, except the usual progressive development of the eye structures. The normal embryos generally begin hatching when about twelve days old, one cyclopean monster hatched at this time but most such indviduals were much later than the normal in coming out. A twelve day cyclopean fish is seen in dorsal view in Fig. 15 and ventrally in Fig. 16. The large cyclopean eye projects forward and occupies the position usually taken by the mouth at this time. A slight indention along its mid dorsal line suggests its double nature, although the ventral view (Fig. 16) shows this same eye to possess only one pupil and lens. ‘The brain of this specimen is practically normal. An embryo with the two eyes intimately approximated is shown in front view in Fig. 17. The eyes are joined and each looks forward in a direction slightly towards the side to which it belongs. A common variety of cyclopean fish is one in which the eye is unusually small and occupies an extremely anterior position; Fig. 18 shows such an embryo. ‘This variety is usually unable to hatch, although a few were assisted in breaking through the membrane. They swam rather abnormally, owing to a twisted condition of the body. A dorsal and ventral view of a cyclopean fish is shown in Plate I, Figs. A and B. This indicates the striking appearance pre- sented by these embryos. b Free-Swimming Cyclopean Fish Many embryos, showing the cyclopean defect in various degrees, hatched normally and were capable of swimming in a manner indistinguishable from ordinary two-eyed fish. These monsters gave many indications of ability to see. ‘They went to the more brilliantly lighted side of the dish with the normal ones. ‘They darted away in a normal fashion when any object was placed in front of the eye, while similar objects put at equal distances from Artificially Produced Cyclopean Fish 299 their tails caused no excitement. In two instances they lived for ten days, which is about as long as the two-eyed embryos can survive without food. At this time the entire content of the yolk- sac has been absorbed. ‘The embryos in nature doubtless begin feeding previous to this stage. The cyclopean individuals appear to be as active as the normal and their ability to live would seem to depend only upon the possibility of their obtaining food. A normal fish eight days after hatching is illustrated by Fig. 23. The mouth projects forward beyond the dorsal tip of the head and the two eyes are lateral in position. A cyclopean embryo eight days after hatching is shown in Fig. 24. Here the two eyes are united and occupy the position which the mouth has in Fig. 23. In Fig. 25 a perfectly cyclopean eye is shown in dorsal view: the same individual is seen in lateral and ventral views in Figs. 26 and 27. This fish swam in a normal manner. In the lateral position the mouth is shown projecting ventrally as a pro- boscis-like structure. This condition is due to the fact that the single antero-median eye occupies the position normally assumed by the mouth and thus obstructs the usual forward growth of its structures. ‘The mouth, therefore, remains ventro-posterior to the eye and grows downward, presenting the proboscis-like appearance. Such a condition recalls in a striking way the nose of the mam- malian cyclops. In mammals the cyclopean defect is accom- panied by a proboscis-like nose situated in the forehead above the median eye. ‘The nose in normal development grows downward to its facial position, but in cyclopia the median eye obstructs its path and forces the formation of the proboscis-like organ in the forehead. The same explanation holds for the fish’s mouth where the eye prevents its forward growth, producing the proboscis- like organ. It is interesting to find that the mouth in cyclopean fish stands in a position so as to fall in the gill series as number one, all the gills and the mouth have the same general direction. I have found that in Bdellostoma the mouth arises in a manner similar to the gills and actually at first arches dorsally and only secondarily arches ventrally. It may have originally been a member of the gill series, as Dohrn (1875) has long thought. It would be 300 Charles R. Stockard Camera sketches of the living embryos in magnesium solutions Fig. 15 Dorsal view of twelve day cyclopean monster, the antero-median eye with furrow indicating its double nature. Fig. 16 Ventral view of the same individual, the eye possesses a single pupil and lens. Fig.17 A twelve day embryo, ventral view showing two eyes intimately approximated. Fig. 18 Fourteen day embryo. Small extremely anterior cyclopean eye with protruding lens, extreme cyclopia. Fig. 19 A five day Monstrum monophthalmicum asymmetricum; the left eye has no mate. Fig.20 A similar twelve day monster lacking its left eye. Fig. 21 Anincomplete diprosopus monster seventy-two hours old. Two brains, two normal lateral eyes and one perfect middle eye, the other middle eye indicated by the circular lens L. Fig.22 The same monster when eighteen days old, three perfect eyes. The embryo hatched three hours after this drawing was made and swam abnormally. Artificially Produced Cyclopean Fish 301 a 302 Fig. 23 Fig. 24 Fig. 25 Fig. 26 Charles R. Stockard Camera sketches of free-swimming fish Normal individual. M, its anteriorly placed mouth. Incomplete cyclops, two eyes joined and occupy the position usually taken by the mouth. Dorsal aspect of a perfect cyclops. Antero-median single eye. Lateral view of same. The mouth M is forced by the eye to remain in a ventral position and hangs down as a proboscis-like structure. Fig. 27 Ventral view of same fish, note perfectly single eye, one lens and one pupil, ys., yolk-sac. Artificially Produced Cyclopean Fish 303 SS, ~~ ‘ ¥ ; fea ee ; CAC iy Lee M MgCl,. The optic process is situated laterally and no indication of a like process exists on the other side. Fig. 30 A slightly oblique section through the cyclopean optic vesicle of a forty-nine hour embryo from +o M MgClby, of.v. optic vesicle. Fig. 31. Cross section through double cyclopean eye of fifty-four hour embryo from oo M MgCh, op.c. optic cup; L, lens thickening of ectoderm; Br, normal bilateral brain. Fig. 32 Section of single cyclopean eye in similar embryo. L, lens; of.c. optic cup, small solid diencephalon above; X, guide figure indicating the plane of the sections. 308 Charles R. Stockard is given in Fig. 33. Such a section is most instructive. The condition of the eye is much the same as that shown by the trans- verse section, Fig. 31. The cup is double and two ventral lenses are present. ‘The section passes below (ventral) the diencephalon so that no part of it shows; the telencephalon is indicated in front of the eyes and a thickening of the forward ectoderm shows the nasal plate, posteriorly or behind the eyes the mid-brain is cut in horizontal section. A sagittal section of a typical cyclopean embryo is shown by Fig. 34. Here we see the eye and the brain in the third dimension. The telencephalon in front, the diencephalon above the eye, and behind this the large mid-brain with a spacious median cavity. In front of the eye is also shown a median ectodermal thicken- ing, the double nasal pit. The eye is single and exactly ventro- median in its position and connects in a more lateral section with the brain at about the point where the telencephalon and dien- cephalon join. The lens and retina are differentiating into their typical structures. One may obtain a clear mental reconstruc- tion of the cyclops monster at this age by comparing Figs. 31, 32, 33 and 34, the transverse, horizontal and sagittal mid-planes of the cyclopean eye. The early stages just described illustrate the cyclopean defect in its various degrees, and the eye throughout its development retains the original condition of singleness or doubleness. No evidence whatever can be found of subsequent fusions during development. ‘Two clearly approximated eyes arise in that con- dition and remain so without fusing to give a double cyclopean eye, and a double eye never attains to the single condition by a more intimate union of its parts. The statement made in my (1907a) former paper, p. 257, that “the fusion of the two components may take place at different periods within a certain limit” is incorrect, as I (1908) have pointed out in a short note on the subject. This statement was one of interpretation and was based on a comparison of late embryos which showed different degrees of cyclopia. It seemed from such an incomplete study that the eyes were more or less double or compound, depending upon the stage in development at which they had become approxi- Artificially Produced Cyclopean Fish 309 So CEP) Fig. 34 a Fig. 33. Almost horizontal section through a double cyclopean eye of a fifty-four hour embryo in 46 M MgClo. See guide figure X for plane of section. Np., nasal plate; L, lens; op.c., optic cup; Tc., telencephalon; Mb., mid-brain. Fig. 34 Sagittal section (guide figure 7) through typical single cyclopean eye showing its ventral position below the diencephalon Dc. The nasal pit, Np. is median; L, lens; op.c. optic cup; Tc., telencephalon; Mb., mid-brain. 310 Charles R. Stockard mated. ‘The point is one which can only be proven by a number of direct observations on all ages of cyclopean embryos and care- ful study of sections; such a study has convinced me that no fusion of the eyes takes place after they are once clearly given out from the brain. It seems advisable for later stages to consider groups of embryos showing various degrees of the cyclopean defect. b Incomplete Cyclopia; Double Eyes Under the term incomplete cyclopia may be considered indi- viduals with eyes abnormally close together although separate Among Fundulus embryos such individuals exist and a series of stages connect these embryos with those in which the two eyes are intimately connected or joined together. An individual of this kind when sectioned will show the eyes as in Fig. 35. ‘This sec- tion is from a four day embryo, the two eyes are united in the median line of the head and both are perfect eyes with a lens, single retina and one optic nerve. The choroid coat as indicated by the heavy line is just beginning to form. Fig. 36 shows a section of two eyes which are more intimately united. ‘This case is the common “‘hour-glass” eye of cyclopia. The two eyes are independent, except for their waist-like connection and each has its lens, single pupil, retina and distinct opticus. The optic nerve of the right component is seen entering the optic cross at the base of the brain. The brain in this embryo is remarkably perfect, as it is in many cyclopean monsters, and [| see no reason whatever for attributing the defect to a “single brain” or any other gross malformation of the cephalic region. Many embryos with deformed :brains possessed two normal eyes and the converse is true, many normal brains were accompanied by cyclopean eyes. Leaving the “hour-glass” eye, we find the double-eye shown in Fig. 37, having a common optic chamber each half of which is supplied by one component. ‘Two lenses and two pupils are present and generally two optic nerves, although they may run so nearly parallel that the two are difficult to distinguish. A single nasal pit is present in the embryo from which Fig. 37 1s a section. All of the cyclopean monsters possess two distinct auditory vesicles. Artificially Produced Cyclopean Fish BUI Fig. 38 is a section through a unique double eye; no other such case was found. The two retinal components are connected along their median dorsal line within the brain and extend down facing one another. ‘They are like the two sides of a leguminous pod; between the two a single lens is placed suggesting the seed in the pod. Enclosing the ventral part of the retinal components is a choroid coat shown in heavy black. This choroidal coat does not fully encompass the retinal areas, a part of which extends dorsally far up into the brain. The anterior end of the eye is V-shaped in section. ‘The optic cup anlagen in this case must have been closely united from their first origin in the brain, since por- tions of the retinal region are still contained within the brain itself, yet during development they did not fuse into a single eye. A single nasal pit is present and the mouth is ventral and _pro- boscis-like. An almost single eye is indicated in section, Fig. 39. The choroid coat surrounds the retina, the latter showing slight traces of its compound nature. ‘Two lightly staining regions of nerve tissue are seen and the entire eye is unusually wide laterally. The single lens is normal. ‘The brain here is also normal and the eye occupies a ventro-median position. A further union of the eyes gives the c Perfect Single Cyclopean Eye and Normal Brain The cyclopean eyes are in many cases perfectly single, resem- bling in all respects, except their position, one eye of a normal pair. They are placed immediately ventral and their antero-posterior mid-plane is in the median line of the embryo. The brain in such a cyclops is often normal in all general respects. Figs. 40 and 41 represent horizontal sections through the brain regions of such a cyclopean fish when seventy-seven hours old. Fig. 40, the more dorsal section, passes through the mid-brain and shows the two lateral, hemisphere-like bodies (corpora bigemina) with well formed cavities. Behind these the section cuts the floor of the hind-brain for some distance and finally crosses it where the head bends. Passing ventrally through a number of sections, we find the one shown in Fig. 41. Here only a small ventral 312 Charles R. Stockard Transverse sections of different degrees of double cyclopean eyes Fig. 35 Section of eyes in four day embryo, the two eyes united. Choroid coat beginning. Fig. 36 Section of ‘“‘hour-glass”’ eyes, the optic nerve of the right component entering the normally bilateral brain. From a sixteen day embryo, the retine and lenses differentiated. r.o.n., right optic nerve; Ch., choroid coat. Fig. 37 Section of eye in hatched embryo. Double-eye with two pupils and two lenses. Retina undifferentiated. Fig. 38 Hatched cyclops, section through the peculiar eye with two components facing and lens between them (see text). Fig. 39 Section through almost single cyclopean eye, only indication of its compound nature paired retinal arrangement. Brain normal. Guide figure X indicates the plane of all sections and the eye position in the several specimens. Artificially Produced Cyclopean Fish 313 ee DO aHOp AGAMA vag 00K? 0010) CSO ee, ae SBN ss S83 S00, yA 700% 314 Charles R. Stockard part of one of the corpora bigemina is cut and the completely single eye with its lens is found lying ventrally and in a median position. ‘The double olfactory pit is seen in front of the eye and somewhat to one side of the head. ‘The posterior part of the section runs below the hind-brain and finally cuts it as the head bends just in the middle region of the well formed auditory ves- icles. The section thus presents the three sense organs, the single cyclopean eye, the nasal pits united into a double pit; the paired ear vesicles alone are in their usual positions. A transverse section through the eye of a four day embryo is illustrated in Fig. 42. The retina is unusually wide laterally but no other indication of doubleness is shown. The choroid coat is beginning to form and the eye is connected with the floor of the brain by a single cellular stalk. ‘The retina at this age is only slightly differentiated and there is no arrangement into layers. ‘This embryo has two distinct nasal plates. Several of the cyclopean fish show the nasal plates separate, although they are usually represented by an anterior double plate near the middle line. A nine day embryo of which Fig. 43 represents a section through the eye has a finely developed brain, well expanded laterally and perfect in general shape and structure. ‘The eye is completely single and the retina is partially formed into layers; the lens is almost transparent and the vitreous humor is being formed about it. The eye has all structures closely similar to those in a paired eye of this age and would doubtless have functioned had the embryo hatched. This specimen has a single nasal pit. Another cyclops of perfect structure when studied in sections at thirteen days old showed the mouth posterior to the eye, hang- ing as a ventral proboscis-like mass. “Iwo nasal plates were present and the eye was single. ‘This eye, Fig. 44, was unusually far forward and although the retina was well differentiated into layers the humor had not perfectly formed behind the lens. The small section of the brain is shown in Fig. 44 to be bilateral and not unusual in appearance. Passing forward through the series of sections to a place where the anterior end of the cyclopean eye- ball stops, a minute lens is found lying in a ventro-median position, Artificially Produced Cyclopean Fish 315 Fig. 45. This lens, although only nine micromillimeters in diameter, has differentiated and shows perfect lens fibers arranged in the usual concentric fashion. It has no connection whatever with the eye, nor with any part of the central nervous system. The small lens doubtless originated and differentiated its tissue in an independent manner. ‘The independent origin and self- differentiation of lenses will be clearly shown in a following sec- tion of this paper. Fig. 45 also illustrates the two lateral nasal plates in section. The cyclopean eye is thus seen to be at times single in nature, showing no trace of a double composition. ‘This may be con- sidered the climax or perfection of cyclopia, if such an expression is permissible. Eyes not completely united, or double-eyes, are the incomplete or imperfect cyclopean condition, while the single condition reduced or distorted may be termed extreme cyclopia. d Extreme Cyclopia: From the Abnormally Small Anterior Cyclopean Eye to Entire Absence of Eyes Many cases are found representing the condition of extreme cyclopia. They may be considered in order, beginning with the least modified. In discussing the living embryo mention was made of those with a small cyclopean eye placed far forward (Fig. 18). Sections of such eyes show them to be of a more or less imperfect nature and sometimes deeply buried in the tissues of the head. Fig. 46 shows a section through the small eye of a hatched embryo eighteen days after fertilization. This eye is placed in the extreme anterior tip of the head and the section shows on the right side pigment spots which lie on the front end of the forehead. The eye is unusually small and the living embryo was abnormal, being unable to swim directly forward. The nasal pits are united in the anterior eye region and a pro- boscis-like mouth is situated ventrally. Two still more abnormal cyclopean eyes are shown in trans- verse section by Figs. 47 and 48, both from thirteen day embryos. In Fig. 47 the eye is close to the single olfactory pit, the retina 1s differentiated into layers, but the lens is larger than the optic cup so that it cannot fit completely into it. The brain of this individ- 316 Charles R. Stockard Sections of perfectly single cyclopean eyes Fig. 40 Horizontal section through mid-brain showing its corpora bigemina, Cb, and floor of hind brain, Hb, in seventy-seven hour cyclops. Fig. 41 A more ventral section of same series, E, the Cyclopean eye; o/.p., olfactory pits united. Hb, hind-brain and Av., auditory vesicle; Cb, floor of one mid-brain lobe. Guide figure X gives plane of each section. Fig. 42 Trans-section of a four day single cyclopean eye in exact ventro-median position. ch, choroid coat; p, pigment spot. — Fig. 43 Similar section of nine day eye. Humor cavity behind the lens. Note perfectly bilateral brain. , pigment spot. Fig. 44 Section of single median eye below perfectly bilateral brain, thirteen days old. Fig. 45 A more anterior section in same series as Fig. 44. The forward tip of the eye ch is seen. A small lens L lies free near the ventral ectoderm; o/.p., olfactory pit; p, pigment spots on anterior end of brain. Guide figure X indicates plane of all sections. Artificially Produced Cyclopean Fish 317 318 Charles R. Stockard ual is abnormal and the eye is out of the median line. The em- bryo of Fig. 48 was abnormal with the brain distorted so that the cyclopean eye was slightly to one side and far out beyond the head. The retina differentiates into layers but the lens lies out of the central position, and would be unable to function efficiently. A peculiar condition is found in the embryo from which sec- tions shown in Figs. 49, 50 and 51 were taken. ‘This very small eye Was again in an extremely anterior position, though almost in the median line. The lens is as large as the optic cup and pro- trudes far out beyond its edge. Fig. 49, the most anterior section of the three, passes through the great circle of the spherical lens and shows it entirely outside the optic cup. On passing back in the series to where the lens is less in size, we reach the anterior edge of the optic cup and choroid coat, Fig. 50. Continuing back in the series of sections, the lens disappears and the optic cup alone is shown in Fig. 51. ‘The lens in this eye is clearly too large for the accompanying cup as was also the case with the two eyes just described. ‘The size of these lenses is, therefore, independent of the size of the optic cup. Lewis’ (’04) idea that the cup regulates the size of the lens does not apply to these embryos, nor does the rule for the amphibian that the origin of the lens is dependent upon the influence of the cup. A step beyond this condition of a small anterior eye with its ill-fitting lens may be illustrated by an embryo in which the eye is a minute choroidal sphere buried in mesenchyme below the brain and in the median line. In life this specimen seemed entirely eyeless, but sections showed this small eye-like structure (Fig. 52) in the position typically taken by a cyclopean eye. Such cases as this emphasize the necessity of sections in order to cor- rectly interpret the conditions of cyclopia and conclusions based only on superficial studies are necessarily unreliable. The nasal pits were in the normal lateral position. Passing back in the sec- tions to the region usually occupied by the two eyes, it will be seen that on one side a typical lens occurs (Fig. 53). The lens is well differentiated and completely isolated from all connections with either nervous or eye tissue. A band of muscle is seen in the figure to touch the inner edge of the lens. Artificrally Produced Cyclopean Fish 319 The occurrence of this lens recalls at once Herbst’s (’o1) argu- ment regarding the independent origin of the lens. He held that “af the lens really developed independently of the optic cup, then in the case of median cyclopia the two lateral lenses should arise in their usual positions; but they do not, and furthermore, the cyclopean cup gets a lens from ectoderm out of the usual lens- forming region.”” The Fundulus embryos show lenses arising at times in their usual places and often in other places, independ- ently of the optic cup. We may suppose that in these embryos certain areas of the ectoderm are at times out of their normal posi- tions, and thus explain the promiscuous distribution of independ- ent lenses. Finally, embryos exist in which no indication of the optic cup can be found, these may be said to have passed beyond the ex- treme cyclopean condition. They are not ordinary individuals that are merely blind, since the mouth is usually distorted and sometimes the snout-like structure which accompanies cyclopia is present. This suggests the possibility that the ‘proboscis- mouth” is not entirely due to its normal position having been usurped by the cyclopean eye. Some of these embryos have free lenses and others no optic parts at all. Figs. 54 and 55 are two transverse sections from the same embryo, the anterior one shows a lens lying against the olfactory pit but free from all connection with the central nervous system. Fig. 55 shows a second lens lying close against the brain tissue. This embryo has no indication whatever of optic cups, and seemed eyeless in life. Other indi- viduals when carefully examined in section had neither an optic cup nor any lens-like structures. We have thus reviewed a series of forms beginning with the usual two-eyed embryos and passing through all degrees of double eyes to single cyclopean eyes, to extremely small cyclopean eyes, to individuals finally with only lenses present and no optic cups and others with neither lens nor cup. 220° Charles R. Stockard The extreme cyclopean condition Fig. 46 Cross-section of hatched embryo, small cyclopean eye located in anterior tip of head. The nose is anterior to this section. , pigment on “‘forehead”” of embryo. Fig. 47 Section of thirteen day embryo. Small cyclops eye with large lens, differentiated retina and abnormal brain partly surrounding the eye; o/.p., nasal pit. Fig. 48 Section of thirteen day cyclops with eye far forward and out of median line beneath an abnormal mass of the brain. Figs. 49, 50 and 51_ Sections of a small anterior cyclopean eye with large lens projecting out of optic cup. The first section Fig. 49, is most anterior, the great-circle of the spherical lens, Fig. 50, tip of lens in the edge of optic cup, and Fig. 51, center of optic cup behind the lens. Figs. 52 and 53 Sections of thirty day embryo which seemed eyeless in life. Brain abnormal. Fig. 52, the cyclopean eye is represented by a choroid vesicle, E. The more posterior section, Fig. 53, shows a perfect lens L, in the usual lateral position, but no optic cup exists. A band of muscle m is between the lens and brain. Figs. 54 and 55 Sections of two lenses L, one forward by the olfactory pit, o/.p., the other more pos- terior and surrounded by brain tissue. No optic cup present in this nine day embryo. Artificially Produced Cyclopean Fish S21 322 Charles R. Stockard INCOMPLETE DIPROSOPUS WITH THREE EYES AND ONE ADDITIONAL LENS A most valuable object for study was an incomplete diprosopus monster which appeared in my solutions. ‘This individual had two heads separated as far as the lateral eye region. It appeared as indicated by Fig. 21 when seventy-two hours old. The two brains are separate, almost back to the auditory vesicles. “Iwo normal eyes are shown in outer lateral positions while between the heads one eye, perfect in shape, is mated with the outer eye of the left head and a circular body occupies the usual position of left eye on the right head. The embryo seemed normal in other respects and was in a vigorous condition. The monster when eighteen days old had developed to the usual size and was still hardy. At this time it presented a striking appearance as indicated imperfectly by Fig. 22. Three large eyes normal in form and capable of movement looked out from the double head. All visible evidence of the circular body shown near the middle eye when seventy-two hours old had disappeared. The middle eye was clearly paired with the left eye of the left head component and the right eye of the right head seemed mateless. A single pair of auditory vesicles were present. ‘The young fish respired and twisted vigorously within the membrane. ‘Three hours after this drawing was made, the embryo hatched and swam about in a circular fashion, the body not straightening perfectly. The free living animal was kept for five days and then preserved for sectioning. The sections show the presence of two brains, one spinal cord and one normal mouth leading into a pharynx with its series of gills, while a second short throat is present in the right head. There are two notochords back to the middle of the yolk-sac and one from there on. The rear end of the medulla becomes single and only one pair of ear vesicles are present. There are two olfactory pits anterior and median to the lateral eyes. Three perfectly normal eyes exist. “They possess clearly dif- ferentiated retina, irides, humor chambers and lenses. “Two of these eyes are connected in the usual way with the brain of the Artificially Produced Cyclopean Fish q22 left head and one with the brain of the right. Fig. 56 is a section showing the middle eye somewhat back of its center so as to bring the edges of the other eyes into the figure. The middle eye 1s more anterior in position than the two lateral ones, owing to the slight obliquity of the left head. A distinct lens is shown in the cup in Fig. 56. On going backwards in the series we reach a sec- tion passing through the middle of the two lateral eyes and the posterior end of the middle eye (Fig. 57). [he section shows dorsally the huge double brain and ventrally a central throat and most interesting of all a fourth lens. ‘This lens lies against the outside choroid coat of the middle eye and is in just the position (recognizing a displacement due to development of the middle eye) to be the lens of the left eye of the right head, if such an eye were present. We thus have in this double head three typical eyes and the fourth represented by a free lens. It was impossible to detect the clear lens in the living embryo which emphasizes again the necessity of sections for a definite interpretation of the conditions existing in these monsters. Conclusions drawn from observations on the living eggs without the comparison of sections may be incomplete. The sections further make clear the nature of the circular outline shown against the middle eye of the seventy- two hour embryo (Figs. 21 and 57). Comparing the figures of sections and those of the whole embryos, it will be remembered that the sides of the sections are transposed, since the drawings of the total embryos are made from a simple microscope and the sections from a compound microscope which inverts the image. This incomplete diprosopus monster increases the series of eye monstrosities so that it passes through the cyclopean group to beyond the normal. The diagram (Fig. 58) illustrates in a simple way the various conditions we have considered and emphasizes the continuous nature of the series. Beginning at one end with eyeless individuals, we pass gradually through a series with small buried cyclopean eyes (which may be indicated in the diagram by a palpebral opening, such as similar mammalian cyclops would show), to the perfectly single cyclopean eye, to the double eye with one lens and pupil, to the hour-glass eye with two lenses and two pupils, to two independent but closely approximated eyes, next to 324 Charles R. Stockard NON GE (NW Fig. 56 Section through anterior median eye and edges of lateral eyes of hatched incomplete dipro- sopus; Br; Bro, the two brains ‘ Fig. 57 More posterior section through middle of lateral eyes, posterior part of middle eye, and an additional fourth lens L. Guide figure X makes both sections clear. Artificially Produced Cyclopean Fish G25 the normal condition and finally beyond to the incomplete dipro- sopus with three eyes and a fourth lens. The idea of arranging monsters in such a series including the normal is due to Prof. H. H. Wilder. V Vi Vil Vill Fig. 58 Diagram of the various conditions shown by the ‘‘magnesium embryos” from entire absence of the cyclopean eye J, to deeply buried eye IJ, perfect single cyclopean eye IIT, double-eye, IV, two approximated eyes V, eyes unusually close together VJ, normal V JJ, three eyes and fourth lens VIII. The normal is a mean from which different degrees of abnormal- ities are but greater or less deviations. It is possible to arrange almost any type of abnormality in such a series. Supernumerary arms or legs on one side might exist in various individuals in dif- ferent numbers down to the single normal one; other specimens could be found showing degenerate or small arms and finally armless or legless individuals are known. MORPHOLOGY OF MONSTRA MONOPHTHALMICA ASYMMETRICA A brief description of the asymmetrical monophthalmica mon- sters in life has been given above, but their true nature and structural conditions are impossible to detect without sections. It is found that here again a continuous series exists, beginning with the ordinary two-eyed individual through all gradations to the complete disappearance of one eye. The section through the middle of the eyes in a normal embryo of thirteen days old is illustrated in Fig. 59. The eyes, of course, are equal in size and alike differentiated structurally. In the salt solutions, however, many embryos occur with one eye perceptibly 326 Charles R. Stockard smaller than its mate. A section through the eyes of an embryo of this kind when seventy-six hours old is shown in Fig. 60. ‘The left eye is decidedly smaller than the right and possesses a cor- respondingly small lens. From the comparative study of a num- ber of individuals it may be safely stated that this difference in size between the two eyes will not be overcome later, nor on the other hand will the small eye degenerate or disappear. “The em- bryo will hatch with its eyes in dissimilar conditions comparable to the state of things shown by this seventy-six hour stage. ‘The brain is normal and two nasal plates are present. An embryo closely similar to the one just described was sec- tioned after hatching. Its large eye appears as in Fig. 61. More anterior sections show a small eye looking forward with a some- what protruding lens in its cup. Behind this small eye is another lens lying free in the ectoderm (shown in Fig. 61). ‘This lens is perfectly differentiated and appears to have arisen independently. A further reduction of the eye is shown by Fig. 62. In this thirteen day embryo the left eye is perfect and the right is rep- resented by a small cellular mass lying close against the brain. The lens of the right side is entirely wanting. In life the head was slightly one-sided, obviously on account of the asymmetrical eye development; no indication of the cellular mass could be detected and the embryo seemed truly one-eyed. A section of another seventy-six hour individual which in life also seemed to be one-eyed is illustrated by Fig. 63. The brain is normal and almost bilaterally symmetrical, an ordinary left eye exists but there is not the trace of an indication of the right optic cup. An ectodermal thickening represents the right lens in process of formation in the position that it would typically occupy. This lens anlage must have arisen independently of a stimulus from an optic cup and is well removed from the brain, so that no direct stimulus from that source can be responsible for its appearance. Other one-eyed individuals showed complete absence of all parts of the second eye, the lens as well as the optic cup failing to arise. The occurrence in the Mg solutions of these one-eyed embryos as well as the cyclopean embryos suggests that the chem- Artificially Produced Cyclopean Fish BaF Monstra monophthalmica asymmetrica Fig. 59 Section through eyes of normal thirteen day embryo. Fig. 60 Section of seventy-six hour embryo with one normal and one small eye and perfect brain. Fig. 61 Section of the normal eye of a hatched embryo; a small eye with a lens is situated more ante- riorly on the other side and behind this is a third lens, L, shown in the figure on the left side. Fig. 62 Section of normal eye in thirteen day embryo, the other eye is represented by the cellular mass, e, close against the brain. Fig. 63 Section of normal eye in seventy-six hour embryo, the brain is bilateral and perfect, but no indication exists of the right optic cup although the ectoderm of that side has formed a lens thicken- ing L. 328 Charles R. Stockard ical influence exerts a peculiar inhibition of that process of out- pushing or separation by which the optic vesicles arise. Such an idea will be more fully considered in the general discussion given below. The unequal eyes may possibly result from an unequal allotment of eye material to one side or the other. A major portion might go to the right side and a minor part to the left, or the entire eye an- lagen might by chance occur on one side. ‘This in a sense would be lateral cyclopia. Such reasoning is of course purely hypothet- ical. INDEPENDENT ORIGIN AND SELF-DIFFERENTIATION OF THE CRYSTALLINE LENS Spemann (’o1), Lewis (’04), and others have concluded from experiments on amphibian embryos that there is no localization of lens-forming material in any given area of the ectoderm. ‘They further held that the formation of a lens is dependent upon a stim- ulation of the ectoderm through contact with the optic-vesicle or cup. Spemann (’o5) in discussing the question of the self-dif- ferentiating power of the lens concluded from a consideration of Schaper’s (’04) experiments on the frog that the lens 1s not capable of self-differentiation, but that a continued influence or contact of the optic-cup is necessary to cause the lens-plate or lens-bud to develop into a typical lens. LeCron (’07) has recently shown that the lens in Amblystoma is not self-differentiating. I (’07d) found in embryos of the blind Myxinoid, Bdellostoma stouti, that a lens-thickening formed in early stages while the optic-vesicles were near the ectoderm. During development the optic cup becomes distantly removed from the ectoderm and the lens-plate disappears as if it were unable to continue development independ- ently of the optic cup contact. On the other hand Mencl (’03) has claimed that the lens in Salmo salar is at times formed independently of the optic cup influence and Spemann (’07) has recently modified his attitude. Spemann finds that in a certain species of frog, Rana esculenta, the lens may arise independently of the optic cup. This lens also Artificially Produced Cyclopean Fish 329 continues to develop and differentiates typical fibers. Most con- clusive evidence favoring the independent origin and self-differ- entiation of the lens is furnished by the Fundulus embryos now under consideration. Attention has been called repeatedly to the occurrence of lenses having no connection with other optical parts. It may be well at this time to summarize these cases which clearly show that in Fundulus the lens may arise independently and continue its devel- opment and differentiation. Fig. 63 illustrates the budding off of the lens from ectoderm on the side of the head which lacks entirely an optic cup. Fig. 61 shows a lens fully differentiated though lying freely in the mesen- chyme of the head. It will be recalled that this is a supernumerary lens; the large and small eyes of the embryo both possess lenses. An optic cup can not be responsible for this third lens. Fig. 57 of the incomplete diprosopus shows the fourth lens of the double head entirely outside the optic cup of the third eye which possesses alens. Figs. 54 and 55 show two lenses in an embryo that pos- sessed no trace of an optic cup. Fig. 53 indicates a lens in its usual position but no optic cup is present. In Fig. 45 a tiny lens is found in front of a cyclopean eye which possesses its own lens. Many other similar illustrations could be given. No one could hold that this indiscriminate collection of lenses, all of which are entirely isolated from any connection with optic cups or other eye parts, as well as in nearly all cases from the brain itself has arisen through direct stimuli derived from the optic cups. It is also evident that the lens after its formation continues to self- differentiate. It seems to me that in Fundulus the case is clearly proven that lens formation does not depend upon a direct stimulus from the optic cup. Such a dependence as advanced by Lewis (’04) for the frog is not, therefore, of universal application, nor is the view tenable that the differentiation of the lens depends upon a con- tinued stimulus from the optic cup. 330 Charles R. Stockard DISCUSSION AND CONCLUSIONS The foregoing facts furnish important information as to the cause and manner of development of cyclopia, and the facts bear directly on previous ideas concerning this subject. By treating the fish eggs with magnesium solutions, it is conclu- sively shown that the experimenter has the power without mechan- ically injuring the egg or embryo to cause what would have been a two-eyed individual to become a cyclopean monster. ‘This undoubtedly is a case of the occurrence of cyclopia through the action of external influences on the developing egg. I conclude, then, that cyclopia does not necessarily result from germinal varia- tions, but I make no claim that it may never arise in such a way. On the contrary, there is no reason why cyclopia should not occur through germinal variations as readily as does any other new fea- ture. ‘The fact that mammalian cyclopean monsters do not sur- vive, or even if it be proven that the free-swimming cyclopean fish are incapable of living or reproducing, does not argue against the possibility that cyclopia may in cases be due to germinal variation. Such a statement is emphasized by a case I (’07c) recently recorded. In a flock of sheep in North Carolina two entirely legless lambs appeared in the spring of 1907. Again in 1908 two other similar lambs have occurred, one being the offspring of a mother which had previously borne a legless individual. These lambs were unable to feed without assistance and in nature would doubtless have died shortly after birth, but their peculiar occurrence in this flock is very probably due to germinal variations, either within the mother or father, or both. Students of inheritance consider sports to be due to germinal variations and the ability of such sports to survive depends merely on their adaptations to the sur- roundings and not in the least on their manner of origin. No reason can be given why a cyclopean individual might not occur as a sport due to sudden germinal variations. From the experi- ments contained in the present paper, however, it may be emphat- ically affirmed that cyclopia is not always due to germinal origin. Spemann (04) through an ingenious ee of experiment, produced double-headed Triton embryos which exhibited various Artificially Produced Cyclopean Fish aon degrees of cyclopia. ‘The eggs of this salamander when constricted about the periphery of the first plane of cleavage with a fiber- like ligature gave monsters with two equal heads. When the lig- ature was oblique with reference to this plane one of the heads was cyclopean to a greater or less degree. Spemann thought the defective head due to the loss of the anlagen of certain parts, con- sequently these parts never began development and organs sit- uated lateral to them developed in contact from the start. In other words parts between the eye anlagen fail to form and thus the anlagen come in contact and so develop from the beginning. ‘This explanation is of course entirely speculative, but it is supported in a manner by experiments which according to Mall (08) Lewis has performed on the fish embryo. Mall states that Lewis found by pricking the extreme anterior end of the embryonic shield in Fundulus eggs that many of the eggs develop into cyclops embryos. It was found in some that the prick had destroyed the “nose” only. ‘This experiment shows conclusively that it is the absence of tissues between the eye arlagen that allows them to come together and unite.” The above explanation no doubt holds for some cases of cyclo- pia produced by cutting or pricking; there it is evident that tissue is destroyed and the destruction of median tissue may cause the regions containing the eye anlagen to unite. It is difficult to apply this explanation to all cases. In the “Magnesium embryos,” why should tissue between the eyes fail to form and not other tissues; why are the nasal pits united in some cyclops and separate in others? A close microscopic examination of the brain floors in cyclopean and two-eyed embryos shows no absence of recog- nizable parts in the former. “The monstra monophthalmica asym- metrica are also to be explained; here one eye in some cases fails to come off from the brain. Is this due to the absence of its early anlage? The very small cyclopean eye sometimes buried deeply in the head, and the eye shown in Fig. 38 which 1s partly inclosed within the brain, as well as the entire absence of an eye, suggest another explanation that may apply to all cases inthe magnesium solutions. The small eyes close together, cyclopia in various degrees, the 332 Charles R. Stockard imperfect formation or absence of one eye and entire absence of eyes are all conditions common to the magnesium solutions and very rare or never occurring in other solutions, nor in the hun- dreds of eggs observed developing in sea-water. The conditions are, therefore, probably due to a common cause, and I suggest hypothetically that this cause is an inhibitory or anesthetic effect of the magnesium on the process of outpushing and separation of the optic vesicles. Magnesium exerts a decidedly anesthetic effect upon both vertebrate and invertebrate animals and is an inhibitor of muscular activity. It might possibly inhibit the giv- ing off of the optic vesicles or prevent their separation in the brain, so that both might come off together as in cyclopia, and it might have caused the eye in Fig. 38 to be arrested when only halfway separated from the brain; the absence of one eye and complete absence of eyes would be perfect inhibition. It is necessary to find a definite point in the strength of the solutions in order to obtain the proper amount of inhibition for many weaker eggs are killed during early stages. The strongest argument against such an hypothesis is the fact that Mg in distilled water solutions fails to cause cyclopia, whereas its anzesthetic or inhibiting powers should be most active in such a solution. Dareste’s (’91) idea that cyclopia is caused by a closed brain or the failure of the anterior vesicle to develop is unsupported, since in Triton with the hollow-brain tube present Spemann finds that the defect occurs. In Fundulus the optic outpush- ings are normally given off while the brain is yet solid, so that according to Dareste all of these fish would be cyclopean in nature. Schwalbe (06). in his Morphologie der Missbildungen des Menchen und der Tiere, considers cyclopia to result from unusual pressure exerted during early stages of development which does not cause the lateral parts to grow together but prevents them from developing at all. This position is somewhat in accord with the hypothesis suggested above. If pressure prevents the grow- ing apart laterally of the anlagen which normally require energy to accomplish their separation, then by anesthetizing a part, one accomplishes practically the same thing as by applying pressure. Artificially Produced Cyclopean Fish 333 ) The part in anesthesia lacks energy to grow out laterally, thus the two eye anlagen remain together in the floor of the brain and come off as one median vesicle either double or single, depending upon the extent of separation possible under the given degree of pres- sure or anzesthesia. Mall (’o8), in his recent memoir on the causes underlying the origin of human monsters, gives an excellent survey and discussion of the evidence furnished by experimental teratology. In the body of the paper is presented a strong case in favor of external influence during development as the chief cause of many mon- strosities. Here we may consider only the discussion of cyclopia. The idea of fusion of the two eye vesicles during their develop- ment is advocated, but the present evidence is against this posi- tion and is in accord with Spemann’s (’04) view of an early defec- tive anlage. Mall also inclines toward the idea of the single brain as being primarily responsible for cyclopia, but it 1s shown by embryos considered here that cyclopia often accompanies perfectly bilateral and bilobed brains, neither does a retarded growth of the frontal process necessarily follow in cases of cyclopia. Experiments uphold the statement “that every egg has in it the power to develop cyclops monsters.’’ ‘The germinal theories of cyclopia are shown by the experiments to be unnecessary as ex- planations of its cause. The possibility of its occurrence through germinal variations, though to my mind extremely slight, 1s not entirely excluded by experiments. The experiments conclusively show the origin of cyclopia through external influences. Much could be said pro and con regarding the significant nature of the cyclopean fish embryos as a specific response to a definite chemical environment. The suggestion is evident, though highly hypothetical, that cyclopia in man and mammals might be due to a similar chemical cause, an excess of Mg salts in either the mother’s blood or the amniotic fluid surrounding the developing embryo. The Magnesium embryo 1s as typical of these Mg solutions as is the now classic lithium larva of the sea urchin produced by Herbst (92, 93) in his Li solutions, or Morgan’s (’04) lithium frog em- bryos produced in a similar way. ‘They all tend to show that dif- 334 Charles R. Stockard ferent chemical conditions may each induce by their actions a specific type of larva from a given variety of egg. SUMMARY 1 The eggs of the fish, Fundulus heteroclitus, give rise to a large percentage of cyclopean embryos when subjected during ee development to solutions of magnesium salts in sea-water. Similar results follow if the eggs are placed in the solutions either before cleavage or when in the two or early four-cell stages, later stages were not tried. ‘This is the first instance of repeatedly causing, by the use of chemical substances, vertebrate monstros- ities such as are known in nature. 2 The peculiar embryos with the median cyclopean eye are able to hatch. Many of them swim about in a perfectly normal manner, darting back and forth to avoid objects placed in their field of vision as readily as do two-eyed individuals. 3. The cyclopean fish is exactly comparable to the monstrous cyclops of man and other mammals. Both have a median eye either double or single in its structure. The nose in the mam- malian cyclops is a single proboscis-like mass above the eye. “The nasal pits in the “ Magnesium embryos” are sometimes united and sometimes separate, but the mouth hangs ventrally as a pro- boscis-like organ strikingly suggesting in form the nose in mam- malian cyclopia. The mouth of Fundulus normally occupies an extremely anterior position but in the cyclopean fish the eye has usurped this place, thus preventing the usual forward growth of the mouth elements and forcing them to remain ventrally as the proboscis-like mass. (See Figs. 25,26, 27.) In cyclopean mammals a similar mechanical explanation accounts for the condition of the nose. The median eye obstructs the path of down-growth which passes normally between the eyes, and forces the nose to form above the eye as a proboscis on the fore- head. 4. A study of more than 275 living cyclops monsters and of many of these in section shows all degrees in the defect. Eyes unusually close together, intimately approximated eyes, the double Artificially Produced Cyclopean Fish | 335 eye in a median position, the single cyclopean eye, an extremely small anterior eye, a deeply buried ill-formed cyclopean eye, and finally an entire absence of the eye. The embryos exhibit these various degrees of the cyclopean defect from the earliest appear- ance of the optic outpushings, and in no case was cyclopia due to a union or fusion of the two eye components after they had originated distinctly. 5 Asecond type of monster designated as Monstrum monoph- thalmicum asymmetricum, the monster with one asymmetrical eye, was also common in the magnesium solutions. ‘These indi- viduals have one perfect eye of the normal pair but the other is either,small, poorly represented or entirely absent. This condi- tion is also present from the first appearance of eye structures and is not due to degeneration or arrest of development. 6 Both types of monsters often form lenses independently of the optic cup stimulus. ‘These self-originating lenses are also capable of perfect self-differentiation, forming lens fibers and appearing as transparent crystalline bodies. Sieh facts oppose the idea that the lens during its origin and development is in a dependent relationship with the optic cup, and show this view not to be of universal application. 7 ‘The experiments conclusively prove that eggs may be in- duced to develop into cyclopean monsters by external influences. These influences do not mechanically injure or destroy certain eye regions as does cutting or pricking. It follows, therefore, that cyclopean monsters appearing in nature are not necessarily due to germinal variations, but are far more likely the result of some unusual external influence during development. 8 The occurrence of the various eye monstrosities shown by embryos which develop in magnesium solutions are all probably due to a common cause and [ suggest the following hypothetically: Magnesium which possesses a decidedly anzsthetic effect on most animals and is inhibitory in its influences on muscular activity may retard through degrees of anzsthesia the optic outpushings in Fundulus embryos and thus account for the total absence of eyes, small eyes, eyes which failed to develop energy necessary for their normal separation and the other unusual conditions which 336 Charles R. Stockard have been considered in detail in the present article. This view, of course, is hypothetical and objections to it are recognized. Cornell University Medical College New York City, October 1, 1908 LITERATURE CITED DareEstE, C. ’91—Recherches sur la production artificielle des monstruosités ou essais de tératogénie expérimentale. pp. 366-383, Paris. Donen, A. ’75—Der Ursprung der Wirbelthiere und das Princip des Functions- wechsels. Leipzig, pp. 1-87. feat C. ’92—Experimentelle Untersuchungen uber den Einfluss der verander- ten chemischen Zusammensetzung des umgebenden Mediums auf die Entwickelung der Thiere. I. Theil. Zeitsch. f. wis- sensch. Zool. iv, pp. 446-518. °93Experimentelle Untersuchungen. II. ‘Theil. Mittheil. aus der Zool. Station zu Neapel, xi, pp. 136-220. *o1—Formative Reize in der thierischen Ontogenese. [Ein Beitrag zum Verstandniss der thierischen Embryonalentwicklung. Leipzig. LeCron, W. L. ’07—Experiments on the Origin and Differentiation of the Lens in Amblystoma. Am. Journ. Anat., vi, pp. 245-258. Lewis, W. H. ’04—Experimental Studies on the Development of the Eye in Am- phibia. I. On the Origin of the Lens. Rana palustris. Am. Journ. Anat., 111, pp. 505-536. Matt, F. P. ’08—A Study of the Causes Underlying the Origin of Human Monsters. Journ. Morph., xix, pp. 1-361. Menct, E. ’03—Ein Fall von beiderseitiger Augenlinsenausbildung wahrend der Abwesenheit von Augenblasen. Arch. f. Entw’mech., xvi, pp. 328-339. Morcan, T. H. ’03—The Relation Between Normal and Abnormal Development of the Embryo of the Frog, as Determined by the Effects of Lithium Chlorid in Solution. Arch. f. Entw’mech., xvi, pp. O01=7 12: ’*06— Experiments with Frog’s Eggs. Biol. Bull., xi, pp. 71-92. ScHaPEr, A. ’04—Ueber einige Falle atypischer Lensenentwickelung unter abnor- men Bedingungen. Anat. Anz., xxiv, pp. 305-326. ScHwaLsE, E. ’06—Die morphologie der Missbildungen des Menchen und der Tiere. I. Teil. Allgemeine Missbildungslehre. Jena, pp. I-230. Artificially Produced Cyclopean Fish 337 SPEMANN, H. ’o1—Ueber Correlationen in der Entwickelung des Auges. Ver- handl. der Anat. Gesellsch., Igol. ‘04—Ueber experimentell erzeugte Doppelbildungen mit cyclopischem Defekt. Zool. Jahrb. Suppl. vii, Pp. 429-470. ’05—Ueber Linsenbildung nach experimenteller Entfernung der pri- maren Linsenbildungzellen. Ausfiihrlich: Zool. Ang., XXvill. ’o7—Neue Tatsachen zum Linsenproblem. Zool. Anz., XXXxi, pp. 379- 386. STocKarD, C. R. ’06—The Development of Fundulus heteroclitus in Solutions of Lithium Chlorid, with Appendix on its Development in Fresh Water. Journ. Exp. Zodl., iii, pp. 99-120. ’o7a—The Artificial Production of a Single Median Cyclopean eye in the Fish Embryo by Means of Sea-water Solutions of Magnesium Chlorid. Arch. f. Entw’mech, xxiii, pp. 249-258. ’o7b—The Influence of External Factors, Chemical and Physical, on the Development of Fundulus_heteroclitus. Journ. Exp. Zodl., ly, pp. 165-201. ’o7c—A Peculiar Legless Sheep. Biol. Bull., xiii, pp. 288-290. ‘o7d—The Embryonic History of the Lens in Bdellostema Stouti in Relation to Recent Experiments. Am. Journ. Anat., vi, pp. 511-515. 08 —The Question of Cyclopia. Science, n.s., xxviii, Pp: 455-456. Wiper, H. H. ’o8—The Morphology of Cosmobia. Am. Journ. Anat., viii, DP tS5 55440: EXPLANATION OF PLATE I. Fig. A Dorsal view of a cyclopean embryo in almost natural colors. The large antero-ventral eye shows a slight furrow indicating its double nature. Fig. B The same embryo when the egg is rolled back towards the top of the page. A somewhat ventral view showing the single pupil and lens, the double condition of the eye is only indicated from above. ARTIFICIALLY PRODUCED CYCLOPEAN FISH PEATES Cuartes R. Stocxarp RicHeoul, del] THE JourRNAL or EXPERIMENTAL ZOOLOGY, VOL. Vi, No. 2 STUDIES ON THE PHYSIOLOGY OF REPRODUCTION IN THE DOMESTIC FOWL I. REGULATION IN THE MORPHOGENETIC ACTIVITY OF THE OVIDUCT?™ BY RAYMOND PEARL In Figure 32, Plate II, of Miss Stevens’s paper on ‘Further Studies on the Chromosomes of the Coleoptera” (vol. vi, no. 1), one chromosome has been omitted; the figure should appear thus: es $ é os ox 32 eisewnere are trying to answer. Ihe zeal tor inquiry in these directions is greatly stimulated by the obvious fact that at some time or other during the history of poultry under domestication there has been.a very great increase in egg production over what obtains in the wild representatives of the genus Gallus. If the thing can be done once, why not again? : 1 Papers from the Biological Laboratory of the Maine Experiment Station, No. 7. Tue JouRNAL or EXPERIMENTAL ZOOLOGY, VOL. VI, NO. 3. STUDIES ON THE PHYSIOLOGY OF REPRODUCTION IN THE DOMESTIC FOWL I. REGULATION IN THE MORPHOGENETIC ACTIVITY OF THE OVIDUCT' BY RAYMOND PEARL INTRODUCTION This paper forms the first in a series in course of preparation in this laboratory, all dealing with various phases of one broad, general problem. It is desirable that at the beginning of such a series a statement should be made outlining the problem under investigation and, in a general way, the standpoint from which it is to be attacked. It is the purpose of this introduction to give: such a statement. | | When the work of this laboratory was organized one general line of investigation which suggested itself as of first-class impor- tance, both from the theoretical and practical standpoint, was the study of egg production in the domestic fowl. A high average yield of large eggs uniform in size and color is a matter of enor- mous importance to the poultry industry. How can it be obtained ? Can high egg producing capacity be bred into a strain? Can feeding produce it? ‘These are the questions which practical poultrymen in the experiment stations, agricultural colleges, and elsewhere are trying to answer. ‘The zeal for inquiry in these directions is greatly stimulated by the obvious fact that at some time or other during the history of poultry under domestication there has been:a very great increase in egg production over what obtains in the wild representatives of the genus Gallus. If the thing can be done once, why not again? 1 Papers from the Biological Laboratory of the Maine Experiment Station, No. 7. ; Tue JourNAL or ExPERIMENTAL ZOOLOGY, VOL. VI, NO. 3. 340 Raymond Pearl It takes but brief consideration of these economic points to con- vince one that behind them lies a very broad and complex biolog- ical problem, on which light must be obtained before there can be any hope of solving the practical questions. This is the problem of the physiology of reproduction in the hen. Egg production is a definite, if complex, physiological process. In the production and laying of an egg a long series of events are involved; a num- ber of different organs of the body play a part. Before we can hope to control egg production with any precision or certainty it is necessary to learn in detail what is the normal course of events in the production of an egg; how and in what ways each of these events may be modified or influenced by external circumstances; and to what extent each of them is an inherited matter. The physiology of the organs concerned in egg production must be worked out in detail. In order that a comprehensive idea may be gained of the scope of this problem, let us examine the following skeleton outline of the factors and processes immediately concerned in egg production and the points which must be investigated in attempting to get light on, these processes. I Physiology of egg production within the individual. A Processes occurring in or relating to the ovary. 1 The development of the egg and its yolk up to the time of ovulation. Resorption of yolk. 2 Ovulation proper. The rupture of the follicle. 3 Fecundity. B_ Processes occurring in the oviduct. 1 Movement of egg to the outside. 2 Formation of albumen. 3 Formation of the several egg membranes. 4 Formation and determination of the shape and color of the shell of the egg. C Intra-individual variation and correlations in regard to the points enum- erated under A and B. Homotyposis. D_ Behavior in its relation to egg production and reproduction in general. 1 Mating instincts and habits. 2 Brooding instincts and habits. Phystology of Reproduction in Domestic Fowl 341 Il Physiology of egg production within the race.” A Variation in egg production. 1 Intra-racial | in regard to each of the points enumerated under I 2 Inter-racial } above. 3 Mutation. 4 Seasonal distribution of egg production. B_ Inheritance of egg-producing ability. Considered with reference to each of the points enumerated under I above. 1 In pure-bred lines. 2 Under hybridization. C Evolution of egg-producing ability. 1 Influence of selection. 2 Egg production in the wild progenitors of domestic poultry. 3. Fixation of egg producing ability as a racial character. IIL The influence of environmental factors (in the broadest sense) on the processes enumerated above. Nutrition. Housing. Meteorological factors. Drugs. 5 Other environmental agents. IV The relation of internal factors to, and their influence upon the processes enum- erated under I and II. FV Pathological and teratological cases relating to egg production. FW DN & This outline, while not as extensive or complete as it might be made, gives a fairly comprehensive view of the general scope of the problem which forms the subject of the present investigation. Each topic in the list suggests, of course, a whole series of problems, but even to enumerate all these would take far more space than is available here. All that is desired at present is that the broad outlines of the general problem on which we are working shall be clear to the reader. On account of the extent of the subject it is necessary to publish the results of the work upon it in a series of separate papers. ‘The skeleton outline given above will serve as a means of coordinating the separate papers, and making clear the ? The general standpoint which regards variation, heredity and other factors of evolution as physio- logical problems has been well set forth by Jensen (’07) and Jennings (’o8). 342 Raymond Pearl relation of each to the general problem. Statements and discus- sions of the subsidiary problems connected with each of the topics in the outline will be given in connection with the detailed treat- ment of those topics. In attacking this problem we are bound to no exclusive method of investigation. The observational, experimental and statistical methods will be used as they appear to be demanded by the exi- gencies of the case. The writer’s standpoint is that the problem is one in general physiology, involving questions having to do both with individual and with racial physiology. On a certain number of the points covered in the outline work has been com- pleted and will be published as soon as possible. On _ other phases of the problem the work is well advanced though not ready for publication. The present paper deals with a definite and circumscribed topic falling under 1B4 and IC of the outline. It is well known by poultrymen that the first eggs laid by a pullet often differ from the normal eggs and from eggs laid later by the same bird in regard to both size and shape. This implies a process of regulation in the continued activity of the oviduct in shaping successively laid eggs. [he present paper deals with the detailed analysis of a clear-cut and unusually pronounced case of such regulatory ac- tivity of the oviduct. THE MORPHOGENETIC ACTIVITY OF THE OVIDUCT A bird’s egg is an object of very characteristic shape. While the form of the egg varies in different species, and also within the single species, all such variation is comprised between relatively narrow limits. ‘The conformation to type in the case of eggs from birds of a single species is in most cases quite close. It is usually still closer in a series of eggs laid by the same bird, as in the case of any of the domesticated fowls. The production of a series of definitely and characteristically formed bodies all conforming closely to a type by an organism implies that the organs concerned in this production have a morphogenetic function along with others. It is a well established fact that the shape of the egg of any Physiology of Reproduction in Domestic Fowl 343 bird is determined in the oviduct. The yolk as it leaves the ovary is spherical in form, except as it may be deformed through the action of gravity. As it passes down the oviduct it is sur- rounded by albumen, and finally at the lower end by the so-called “shell” membrane, or membrana testacea. It is probable that the egg is not given anything approaching its characteristic form until after the formation of this membrane. As to exactly where and how the egg is given its form by the oviduct there is some difference of opinion and definite evidence is lacking. Szielasko (05), who has paid particular attention to this problem, after reviewing the older literature of the subject expresses the opin- ion that the egg is given its definite form in the uterus. He says on this point (p. 289): ‘Das Geprage, welches die Eiform der verschiedenen Species aufweist, kann in der Tat, wie Grassner vermutet, nur von dem Uterus verliehen sein; denn solange das E1 im Oviduct verweilt, ist seine Form variabel, da es jeder Um- hillung entbehrt. Die erste Hille, membrana testacea ge- nannt, wird dem Ei erst im untersten Abschnitt des Oviductes un- mittelbar vor der Miindung desselben in den Uterus—im sogen- annten Isthmus—-umgelegt. Auch durch diese Membran wird dem Ei noch keine bestimmte Gestalt gegeben. Diese resultiert erst aus der Umlagerung der harten Kalkschale, welche im Uterus geschieht. Hier ist also das formgebende Organ, hier muss demnach die Untersuchung angreifen.”’ This was also the early view of the matter. Wahlgren writing in 1871 states as a matter of common opinion that: “Hier (.e., in the uterus) erhalt das Ei seine Schale und seine Form.”’ The most recent worker in the subject, Thompson (’08), while making the statement (p. 112), “The egg, just prior to the form- ation of the shell, is, as we have seen, a fluid body, tending to a spherical shape and enclosed with a membrane,” which would cer- tainly seem to imply that he supposed the definite shape to be given to the egg in the uterus, proceeds to develop a theory of the method by which the egg gets its form which seems, as he states it, to in- volve the activity of nearly the whole oviduct in the process. Direct observation shows that after the membrana : x ae SP. AMT. SP. AMT. all < Py SP. AMT. SP. AMT. 5 ¢|/S2).63 | ORIGINAL DIA.. FINAL DIA. | A J || O & | ORIGINAL DIA.| FINAL DIA. Pa lewew ih thes | Gre la 2G pire hei [2/2 cal 4 | fecha 63 53 0-3 -024 028 65 53 T-Onigu eee OLe -O14 45 I=2 | .023 | .033 TS 79> (23) — fy --933 -036 | | 77 65 3-4 046 054 56 | 5-5 -065 | 089 80 | 70 | 4-5 | .056 064 | | 90° | So | 4-4 ||, | ong .050 | 7x || 5-6] .o61 | -078 go | 82 4-5 | .050 | .055 78 || 6-6| .067 .077 95 | 91 4-5 O47 | .050 85 5-6 | .058 | .065 105 | 95 | 2-4 | 029 .032 | 90 || 3-4 | -033 | 039 10s | 87 | 4-4 | .038 | .046 | | | 110 | 100 | 1-3 | .018 | .020 | | | EUS eLCO ma Ea 4. -030 | .035 | 98 | 3-4] -030 | 036 120 | 113* | 1-0.5| .006 | 007 | 108 || 1-3 | .017 | .019 Av. 91.5 | 81.5 3 .033E .003 | .038- .003 | 78.9 | 4.1 | .044-b0.005 | .055++0.006 J } } Paes | *Hole in disk due to parasites. A comparison of the regeneration in the several groups of medu- sz is facilitated by table VIII, which is a tabulation of the aver- ages contained in tables III to VII. The first column shows the number of individuals composing each of the groups, the second column indicates the number of mouth-arms amputated from each 452 Charles R. Stockard * specimen, column three the original diameters in millimeters of the medusz disks, column four of the upper half of the table the diameters after 17 days and the lower half contains the diameters after 24 days. The numbers in column five are obtained by sub- .tracting those in column four from those in three and represent the average decrease in diameter of the disks. Column six indi- TABLE V Regeneration of the oral-arms in Cassiopea, when four arms are removed | | | z eat fe ics g [38 |< 2 ee eee ae | Z29|/20|/, 2) sp. amr. | sp.amr. | 2 5), * ©) sp. amr. SP. AMT. = ‘i | : | Ss > < . | ° = 2 | ORIGINAL | FINAL | < al . a | ORIGINAL FINAL ags| 4 2\8 ae | ® Ble 2.8] wu uw Heh cerace| DIA. | DIA. Nie te BS a DIA. | DIA. 2 ce | 8 a | 2 a) 2 a |; a = } a ee | 65 | 54 | 2-2-2-3 | .035 | 042 46 4-4-4-4 .062 | .087 70 | 59 | 4-4-4-4 | .057 .068 | 73 | 67 | 4-4-4-3 | -051 _ +056 57 | 2-3-3-4 | 041 053 5 WN OP EN Oe | .065 58 4-5-5-6 | 067 | .086 US |e GS | ere bry |i sclek fe 043 | 80.) 71 =| 2=2=2-1 |) 2022 -025 | 61 De | 038 | .049 80 | Gc 2=2—2-0 O28 -035 | | 93 81 | 2-2-3-3 | .027 O31 70 ASSES | O51 068 98 go | 4-4-5-5 046 -050 79 | 3-4-5-5 | 043 054 98 | 76 | 1-1-2-2 | .or5 .020 | 105 || 93 4-5-5-5 | .045 -O51 | 89 5-6-6-7 | .057 .067 115 | 1o5F | 3-4-4-4 | 033 | 036 98 | 3-3-5-6 | .037 043 115 | 92 | 4-4-5-5 | -039 -049 120 103* | 2-3-4-4 | .027 5Og2 | 98 hs || sCuS | .043 Av. 90.1 a7e7 || 363 -037 .002 .043-+ .003 || 72.9 4:3 .048-+0.003 | .061-0.004 *Hole in disk center. cates the actual percentage of decrease in diameter in terms of the original size. ‘The seventh column gives the average lengths of new buds in each group. Column eight shows the specific amounts of regenerated tissue calculated by dividing the numbers of col- umn seven, the length of new buds, by the numbers in column three, the original disk diameters. Column nine gives the spe- 453 Studies of Tissue Growth *payourig pur iesuly Ur png eniut oy} Aqyensn ‘apis [e1UeA uO pound sauod [eseq adie] sv UOTRIOUASaI UISaq suIYL. *pajeqndure or0M Ay pur AT[euIIOU suTIe UoAas ATUO prey UauTIIadcy +o0*0-Folo: 6So0° gSo° glo: Llo: 8go° 980° ofo: LLo: glo “vid ‘IVNII “LWNY “dS £00'0 = 7zSo° Lt¥o: tto°* tg0° oSo: Sgo° 690° IZO° $So: gSo° ‘vid ‘IVNIOINO "INV “dS £=9-9-9-S-¥ L-S-S-S-b-¥ ETT TELN 4-9-9-S—S-£ L-L-L-9-S-S 9-9-9-5-$-5 t-t-Z-I-I-I b—b-b-4-4-£ +—-b_b_-b_£_-£ €1 anol ‘WN NI sang -WUV JO HIONGT £1 anof “vId ASIa 700° =-+to° Sfo° tho: S£o- tbo: Sfo- Ifo" 190° | 6to° | 990° SSo- tI0° | gSo- tto° oSo° “vid ‘IVNIJ “LNV “dS 700° =--L£o° Ifo" Sfo° Lzo° gto: ofo" ¥z0° 1So° tbo: $So- 1So° 110° gro: Lto: Ito* *vId IVN -IOTHO “LWW “ds ue +-b-b-+_-£-£ Ga A Aes Gossett +—-b—b-b_-+_+ 4 £-£-£-£_£-£ Ret e= tome g-S-S-S—b_- era cacece ¢-S—b—b_-b_b +—-y-v—b_-b_-£ 1-1—1-1--s20=980 +——+-£-£-z y-v—t—t-0—6 Sat ae 9 anal ‘WW NI Sand-Nuv 40 HLONAT | bbl 9 anof ‘NW NI “vid ASI ot AVN | | "WN NI | TA YTaVL parouas asD SuLsD x1S Udy ‘vad o1ssv-) ut sutsv-j040 ay1 fo uolwsauanay Charles R. Stockard 454 junosor AvW UONPIauasaI [RUOTIIPpe STYT, *suIIe-[kIO VY] JO UONRI9uader pides ApreTNIed ay} 104 ‘apra ‘uur L Aq Suoy ‘wu $z st anssty Mau ay} ‘Japs0qg YsIp ay} Jo jred & Suyesrauadar osye st uaunads sty Tx $00'0=7zgo° go" 190° £go" 1Lo° 6Lo° Szi° olo* for" £g0° *vId IVNIA “INV ‘dS 4o0'0 + 850° 6to° gto: 1S0° ‘vid IVN -ID1HO “INV “ds oh L-L-g-9-9-S—S-S L-L-9-9-S-S—>-» Pi pi5 5-0-8 g-9-5—S—S-S—b_-b 9-9-9-S—S—S—S_S o1-g-g-8-g-8-L-L as 2 tere $-S-S_$_S_+_t_ Pe as A Or 2 €1 anol ‘WN NI | '€1 anat 9°99 ob ‘WW NI \SaNad-NWuv AO HLONAT*VId ASIC) | foo" = 6S0° | gto: 60° ogo" Lyo: | gto: ogo" 190° Lgo° 6g0° | oso" | ESo- Zgo° | Zgo" | tro: | “vId £00° gto’ 17 6f0° g-S—S—S—b-b_-4_ £fo° GN tests gto" $+9-9-9-S—S—b_b_£ gto: Sh V6 8 ofo' boy -7-2—7 6to° $-S—S-S_S_+_+_+ oSo: fo 5 =v $So° 6 $65 SF. glo: L-S+g-S-9-S*9-9-9-S-S 1t0° <———p_¢-€—7_1 zo" bby Ff zz 990° §-$—$—S_-§_S 4 690° $-S-S-S—b_-+_-b_- ££o° tte —=—t—1 =I ‘vId “IVN | g anal JIVNIA “LNW “dS-IDIYO “LNW ‘dS “WW NI SGNad-NWav JO HLONAT S-zE z°6g “AY ool ozI 96 S11 £g for 0g oor Pye oy LL S6 gl $6 Segor. ie eg #89 | 8 S9 | og pts ah ie gs zl SS So 6b | Sg g ann{ | ot AVN ‘NW NI | ‘WW NI ‘VIG MSId| ‘yId 4XSIa panouas 340 SULID 1y31a uaym ‘padot IIA W1ave SSD ul susv-jo410 ay1 fo uolwsauaday Studies of Tissue Growth 455 cific amounts of regeneration when calculated by dividing the lengths of new buds by the final disk diameters. The different results shown by columns eight and nine are obvious and empha- size the error of using the final diameters in the calculations instead of the original diameters. The meduse most injured, although in this experiment they are actually regenerating at the fastest rate, are also most rapidly decreasing in size and so give a dispro- portionate divisor in the calculations if their final size is used. TABLE VIII Tabulated summary of Tables III to VII showing the rates of regeneration and the decrease in size when Cassiopea xamachana regenerates different numbers of oral-arms. The upper half of table shows the conditions after seventeen days, the lower half after twenty-four days [DIAMETERS | 2 ACTUAL LENGTH = g| ORIGINAL | AFTER 17 PERCENT- SP. AMT. OF | SP, AMT. OF No. or |% 8 LOSS IN | OF NEW < © |DIAMETERS.| AND 24 AGE Loss| REGENERATION |REGENERATION SPECIMENS.|5 = SIZE, IN TISSUE 2 | IN MM. DAYS, IN IN SIZE, ORIGINAL FINAL 3% MM. IN MM. z MM. DIAMETERS, DIAMETERS. 12 I 88.9 82.3 6.6 7-4 3-4 ©.038-0.004 | 0.041+0.004 13 2 gI.5 81.5 10. 10.9 3.0 -0330.003 | .037+0.003 14 4 SORE Tifa ateed: 13.8 38 .0370.002 | -042+ 0.003 14 6 89.4 74-4 | 15. 16.8 3-3 -037+0.002 | -044-+ 0.002 14 8 89.2 eas Ne ORG] 18.7 4.1 |. .046+0.003 | .057+-0.003 —| ————————EEE 9 I 91.3 Tess | aks 13-7 | 4-6 | .o50+0.004 | .058+0.004 8 2 94.6 78.9 iay/ 16.6 4.1 -0430.005 | .052+-0.006 9 4 Ol Om 7269 18.7 20.4 4.3 -047-++0.003 | .059+0.004 9 6 91.4 C8e2 ih ezai.2 25-4 4.8 -052-+0.003 | .070+0.004 9 8 92.6 | 66.6 26 pie Ihe Gos! -057+0.004 | .o80-0.005 Table VIII is divided into two parts the upper part shows the results of the entire experiment 17 days after the operation. The lower part gives the results shown by the number of individuals indicated 24 days after the operation. The facts of chief importance contained in the tables are: the decrease in size of the medusz during the experiments in direct relation to the number of arms being regenerated (Fig. 8), and the absence of any significant relation between the number of regen- 456 Charles R. Stockard erating arms and their rate of growth, with the possible exception of those growing eight arms. Plate I strikingly illustrates this decrease in size in six individuals originally of equal size, and on closer examination also shows the difference in lengths of the re- generated buds in the several specimens shown. In addition to these facts it should also be noted that the spe- cific amounts and rates of regeneration when calculated on the basis of the original disk diameters gradually increase from those medusz growing two new arms to those regenerating eight. When the specific amounts of regeneration are calculated on the basis of the final disk diameters a similar but much more exaggerated | Arm Fig. 8 Curves aa ae the decrease in size of meduse disks when regenerating different num- bers of oral-arms; 1, 2, 4, 6 and 8. The upper curve indicates the condition after 17 days, the lower curve after 24 days. The numbers on the vertical line indicate the actual loss in diameter in millimeters. increase is noted in those medusz growing the larger number of arms. (This exaggerated increase is only apparent and is due to the fact that those meduse regenerating many arms have de- creased in size to a greater extent than those regenerating fewer arms.) ‘The above increases in specific amounts of regeneration are, however, insignificant as is readily seen by comparing the prob- able errors. The probable error of the difference between any two groups may be represented by the formula | Delremernn Vel ae) Dy 1.e., the probable error of the difference in specific amounts of regeneration between any two groups is + or — the square root Studies of Tissue Growth 457 of the sum of the squares of the errors of these two specific amounts as given in the table. One finds on making the calculations that the differences are scarcely equal to or even less than the probable errors when comparing the groups regenerating one, two, four or six arms. If any of these four groups be compared with those individuals regenerating eight arms the latter group seems to show a distinct advantage in its rate of growth which is almost three times greater than the expected errors of the differences in the several cases. It may be stated, therefore, that the data for Cassiopea shows no relationship between the degree of injury and the rate of regeneration with the possible exception of those indi- viduals regenerating all eight of their oral arms. ‘The latter class seems to grow new arms at a more rapid rate than do specimens injured in any other manner. In consideration of the limited number of individuals and the wide variability within the group even this latter difference may not be generally found. Zeleny (’07) concluded from a study of a small number of these medusz that the rate of regeneration was fastest in those specimens having lost six of their oral arms and slower in individuals that had lost more arms, as well as in those growing fewer. It is obvious that my more extended series of experiments on Cassiopea fails to show any such advantage for those regenerating six arms over those growing eight arms. Since Zeleny’s series is only one-third as great as mine the differences in rates shown in his table are probably, like mine, not significant. The lower half of Tabie VIII shows that after 24 days the dif- ferences in ratés between differently injured medusz are still not significant. We may now proceed to a consideration of the regeneration rates of the arms in brittle-stars injured to different extents. A large number of these animals were lost in unsuccessful aquarium experiments. Others were successfully kept in floating “live- cars’? where they flourished and increased in size. The “live- car” experiments were started with 150 individuals, 135 of which were available for final measurements. Zeleny (’03) experi- mented on the brittle-star Ophioglypha and in fact on this form first obtained the data from which he suggested the principle that 458 Charles R. Stockard appendages regenerate more rapidly when three or four are re- moved than when one or two are cut away. Zeleny’s final meas- urements were made on only 36 animals and these had regenerated very short arm buds. After 46 days only a few of his individuals had new arms even 5 mm. in length. These facts tend of course to increase the probability of error. The experiments now to be recorded continued for 49 days at the end of which time 135 specimens were measured. ‘The aver- age length of the new arms ranged from 29 to 46 mm. ‘The error in measurements is greatly decreased in arms of such length. (An error of I mm. here would equal an error of only 0.1 mm. in Zeleny’s measurements. A 1 mm. error is unlikely to occur but errors of 0.1 mm. in measuring brittle-star arms are difficult to avoid.) Fortunately two species of ophiurans were employed in the experiments since it happens that they differ slightly in their responses to different degrees of injury. “The case of Ophiocoma riisei, a large black spiny form with reddish tube feet, may be con- sidered first. Perfect animals were selected and grouped into five lots the individuals of the lots being of the same average size. All arms were amputated I cm. out from the disk with sharp scissors. ‘The first lot had one arm cut from each brittle-star, in the second two arms were cut off, the third lost three arms, the fourth four and all five of the arms were cut from each individual in the fifth lot. After 49 days the specimens were expanded and killed in fresh water and chloretone and preserved in alcohol. The new arms plus the old 1 cm. stump were not so long as the original arms nor were they equal to the old ones in thickness. A correct comparison of all the groups is made by considering the averages of actual new arm lengths shown in Table IX. Origi- nally the average diameters of all the lots were practically the same, therefore, the arm lengths are to one another as the quotients (specific amount of regeneration) would be if they were divided by the originally equal diameters. I have calculated the specific amounts on the basis of final diameters merely to show the erro- neous impression obtained by using such a method. Studies of Tissue Growth 459 The upper half of Table IX summarizes the data from O. riisei and presents the following points of interest. First, individuals of Ophiocoma ritset when kept under identical conditions increase in body size the slower the more arms the individual is regenerating. This fact is not likely due to the incapacity of the more injured specimens to secure food since the compartments in which all were confined are small and the individuals seemed equally able to traverse this limited feeding ground. A probable explanation is that the new regenerating tissue possesses an excessive capacity for the assimilation of nourishment and consequently those speci- ‘mens regenerating more new arms were less able to increase in body size than those regenerating fewer. Table IX shows secondly, that the rate of regeneration of each arm bears no relation to the number of regenerating arms, or in other words, the extent of injury. Column five gives the specific amounts of regeneration for each arm when calculated by dividing the average arm lengths by the final disk diameters. The last two figures in the column indi- cate the error of such calculation. Calculating the specific amounts of regeneration on the basis of the original average diam- eters which were practically equal in all the groups we obtain a series of numbers bearing the same relations to one another as are shown by the numbers in column four. Ophiocoma echinata, a spiny, grayish, mottled brittle-star, was experimented upon in exactly the same fashion as Ophiocoma riisei. Its response to different degrees of injury was much more pronounced than that of the species riisei. Again five groups of individuals of the same average size were selected and operated upon so as to remove different numbers of arms 1 cm. from their bases at the disk. Referring again to Table [X the lower half represents a tabulated summary of the data from Ophiocoma echinata. ‘Two facts of importance are here also to be recognized. First, the fourth column giving the average arm lengths for each group shows that the rate of regeneration decreases as the extent of injury increases. Each new arm grows fastest from those individuals regenerating a single arm and successively slower in the groups growing two, 460 Charles R. Stockard three, four and five new arms. ‘These differences in rates are significant when compared with their probable errors. Secondly, it is to be noted that the average of final disk diameters in Ophiocoma echinata is practically equal in all of the groups. This fact might be reconciled with the smaller increase in disk diameters shown in the more injured groups of Ophiocoma riisei, TABLE IX Tabulated summary showing the rates of regeneration in brittle-stars when regenerating different numbers of arms O phiocoma rtiset AVERAGE NO. OF FINAL DISK) NO. OF LENGTH OF SP, AMT. OR RATE SPECI- DIAMETER | ARMS RE-} ARM-BUDS ar REGENERATION. MENS IN MM. MOVED. AFTER 49 DAYS,IN MM. 14 18.3 I 42.9 2.340.104 13 18.1 2 45.8 2.530.060 15 ce 3 41.8 2.420.061 14 Wyfon 4 41.5 2.400.043 15 16.5 5 39-6 2.40+0.049 O phiocoma echinata 13). |) 16.6 Ue BaGtg 2.220.052 TG? shea 2 ly any 2.07+0.089 14 16. Be 33-5 2.09+0.060 14 16.5 4 ahieel 1.90+0.034 13 16. 5 28.9 1.810.040 where the rate of regenerative growth is the same in all groups, by supposing that the decrease in growth rates of arms in more injured individuals of Ophiocoma echinata is sufficient to allow the disk to increase in size as readily as those disks which are growing only a few arms but at a more rapid rate of regeneration. In Ophiocoma echinata arms grow 30 per cent faster in those speci- mens regenerating only one arm than in those growing five. It is admitted, however, that the actual amount of regenerating tissue Studies of Tissue Growth 461 is greater in the individuals growing five arms than in those grow- ing fewer. The above reasoning depends largely upon the rate of regeneration itself as a factor acting upon the old body tissue to inhibit its growth or to cause it to decrease in size. I do not believe that this is entirely important and offer the above suggestion only as a possibility. ‘The facts furnished by the medusa, Cas- siopea, indicate that the decrease in body size is in a greater pro- portion than the increase in growth rates of new arms in speci- mens extensively injured as compared with those less injured. We have now considered the relation between the degree of in- jury and the rate of regeneration in three different species of ani- mals. The three species clearly show that the extent of injury . fails to exert an influence in any one definite direction over the rate of regenerationinallanimals. Former experiments which have seemed to indicate that the rate of regeneration is increased in animals injured to greater degrees have either been performed on crustaceans where growth is not continuous and where the influence of regeneration on the molting cycle introduces a com- plication, as Emmel’s work so clearly demonstrates, or else have been conducted with too small a series of animals to justify general conclusions. Scott’s study (’07) of the rate of fin regeneration in more than 100 individuals of Fundulus heteroclitus shows by careful calculations that the degree of injury exerts no influence either to increase or decrease the rate of regeneration in this fish. These experiments on the fish, along with those of Zeleny on crus- tacea, an ophiuran and a medusa, Emmel’s study of larval lob- sters and my experiments on Cassiopea and two Ophiurans would seem to justify the following conclusion. By varying the extent of injury in several animal species there is no definite influence exerted 1n any one direction on the rate of regenerative growth. Morgan (’06, p. 460) draws a resemblance between the differ- ences in rate of regeneration at different levels on an appendage or body and Zeleny’s idea regarding the relation between the regen- eration rates of the new parts and the number of parts removed. “Tf the distal end of the tail is removed it regenerates more slowly than when more of the tail is cut off. Thus the more the material removed the greater the rate of regeneration of the new part. 462 Charles R. Stockard Stated in this form the two results appear to be identical.” This resemblance is quite true for the evidence furnished by Zeleny but the more recent work fails to accord. In the larval lobster and some ophiurans it is not true that ‘“‘the more the material removed the greater the rate of regeneration of the new part.” The influences exerted at different levels over the rate of regen- eration cannot be identified with the influences due to different degrees of injury. It must be recalled that Zeleny (’03 and ’o5) claims that each appendage regenerates at a more rapid rate when several are removed than when only one is amputated. This has been shown not to be true for all animals but, on the other hand, the total amount of tissue regenerated from several arm stumps is greater than the amount from one even though the single arm may be regenerating at a more rapid rate. The more mate- rial removed up to a certain limit the greater will be the mass of newly regenerated tissue in a given time, irrespective of whether the greater amount of material is removed by cutting at a deeper level or by amputating a larger number of appendages. The statement in this form is supported by the present evidence. VII THE RELATION BETWEEN THE RATE AND AMOUNT OF REGEN- ERATION AND THE PHYSICAL CONDITION OF THE ANIMAL BODY In the foregoing pages it has been repeatedly mentioned that individuals regenerating several appendages are at the same time either decreasing in actual body size or are increasing in size slower than other individuals which are replacing fewer lost parts. It seems expedient now to consider such cases collectively in order to determine whether there is any actual tendency on the part of regenerating tissue to appropriate nutriment at the expense of the general body vigor. Emmel (’06) showed that the process of regeneration retarded the growth of young lobsters sometimes as much as 24 per cent. He demonstrated that the retardation was due to the process of regeneration and not to mutilation or other causes: Since “the average length of the molting period for those lobsters in which the mutilations were not pies ded by the regeneration of the Studies of Tissue Growth 4.63 limbs, was not only less than the length of the molting period for the regenerating specimens, but also in a large proportion of cases was even shorter than the molting period for the normal lobster, ”’ Emmel finally concludes, that the process of regeneration by retard- ing both the frequency of molting and the increase in size retards the growth of lobsters. After an examination of the relation between the increase in size in normal and in regenerating salamanders Morgan states (06): “That a newly regenerating part has the power to take from the blood the materials that it needs for growth, even when the amount present in the blood has fallen so low that the rest of the tissues cannot maintain themselves, but break down to supply the blood with a certain amount of nutriment. If this idea expresses approximately the relation that exists, it follows that while the new part requires a certain amount of food in order to continue growing, it can take advantage of a condition that the older or differentiated tissues cannot make use of; in fact, when the latter slowly lose ground. . . . Since in regeneration the new part is formed directly out of the old tissues we may assume that this property (excessive capacity of assimilation) of young parts is something connected with their lack of differentia- tion, which is lost when differentiation takes place, and is regained again when the differentiation is lost.”’ Morgan’s ideas are most suggestive when considered from the standpoint of the conditions found in malignant growths. In Ewing’s (’08) survey of the latter subject he calls attention to the fact that the energies of cells are normally divided between pro- liferation and specialized function, between work and growth, both being limited by blood supply. Examples of cells set apart for growth are the germ center cells of lymph nodes, the cells at the bases of intestinal villi and the basal cells of the epidermis. Ewing states that “it is just from these cells, subject to marked variations of the demands for growth, that tumors arise. It is clear that deficient demands for function on the part of derived cells would leave their energies unconsumed and further available for growth. These conditions surround the inception of cancer in the atrophying breast.” 4604 Charles R. Stockard Adami (’o1) has also presented the general importance of this point of view, designating the tumor process as the cumulative “habit of growth replacing the habit of work.” Considering both these statements and Morgan’s, one may express the case as follows. Undifferentiated tissue is that which has not begun to function and so employs all of its energies in growth. When a limb is removed the tissue at its base can no longer function so that it gives up its differentiation and begins to growagain. I agree with Morgan when he states that “‘ because a tissue had become differentiated it has not lost the potentiality of becoming young again, provided it gives up its differentiation.” However, neither Morgan nor Emmel have drawn any similarity between the action of regenerating tissues and malignant growths. The one difference I wish to point out, however, between regen- erative growths and tumor growths is that the former sooner or later stop on account of differentiation and function having begun while the latter are not inhibited by such forces and so continue to grow indefinitely. I believe that additional evidence in support of the above views is furnished by the medusa and perhaps also by the brittle-star. The medusz in all cases were unfed and decreased steadily in size during the experiment. ‘The decrease in size was without exception greater in those specimens which were regenerating a larger number of arms when compared with others regenerating fewer. In passing through the series each group was smaller than the preceding groups regenerating fewer oral-arms but always larger than the following groups which were regenerating a greater number of arms. ‘The series is illustrated in plate I, a photo- graph of six medusz each regenerating successively greater num- bers of parts. ‘The specimens were all equal in size at the begin- ning of the experiment but are now successively smaller as the number of regenerating parts is increased. It may be argued that the greater removal of tissue lowers the possible food supply to be drawn on by other parts. It is equally true, however, that when more parts are removed fewer remain to require nourishment from any source. The oral arms which are supplied with nematocysts normally Studies of Tissue Growth 465 move to surround the prey and are not entirely passive but prob- ably in comparison with the disk tissue use a proportionate share of the available material in starved individuals. It would seem in consideration of the great loss in size that the new regenerating arms require an unusually large amount of nutriment. ‘The medusz growing six new arms, for example, show a much greater decrease in the size of their disks over those growing two new arms than would be balanced by the difference in amounts of new tissue in the two cases. In other words, the regenerating tissue, if it be the real cause of loss in body size, exerts a peculiarly great exhaustive influence. The influence is in fact almost malignant in nature. The brittle-stars do not fall completely into line with the above discussion but on closer examination they also seem to supply facts in this direction. In Ophiocoma rusei the rates of regenera- tion for individual arms in specimens injured to various degrees are practically equal (Table [X). ‘The influence of the new tissue would be expected to show itself most markedly in the specimens growing many arms and gradually less in those growing fewer. This is found to be a fact. All individuals were feeding and grow- ing during the experiments but those regenerating five arms in- creased least in size although they had the smallest total amount of tissue to feed and those growing only one or two arms increased most. The other species of brittle-star, Ophiocoma echinata, regen- erates each arm at a rate varying inversely with the extent of injury. When five arms are regenerating each arm grows only 78 per cent as fast as when only one arm is being regenerated. ‘The increase in size of these ophiurans was uniform in all the groups. A pos- sible adjustment of this fact to accord with the medusa and Ophio- coma riiset might be accomplished by assuming that the more rapid growth of the smaller number of arms inhibited the general body increase in size to the same extent as did the larger number of less rapidly regenerating arms. By increase in size is meant the increase in the disk size. When several appendages are re- moved there is less tissue to draw on the food supply and the disk might be expected to increase more rapidly in size under such a 466 Charles R. Stockard condition than it would when four or five long arms are present to be fed. Recognizing this, it might be possible for the speci- mens regenerating four or five arms to increase in size so as to equal those regenerating fewer arms and at the same time the larger number of regenerating arms may have inhibited the increase in size to a greater extent than did the fewer new arms. ‘The actual amount of new tissue formed is much more in the specimens grow- ing the larger number of arms even though the rate of growth for each arm 1s less. It seems then that the regenerating tissue in medusz and ophiu- rans exerts a debilitating influence over the old body tissue in consequence of an excessive power to absorb nutriment. This excessive ability to appropriate nutriment seemingly possessed by regenerating tissue is most significant and deserves careful investigation. Experiments are now under way which I trust may add something towards an analysis of this problem. VIII SUMMARY AND CONCLUSIONS 1 Circular preparations made from the disks of the medusa Cassiopea, regenerate tissue at equal rates whether in periodic pulsation or in a condition of rest. Circular preparations in which one-half pulsates and the other half is at rest regenerate tissue at equal rates from the two halves. ‘The halves are as near as pos- sible identical, being equal portions of one individual still organ- ically connected. ‘Therefore, the process of regeneration in Cas- siopea not only takes place independently of functional activity but the rate of regeneration is also uninfluenced by such a factor. 2 Peripheral pieces of the disk of Cassiopea cut in sundry patterns, bias-strips, equilateral triangles and V’s show decided regulatory ability and tend to assume the original circular shape of the entire disk in the most direct way that their forms will per- mit. ‘The attainment of a circular form either a disk or a cup- shape inhibits the process of regeneration in the pieces, yet regen- eration will continue for a much longer time if such shapes be prevented. ‘The factors here evinced are probably comparable to those which co6rdinate the growth of tissues and organs in such a manner as to insure the specific body form. Studies in Tissue Growth 467 3. The rate of regeneration from a peripheral cut on the Cas- siopea disk is faster the nearer the disk center the cut is made. In the brittle-stars Ophiocoma riisei and Ophiocoma echinata new arms regenerate faster as the old arms are cut off nearer their base of attachment to the body-disk. The nearer the distal end a portion of arm is amputated the slower will a new part regenerate. These experiments and those of several other workers all show that the rate of regeneration in diverse species of animals varies with the level of the cut, being faster as the cut surface is nearer the body center. 4 The rate of regeneration does not bear the same definite relation to the extent of injury in all animal species. The medusa, Cassiopea, regenerates each oral arm at a rate which is independent of the degree of injury when replacing either one, two, four or six of its arms. If, however, eight arms are amputated each arm is regenerated at a rate which, after taking account of the probable error, is significantly greater than the regeneration rates in meduse injured to any less extent. The brittle-star, Ophiocoma riisei, regenerates either one, two, three, four or all five arms at rates which are not significantly different. In other words, there is no relation between the rate of regeneration of the individual arms and the degree of injury in this species. The rate of regeneration for individual arms in Ophiocoma echinata, another species of ophiuran, is fastest when only a single arm is regenerating and successively slower when two, three, four and five arms are being replaced. The rate of regeneration is slower the greater the extent of injury. The facts show that the rate of regeneration does not increase with an increase in the extent of injury in all animals but may actually respond in an opposite manner, or the rate of regeneration may even be independent of the extent of injury. 5 The unfed disk of Cassiopea decreases in size during regen- eration in direct relation to the number of regenerating arms. Thus while the disks which are regenerating eight new arms grow them at the most rapid rate these disks are also decreasing in size most rapidly. 468 Charles R. Stockard In Ophiocoma riisei when all of the individuals are growing, those regenerating a larger number of arms increase in size slower than the specimens regenerating fewer arms. Ophiocoma echinata regenerates each arm faster when only a few arms are cut and such individuals increase in size at about the same rate as do those which are regenerating each arm more slowly although more arms are being replaced. Regenerating tissue possesses an excessive capacity for the absorption of nutriment and may do so even to the detriment of the old body tissue. Studies of Tissue Growth 469 LITERATURE CITED ApamI, J. G., ’01—The Causation of Cancerous and Other New Growths. Brit. Med. J., London, 1go1, i, 621-628. Emme, V. E., ’06—The Relation of Regeneration to the Molting Process in the Lobster. 36th Ann. Rep., Inland Fisheries of Rhode Island, 258-313. ’07—Relation Between Regeneration, the Degree of Injury and Molting in Young Lobsters. Science, n. s. xxv, 785. Ewine, J. ’0o8—Cancer Problems. Arch. Intern. Med., i, 175-218. Kine, H. D., ’983—Regeneration in Asterias vulgaris.. Arch. Entwickl.-Mech., Vil, 351-363. Morean, T. H., ’02—Further Experiments on the Regeneration of the Tail in Fishes. Arch. Entwickl.-Mech., xiv, 539-561. °06—The Physiology of Regeneration. J. Exper. Zodl., iii, 457-500 Scott, G. G., ’07—Further Notes on the Regeneration of the Fins of Fundulus heteroclitus. Biol. Bull., xii, 385-400. StocKaRD, C. R., ’07—Preliminary Report on Regeneration in Cassiopea. Year Book, No. 6, Carnegie Institution, 118-119. ’*o08—Studies of Tissue Growth, I. An Experimental Study of the Rate of Regeneration in Cassiopea xamachana. Carnegie Institution, Publication No. 103, and Science, n. s., xxvii, 448. ZELENY, C., ’03—A Study of the Rate of Regeneration of the Arms in the Brittle Star, Ophioglypha lacertosa. Biol. Bull., vi, 12-17. °o5—The Relation of the Degree of Injury to the Rate of Regeneration. J. Exper. Zodl., 11, 347-369. °o7—The Influence of Degree of Injury, Successive Injury and Functional Activity upon Regeneration in the Scyphomedusa, Cassiopea xamachana. J. Exper. Zodl., v, 265-274. EXPLANATION oF Piate I Six specimens of Cassiopea xamachana regenerating successively greater amounts of material. At the beginning of the experiment the individuals were equal in size, 24 days later the photograph shows that they have decreasedin size in direct relation to the amount of material beingregenerated. The specimens which are growing eight arms and have decreased most in size, although weak and emaciated, regenerate each arm at a faster rate than do those growing fewer arms. STUDIES OF TISSUE GROWTH Cuartes R. Stocxarp Tue JourNAL or ExperIMENTAL Z66LOGY, VOL. VI, NO. 3 FACTORS OF FORM REGULATION IN HARENACTIS ATTENUATA I WOUND REACTION AND RESTITUTION IN GENERAL AND THE REGIONAL FACTORS IN ORAL RESTITUTION BY CMe CHILLED (With Twenty-rour Ficures) The present paper is concerned with a part of the data obtained by the writer on form regulation in Harenactis attenuata (Torrey, ’02) during a stay of several months in the autumn and winter of 1905-06 at the laboratory of the San Diego Marine Biological Association at La Jolla, California. ‘This actinian occurs in great numbers in the fine sand or mud of the tide-flats of False Bay and San Diego Bay and is extremely hardy in the laboratory. Its usual form and habit and the regulatory changes which occur under certain changed environmental conditions have been de- scribed elsewhere (Child, ’09) and only one or two points require mention here. Figs. 1 and 2 are diagrammatic outlines about two-fifths of the natural size, of the shape of individuals in the extended condition, Fig. 1 being a condition approaching maxi- mum extension and distension in large individuals and Fig. 2 a size and shape nearer the average. Extension and distension to the degree indicated in these figures occur only when the animal is in its burrow in the sand. When removed from its burrow and kept without sand during several months it gradually undergoes regulation into a shape resembling that of the “sessile” actinians (Child, ’09). Under the usual conditions of life the circular mus- cles of the body-wall are well developed and the mesenteries bear powerful longitudinal retractor muscles inserted distally on the walls of the oesophagus, whose contraction invaginates the disc and tentacles (Fig. 3). The mesenteries, the mesenterial fila- ments and the longitudinal muscles, particularly the latter, occupy THE JourRNAL OF EXPERIMENTAL ZOOLOGY, VOL. VI, NO. 4- 472 C. M. Child Factors of Form Regulation in Harenactis Attenuata 473 a considerable space in the enteric cavity, so that even when the water has been almost entirely forced out of the cavity by extreme contraction, the body still retains its cylindrical shape, because the cavity is filled by the large muscles and other organs. The twenty-four tentacles vary in length from twenty to forty milli- meters according to the degree of distension and various other conditions. Apparently there is no close correlation between the size of the animal and the length of the tentacles, for in small individuals the length of the tentacles is very commonly nearly or quite as great as in large (Cf. Child ’osb, PP: 272-274). For certain lines of experiment the species has proved to be most favorable material, though the power of invagination of the oral end and the presence in the enteron of the large muscles are complicating factors in many cases. In my experiments the .animals were kept in bowls containing one to two liters of water according to the number of animals. ‘The experiments extended over four and a half months and many individuals were kept during the whole time. No attempt was made to feed the ani- mals: probably they obtained a certain small amount of food from the water, which was renewed every few days, but a considerable decrease in size was observed during the experiments. Aside from this decrease in size and the change in shape (Child ’og), however, a large number, both of whole individuals and of the products of experiment were apparently in perfectly good condi- tion at the end of the time. In the present paper the wound reaction and the course of resti- tution under certain varied conditions, including section at dif- ferent levels are described. “The figures are diagrammatic: Figs. 1 and 2 are about two-fifths natural size, the other figures except Figs. 4 and 5 and 23 and 24 about one-fourth to one-third above natural size. [i Eb REACTIONS 10) THE WOUND In Harenactis, as in other actinians which I have examined, the first reaction to the wound consists in contraction of the regions adjoining the wound. ‘This reaction seems to be characteristic 474 C. M. Child of all parts of the body. Usually it results in a decrease in the size of the opening made by the wound, but it is not at all difficult to cut the body-wall in such manner that closure of the wound is impossible because of the contraction which occurs. In these cases the wound reaction occurs in exactly the usual manner, but the conditions of the experiment determine the character of the result as regards closure or non-closure of the wound. ‘The case is very similar to that of Cerianthus which was considered in an earlier paper (Child ’o4a). In both species, as well as in many other actinians, the same reaction which brings about closure of the wound in certain cases, renders it impossible in others. Under certain other conditions closure of the wound may occur in such manner that return to the usual form is impossible (Child ’o4d, pp: 205-207; ’o8b, pp. 41-45). The actual closure of the wound by new tissue is the result of proliferation and growth of cells adjoining the cut surfaces of the body-wall mesentery or other organ. But in Harenactis, as in Cerianthus (Child ’o4a, pp. 66-74, ’o4b, pp. 276-279, ’08, pp. 30-32), the outgrowth from a cut surface of new tissue with a free margin does not occur to any appreciable extent except under certain special conditions. ‘This point cannot be too strongly emphasized since it is one of the most important features of regu- lation in these and many other actinians. In the papers referred to above, I have shown that the growth of the new tissue begins in the angles of the inrolled portions of the cut surface, and also, and this is the most important point, that it ceases almost imme- diately unless the tissue is subjected to some degree of stretching or tension. ‘The indefinite outgrowth of new tissue from a cut surface until it meets new tissue growing out from another surface does not occur in these forms. In short we may say in general terms that the growth and differentiation of new tissue in the body- wall, the tentacles and the mesenteries 1s possible only when the regions concerned are subjected to a certain degree of mechanical tension or stretching. It 1s unnecessary to cite at length the data concerning this point for Harenactis since they are quite similar to those presented in my earlier papers for Cerianthus. Closure of the wound in Harenactis as in Cerianthus can occur only under Factors of Form Regulation in Harenactis Attenuata 475 the following conditions: First, approximation of at least certain parts of the cut surfaces, in such manner that when the cells of these regions become “embryonic” after the wound has been made union may occur; and second, the existence in the thin mem- brane thus formed of a certain degree of mechanical tension or stretching. ‘The diagrammatic Figs. 4 and 5 will serve to illus- trate the point: Fig. 4 represents a region of the body-wall adjoin- ing a wound, the dotted area indicating the region in which the loss of the original differentiation occurs, i.e., in which the cells become “embryonic.”’ In case the cut margin is straight as in Fig. 4 and no other cut surface is in contact with it, no further growth occurs and the wound is never closed. ‘The cut margin simply heals over, though as I showed for Cerianthus (Child ’o4a, p. 70) it retains for a long time’and probably indefinitely, the power to unite with another cut surface if contact between the two occurs. As a matter of fact, however, in most cases except extensive longitudinal wounds, the contraction of the regions about the wound brings about more or less “ puckering” of the cut margin. In cases of transverse section of the body the cut end commonly appears very much as if drawn together as the mouths of bags are often closed by a “draw-string.” Fig. 5 is a diagram of a portion of such a cut surface in a state of moderate contraction. Various angles are formed between the different parts of the surface and there is of course more or less elevation or depression of different parts perpendicular to the plane of the figure. Under these con- ditions the cells immediately adjoining the cut lose their original differentiation as in the preceding case, but the process does not stop here. These cells form thin membranes across the most acute portions of the angles between different parts of the cut sur- face, i.e., these angles gradually become less deep by the extension from their apices of the thin membranes of new tissue. The dotted lines in Fig. 5 indicate various stages in this process. It will be noted that no appreciable growth occurs where the cut margin is convex. Ina case where the wound was so widely open as in Fig. 5 growth of the new tissue would probably not proceed much beyond the stage indicated in the figure, 1.e., closure of the wound would never occur unless other conditions arose. Commonly, 476 C. M. Child however, the contraction following the wound is so great that the various parts of the circumference of the wound are closely approx- imated. In such cases the margin of the thin new tissue, after it has filled the various angles of the puckered margin, forms a circle. If this circle is not too large, i.e., if its curvature is suff- ciently great, growth of the new tissue continues, the circular opening becomes smaller and is finally closed (Fig. 6). If the opening is above a certain size the new tissue ceases to grow and an opening remains permanently, or until other conditions are estab- 6 lished. In my earlier paper I called attention to the similarity between the growth of these thin membranes of new tissue and the behavior of a fluid or semi-fluid film, and suggested that sur- face-tension probably constituted a factor in the process of out- growth (Child ’o4a, pp. 66-74). The fact that the rapidity of growth of the new tissue increases with the concavity of the free margin or the acuteness of the angle between two portions of the margin (Fig. 5) and, on the other hand, ceases as the margin Factors of Form Regulation in Harenactis Attenuata 477 approaches a straight line or becomes convex, 1s of special inter- est, in that it shows very clearly the dependence of the process upon physical conditions of some sort. Moreover, the thin mem- brane when formed is always under a certain degree of tension, and this condition together with the other facts again suggests surface tension. While this factor is apparently sufhciently important to determine under certain conditions whether growth of new tissue and wound closure shall occur or not, the process of growth itself, when it does occur, is of course very different from the behavior of a fluid or semi-fluid film. After the wound is closed the mechanical conditions of tension favorable to further growth arise from the distension of the enteron with water. Thus far it has been impossible to discover any very great regional physiological difference either in the method or the rate of wound closure, though apparently the rapidity of the reactions is greatest in the oral regions and decreases aborally. In a given region of the body the method of closure of aboral wounds does not differ essentially from that of oral wounds. Certain inciden- tal regional differences in the method and rate of wound closure do, however, appear; these are due primarily to the anatomical structure of the animal, and secondarily to the general occurrence of contraction of the tissues as a wound reaction. ‘The factors chiefly concerned in these differences are briefly considered in the following sections. I The Mesentertes The contraction following the wound involves not only the body- wall but any mesenteries which may have been injured. In a terminal wound in the cesophageal region, for example, the mesen- teries extend from body-wall to cesophagus, and after section the mesenteries contract, as well as the body-wall and the cesophagus. It is this contraction of the transverse cut surface of the mesen- teries that brings the cut margins of the body-wall and the cesopha- gus together with such uniformity in cesophageal regions (Figs. 7 and 8). In consequence of the presence and arrangement of the mesenteries, it is impossible for the cut end of the body-wall to close over the end of the cesophagus. In all cases where section 478 C. M. Child of the body occurs in the cesophageal region, whether it be at the oral or aboral end of a piece, Figs. 7 and 8, Fig. 13, the body-wall and cesophagus unite so that the cesophagus remains widely open to the exterior. Aboral to the cesophagus the mesenteries play little or no part in closing terminal wounds since they hang free in the enteron and their contraction after injury can produce no marked mechanical effect on the body-wall. In these regions oral or aboral ends close MANO 9 f LT Tee by approximation and inrolling of the cut margins (Figs.gand 10), but the closure requires a longer time and is imperfect or retarded much more frequently than in the cesophageal region. In cases of partial transverse section of the body, i.e., of trans- verse lateral wounds, the method of wound-closure depends on whether the cesophagus is involved in the wound or not. When such a wound in the cesophageal region is deep enough to cut through a part of the wall of the cesophagus, the cut surfaces of the mesenteries, oral and aboral to the wound, contract exactly as in the case of a terminal wound, and cesophagus and body-wall Factors of Form Regulation in Harenactis Attenuata 479 are drawn together and unite (Fig. 11). In this manner the lateral mouths and lateral partial discs are formed. If, however, the lateral wound is not sufficiently deep to involve the cesophagus, or if it is in the subcesophageal region, closure occurs, if it occurs at all, by approximation of the cut margins of the body-wall (Fig.12). If the wound is not deep enough to sever the mesenteries of the side of the body, their contraction aids in bringing the cut edges together and closure occurs in a very short time. ‘This is always the case in transverse wounds in the cesopha- geal region which do not involve the cesophagus, and in the sub- cesophageal region when the mesenteries are not completely severed. If, however, a wound in the subcesophageal region is deep enough to sever some of the mesenteries their contraction can no longer aid in bringing the cut margins of the body-wall together. As a matter of fact, such wounds are commonly rather slow in healing and closure usually occurs from each end of the transverse wound toward the middle. Since the body is cylin- drical it is evident that a lateral transverse wound varies in depth from each end toward the middle, being deepest in the middle. In those regions of the wound where the mesenteries are not com- pletely severed their contraction will aid in bringing the cut edges of the body-wall together, while in other regions no such factor will exist. Consequently closure at the ends of such a wound will occur much more readily and rapidly than elsewhere. Frequently the middle regions of these lateral wounds remain open for a long time, since the cut surfaces are not sufficiently approximated for the formation of new tissue between them. Longitudinal wounds of any considerable length in the body of Harenactis very often remain open indefinitely. The margin of the body-wall on each side usually rolls inward spirally so that approximation of the cut edges is impossible, or in some cases one margin precedes the other and the whole body rolls into a single spiral with longitudinal axis. In either case closure of the wound can never occur. “The usual failure of the cut margins of the body- wall to approximate each other after a longitudinal wound is due simply to the fact that there is nothing in the structure of the body which serves to draw the margins together or to prevent their 480 CM Gis attaining some other position. In the case of a terminal or lateral transverse wound, continued spiral inrolling of the cut margin of the body-wall is impossible both because of mechanical conditions in the wall (1.e., the region nearest the wound contracts most strongly) and because of the presence of more or less voluminous organs inthe enteron. Moreover, as pointed out above, the mesen- teries are important factors in bringing the cut margins together in certain cases. In the case of longitudinal wounds, however, none of these factors can serve to bring the margins together, con- sequently closure of such wounds occurs only rarely. In the case of short longitudinal wounds the mechanical conditions on the body-wall prevent any great degree of inrolling, and the cut mar- gins, especially near the ends of the wound, usually approach each other sufficiently to permit the formation of new tissue and closure. On the other hand, it is possible to induce experimentally peculiar methods of wound closure and union of the cut margins in such manner that anything like return to the usual form 1s impossible. Such results can be attained simply by altering the relations between the various mechanical factors involved in approximation of the cut surfaces. One case of this kind, which is of especial interest, will serve to illustrate the point. Pieces of the body from the region between the lines ¢ and d, Fig. 2, contain the large retractor muscles and when contraction occurs after isolation of the pieces (e.g., by two transverse cuts) parts of the muscles and mesenteries usually protrude from one or both ends of the piece (Fig. 14). In such pieces it is possible, with a little care, to remove completely the retractor muscles and the parts of the mesenteries in which they are imbedded. In this operation two results important for the further history of the piece are attained, viz: first, the mass of the enteric organs in the piece 1s much reduced; and, second, the removal of the axial portion of each mesentery leaves a longitudinal wound involving the whole length of each mesentery (see the dotted lines in Fig. 14), besides the transverse wounds made when the piece was isolated. In these pieces the wounded edges of the mesenteries contract and the result is the approximation of oral and aboral cut margins of the body-wall and their union (Fig. 15). This method of closure Factors of Form Regulation in Harenactis Attenuata 481 gives rise to a ring without any opening between the enteron and the exterior (except of course the cinclides), without any oral and aboral end—since these two ends have fused, and in fact without any of the features of shape and relation of parts characteristic 43 of the species. ‘There is no difhculty in obtaining these pieces. They result almost invariably when the conditions described above are established. ‘Their further history is of great interest but will be considered in another connection. Extrusion of the cesophagus sometimes occurs in short pieces from the cesophageal I5 region (e.g., between a and b, Fig. 2). If the extruded structures are cut away rings are formed as in the cases just described and for the same reason. The relation between the mesenteries and closure of the wound in Harenactis does not differ very widely from that existing in Cerianthus (see Child ’o04a, wound-closure after different kind 482 C. M. Child of wounds; ’04d, wound closure in cesophageal pieces; ’o5a formation of lateral mouths). 2 The Esophagus In the cesophageal region the cesophagus and the mesenteries supplement each other: the walls of the cesophagus constitute one of the points of attachment of the mesenteries, and the contrac- tion of the latter draws the cut margins of cesophagus and body- wall together. Both are essential factors in the result, viz, the union of the cut edges of cesophagus and body-wall. Figs. 7-12 and Fig. 13, together with the explanations of these figures in the preceding section, will sufiice to make clear the relations of parts. 3. Mass Effects of the Enteric Organs The mesenteries, with their large longitudinal muscles and mesenterial filaments constitute a mass of considerable size in the post-cesophageal region of the enteron. In most cases of section of the body in regions between the aboral end of the cesophagus and the attenuated region, 1.e., between c and e Fig. 2, the contrac- tion following the wound is sufficient to force the mesenterial organs out through the cut end to a greater or less extent, where they form a plug which often delays or prevents closure of the wound or in other cases becomes infected and frequently brings about the death of the whole piece. In pieces with cut ends both orally and aborally (Fig. 14) the mesenterial organs often protrude from both ends, and such pieces are absolutely incapable of clos- ing and undergoing regulation. ‘The effect of removal of these organs, 1.e., the formation of “‘rings”’ (Fig. 15), was discussed above. If the organs are not removed the extruded portion is sometimes gradually constricted off from the remainder by the continued contraction of the cut end of the body-wall about it. If the exter- nal decaying portions do not infect the rest of the piece before they separate regulation proceeds in the usual manner after the plug has dropped off. On the other hand, in some cases, especially when the piece is intact aborally, more or less elongation usually follows the extreme contraction produced by the wound, and in Factors of Form Regulation in Harenactis Attenuata 483 some of these cases the mesenterial organs which were extruded are again drawn into the enteron and the cut end closes over them without any marked delay or other departure from the usual method. Since the degree of contraction of the body-wall following a wound is in some degree proportional to the intensity of the stim- ulus, pieces cut at both ends extrude their mesenterial organs more frequently and to a greater extent than those with one end intact. Not infrequently, if contraction is considerably greater or more rapid at one end than at the other such pieces turn completely inside out. In this position regulation in the usual manner is of course impossible, and so far as I am aware, pieces which have turned inside out do not succeed in turning back again, but remain in this position until death occurs. As was noted above in Section 1, even the cesophagus may be forced out of the end of the piece in some cases: this occurs only in short pieces within the cesophageal region, 1.e., in which the body-wall and cesophagus are cut at both ends (e.g., pieces between levelsaand b, Fig. 2). If the cesophagus is cut away such pieces form rings; if it remains, closure of the wound and restitution are impossible. The mass of the enteric organs is a much more important factor in determining the occurrence or non-occurrence and the method of restitution in Harenactis than in Cerianthus where these organs are relatively much smaller and very rarely protrude from a cut end. 4. The Nature and Significance of the Wound-Reactions It is evident that the contraction of the cut surface which follows a wound is a general property of the tissues of Harenactis. It 1s not in all cases simply a muscular contraction, for it may occur in regions where muscle fibers are not differentiated or at mght angles to the direction of the fibers. Of course where muscle fibers are injured, they become involved in the reaction. In the case of Cerianthus, where the wound reaction is also apparently independent of muscle fibers and otherwise very similar to that in 484 C. M. Child Harenactis I suggested that the contraction was due primarily to a difference in the elasticity of the different layers of the body- wall, though undoubtedly complicated and modified by other factors (Child ’o4a, pp. 55-65). It is possible the physiological as opposed to the purely mechanical aspects of this reaction were not sufhciently emphasized in that paper, though my work on Harenactis has not brought about any essential modification of my views concerning this point. Here, as in Cerianthus, the in- rolling of the body-wall which follows a wound may be interpreted as the result of a difference in elasticity of the different layers of the wall, though it is probable that the physical condition of the wall may be altered by various conditions, including the wound itself. But the point which seems to me of greatest importance is that the reaction does not appear to possess an adaptive character, either in Harenactis or in Cerianthus. Attention was called to this point in my earlier paper (Child ’o4a, pp. 62-65), but a brief further consideration seems desirable. I am unable to find any basis for the conclusion that the contraction following the wound is an adaptive reaction directed toward bringing the cut margins together and so producing conditions which permit the closure of the wound. Closure may be prevented even more easily in Harenactis than in Cerianthus by the form or position of the wound. In every case where a wound is made the wounded surfaces of the tissues contract, whether closure of the wound results from contraction or not. Whether the wound is closed and how it is closed depend, not upon the contraction which follows the wound, but upon the conditions under which that contraction occurs. In the above consideration of the mesenteries, the cesophagus and the mass of the enteric organs as factors in the closure it is sufh- ciently evident that the contraction may lead to very different results under different conditions. In the cesophageal region, for example, union of the body-wall and cesophagus occurs (Fig. 7, 8, 11), not because this is the process best fitted to bring about return to the “normal form,”’ but because any other method of closure is physically impossible under the conditions. The contraction of the cut mesenteries must bring the margin of body-wall and Factors of Form Regulation in Harenactis Attenuata 485 cesophagus together. ‘This result occurs with the same certainty at the aboral as at the oral end (Fig. 13), though closure of the aboral end in this manner renders a return to the normal form absolutely impossible and the animal is condemned by its own reactions to a death by starvation since there is no opening between the enteron and the exterior. On the other hand, in the subcesophageal region the plug of muscles and mesenterial flaments which is forced out of the wound by the contraction (e.g., Fig. 14) often serves to prevent absolutely the closure of the wound. ‘The formation of the “rings” (Fig. 15) when these plugs are cut away has been described above: these rings are as naturally and necessarily the result under certain conditions of the contraction following the wound as is the “nor- mal” method of closure under other conditions. Contraction follows longitudinal wounds just as it does trans- verse, but in longitudinal wounds of considerable length closure almost never occurs, simply because there is nothing in the struc- ture of the animal which serves to bring the cut edges together. Often in such cases the body rolls up in a spiral about either a longi- tudinal or a transverse axis and closure becomes absolutely 1mpos- sible. Figs. 10-24 of my Cerianthus paper (Child ’o4a) show some of the reactions to longitudinal wounds in Cerianthus. In Harenactis the results are in general similar, though more or less modified by the greater volume of the enteric organs. In short, if we limit our consideration of the wound-contraction to certain cases it may seem to be more or less “teleological,” but if we include all cases it becomes evident at once that the result of the reaction differs very greatly according to conditions, even in some cases making continued existence impossible. The growth of new tissue following the wound does not appear to be an adaptive or teleological reaction any more than the con- traction of the tissues. The conditions under which it occurs have been described above; apparently it occurs wherever these conditions exist, without any relation to the result produced. The growth of new tissue between the cut margins of cesophagus and body-wall occurs as readily at the aboral (Fig. 13) as at the oral end of a piece, although in the former case continued existence 486 TeeeChs becomes impossible because of this growth. “The same may be said concerning the formation of the “rings” (Fig. 15) and various other “‘abnormal”’ results. On the other hand, closure of the wound by new tissue may be delayed for months, or may fail entirely to occur simply because the wound possesses a certain direction or because of certain relations existing between it and parts of the body. The facts cited seem to force the conclusion that various reac- tions which result under certain conditions in the closure of the wound by new tissue are not adaptively directed toward “restora- tion of the normal form” or any other end. On the contrary, they are very evidently in no way concerned with such restoration for they occur just as readily, provided the proper physical con- ditions are present, in cases where their occurrence renders restor- ation of the normal form impossible and leads inevitably to death. It is of interest in this connection to note that Moszkowski in a recent paper takes a very different position. In summing up, he says: “Die Ersatzreaktionen bei Actinien lassen jedenfalls erkennen, dass dem Actinienkorper ein immanentes Bestreben innewohnt, erlittene Verletzungen wieder gut zu machen, und zwar stehen diesen Formen mannigfache Arten des Ersatzes zu Gebote. Welche Ersatzreaktion gewahlt wird, hangt von der Hohe ab, in der operirt wird. Es scheint aber, dass ein weiteres Bestreben vorwaltet, immer diejenige Art des Ersatzes zu wahlen, bei der die Restitutio ad integrum am schnellsten erfolgt. Auch wenn verschiedene Arten des Ersatzes in Konkurrenz miteinander treten, wird immer diejenige obsiegen, welche die rascheste Erreichung des zieles garantiert. Es liegt den Ersatzreaktionen bei Actinien also ein exquisit teleologisches Prinzip zugrunde, wobei es vorlaufig noch unausgemacht bleiben soll, ob dieses Prinzip als ein primares oder sekundar erworbenes anzusehen ist”’ (Moszkowski ’07, p. 432). It is impossible to discuss this paper at length here, but the actual results of experiment do not differ very widely from those recorded by others with other species. After careful study of the paper I can say only that the facts cited do not seem to me to afford a basis for conclusions such as Factors of Form Regulation in Harenactis Attenuata 487 that quoted. I believe that much more extensive series of experl- ments are necessary as a foundation for such a conclusion and it is my experience that the more varied and extensive the experimenta- tion, the less teleological do the reactions appear. In certain cases, at least, the “choice” of a method of reaction, of which Moszkow- ski speaks, is nothing more than the fact, clothed in teleological language, that the structure of the animal is such that only a cer- tain reaction is possible. It might be said for Cerianthus and Harenactis, for example, that in the cesophageal region, the reaction is chosen which brings the cut edges of body-wall and cesophagus together and so hastens restitution. However, when we find that such a reaction occurs aborally as readily as orally although in this case it renders continued existence impossible, it becomes evident that there is nothing adaptive or teleological about it. When we actually analyze it instead of assuming its teleological character, we find that it results primarily simply from the fact that all tissues of these species contract when wounded and freed from the tension resulting from enteric fluid. This being the case, the anatomical arrangement of parts deter- mines that the method of closure shall be different in the cesopha- geal and the subcesophageal regions. The reaction occurs in the characteristic manner in each region, whether it leads to the death of the individual or to complete restitution. Something more convincing than the facts recorded in this paper of Moszkowski’s are necessary before we can accept his conclusions. II THE DIFFERENTIATION OF NEW PARTS The visible stages of the differentiation of new structures re- quires only brief consideration because Harenactis does not differ widely from Cerianthus (Child ’03a). The disc and tentacles are formed by redifferentiation of the most distal portions of the body-wall of the piece, together with the new tissue which closes the wound (Figs. 16-17). Whenthe level of section is in the cesoph- ageal region the union of the cut edges of body-wall and cesophagus leaves the distal end of the old cesophagus widely open as a mouth (Figs. 8 and 16), but in the subcesophageal region the cesophagus i ee €. M. Child is formed anew from the central portion of the new terminal re- gion (Fig. 17). “The new tentacles arise from that region of the body-wall which was originally just aboral to the wound, not from the cut surface itself (Figs. 16 and 17). The diameter of the disc is at first considerably less than the diameter of the column, but as growth proceeds it increases (Figs. 17 and 18). Except in certain special cases to be considered later the number of tentacles is always twenty-four, 1.e., the same as the number of mesenteries, and in restitution as in ontogeny, the tentacles arise over those regions which in the new position of the body-wall following inrolling become the oral ends of the intermesenterial chambers. Since Harenactis possesses a definite number of mes- enteries, twenty-four, which extend the whole length of the body, AN 1 TM new mesenteries are not formed in restitution, though extensive redifferentiation of the old mesenteries may occur, e.g., in the case of restitution of an oral end in the subcesophageal region, where the formation of an cesophageal region is involved. In Cerianthus the new tissue, or at least the regions where growth is most rapid, differ in color from others, but in Haren- actis new tissue or regions of rapid growth very soon become indis- tinguishable from the other parts. In the case of oral restitution it is quite impossible to determine how far the formation of new material is localized in the oral region after the formation of disc and tentacles. In the healing of wounds and restitution where the formation of new tentacles and disc does not occur, i.e., in lateral wounds not involving the cesophagus and in aboral restitution, the new tissue formed is at first distinguishable from the old by greater Factors of Form Regulation in Harenactis Attenuata 489 transparency and delicacy, but there is no sharp boundary between it and the old and the differences of the earlier stages soon dis- appear, so that in these cases also it is impossible to determine exactly to what extent localized growth occurs. ‘The foot-region does not differ from other parts of the body-wall in its gross ana- tomical characteristics, therefore the only certain proof of the restitution of the foot-region is attachment. Localized growth of new tissue can be greatly increased by repeated operation, 1.e., the growth-reaction after each operation is much the same. It is possible in this manner to alter the shape of the body to a very considerable degree. Figures 19 and 20 show the usual method of closure of a lateral transverse wound not involving the cesophagus and not renewed after healing has occurred the first time. ‘The shaded region in the figures indicates approximately the extent of the new tissue. In Figs. 21 and 22 two cases are shown in which the lateral cut was repeated a num- ber of times, in each case after the preceding wound had healed. Each healing involved more or less localized growth in the region of the wound, and deformation of the body was the result. The two figures show different processes of deformation depending on the character of the wound. In Fig. 21 the wound was deep 490 C. M. Child and terminal cut edges of considerable extent existed after each operation. Each time the wound healed the new tissue extended a little further along the cut edges of the mesenteries, so that sepa- ration of the regions oral and aboral to the wound was gradually taking place. In Fig.22, on the other hand the cut was not deep and affected chiefly the body-wall, the mesenteries being but little involved. Here the growth which follows each operation brings about a bulging of the body-wall instead of a depression, as in the preceding case. In these two cases the reaction is exactly the same, so far as can be determined, and the difference in the result is due simply to the different physical conditions undet which the reaction occurs. In cases like Fig.22 more or less regulation of the shape usually occurs if the animal is finally left undisturbed, but a shape like that in Fig. 211s, at least relatively, permanent, though regulation may perhaps occur very slowly. Factors of Form Regulation in Harenactis Attenuata 491 III REGIONAL FACTOR IN ORAL RESTITUTION I The Limits of Restitution and the Limits of Ex periment The high degree of contractility of the body-wall, and the vol- ume of the mesenterial organs constitute together an obstacle to the isolation of small pieces of the body of Harenactis. In the middle region of the body, between the cesophagus and the attenu- ated aboral region itis usually impossible to obtain normal restitution from pieces of less than one-third or one-fourth of the whole length of the body. In such pieces a wound is present at each end and the immediate wound-reaction is so violent that the mesenteries and muscles are forced out of one or both ends of the piece by the contraction following the wound, and form a plug preventing closure, or die and disintegrate and apparently infect other regions. or occasionally the whole piece turns inside out and remains in that condition until death. If the protruding plug be cut off the stimulation produces further contraction and other masses of the mesenterial organs take the place of those removed. The result of the removal of all these organs, viz: the formation of “rings” (Figs. 14 and 15) has already been mentioned. In the cesophageal region short pieces with a wound at each end often extrude the cesophagus. All of these conditions make it practically impossible to determine the regulatory capacities of most regions of the body by isolation of small pieces. ‘The phe- nomena in rings, which will be described later, afford good reason for believing that even very small pieces of the body are physio- logically capable of complete restitution, but usually fail to under- go such restitution simply because closure of the wound is physi- cally impossible after the extrusion of the enteric organs. In the attenuated aboral region the volume of the enteric organs is not great and closure of the wound usually occurs even in very short pieces. Here, however, as will appear below, the physiological capacity is lacking or very much less than elsewhere and restitu- tion is much retarded, or does not occur at all. In pieces with a wound at only one end, 1.e., pieces possessing the original oral or aboral end, the contraction following the wound 492 C. M. Child is usually not as violent as in the above cases and the enteric or- gans are extruded only slightly and temporarily, or not at all. In these cases restitution is complete at the same levels at which it is impossible in pieces with both ends cut, though the physiological . capacity may be the same or almost the same in both. The non-teleological character of the regulatory reactions fol- lowing removal of a part appears very clearly in these relations. In the regions where the body is physiologically capable of com- plete restitution, even in small pieces, the physical conditions resulting from the wound interfere with and prevent restitution in certain cases. In the extreme aboral region, on the other hand, where the physiological capacity for restitution is very slight no such physical obstacles to closure of the wound exist. 2 The Experimental Data Regional Difference in Rapidity and Amount of Restitution As was noted above, the presence or absence of the cesophagus determines to a certain extent the method of wound-closure and the time at which the restitution proper may begin. ‘This, how- ever, is merely an incidental anatomical factor: another regional factor more fundamental in character and quite independent of the cesophagus exists, as is evident from the decreasing rapidity of restitution with increasing distance from the oral end. The time of appearance of the tentacles and the rapidity of their growth are the most conspicuous features of this regional differ- ence. Measurements of the length of the tentacles cannot of course possess any exact value for the length of the tentacles varies within wide limits according to the degree of distension. In individuals kept under similar or identical conditions, however, the degree of distension is likely to be more or less similar, and observation and measurement of a large number of individuals have shown very clearly that a comparison of tentacle-lengths 1s possible. Such measurements do not show anything that cannot be discovered by direct observation; they merely afford a means for recording briefly the results of observation. Factors of Form Regulation in Harenactis Attenuata 493 TABLE I Time of appearance and length of tentacles in mm. at different levels SERIES | 3 DAYS 8 pays 14 DAYS 26 DAYS 136 DAYS Th, (CRO) heeemonccnsee | I-2 | 2 47=10,) 1 W207 12-20 6-12 10 (CLS) Repednecerers ip ton's \e ies 10-15 | 12-15 8-10 iNOl “(G) IO) esse caodeese ae | 5-7 5-10 12-18 70s IN? (Gh ERO) Bees opea conor ba ros O°. 2 10 6 Wo Gp hiro eceeeacodecae | O. | °. °. o-5 degenerating Table I gives the results obtained in five series of three pieces each. ‘The level of the oral end of each series is indicated in the table by the reference to Fig. 2. In all cases the pieces include all that portion of the body proximal to the level of the section, consequently they are of very different sizes: in Series I, for exam- ple they include practically the whole body except the extreme oral end, and each of the following series includes less of the body than the preceding until in Series V only the attenuated aboral end is included, and the pieces are only a small fraction of the size of Series I. ‘The tentacle measurements given in the table are maximum and minimum measurements for all pieces of the series: the tentacles of a single piece are usually very nearly equal in length. In Series IV only one piece out of three closed and produced tentacles: in Series V two pieces closed but only one pro- duced tentacles. As regards the time of appearance of the tentacles, the table is not exact since the pieces were not examined every day, but it shows clearly enough that the length of time between the opera- tion and the appearance of the tentacles increases with increas- ing distance of the level of section from the oral end. ‘That the presence or absence of the cesophagus is not responsible for this difference is evident from the fact that a difference exists both between levels within the cesophageal region (Series I and II) and between those entirely aboral to the cesophageal region (Series IV and V). The tentacles in all series attained their maximum length in fourteen to twenty-six days after the operation. As a matter of 404 C. M. Child fact it so happens that approximately the maximum lengths appear in these series either at fourteen or twenty-six days, thus making it unnecessary to give in the table the measurements of intervening or immediately following dates. In other words, the tentacles in Series I reach their maximum length fourteen days after sec- tion and are considerably reduced at twenty-six days. In the other series the maximum is reached somewhere between fourteen and twenty-six days, but reduction of the tentacles does not begin until later. In Series II the maximum length is almost attained at fourteen days, in Series III the tentacles are scarcely more than half their maximum length after fourteen days, in Series IV al- most the whole growth Be the tentacles occurs after fourteen days and in Series V the tentacles do not appear within fourteen days (as a matter of fact, not until nineteen or twenty days) and soon cease to grow. Series I to III show very clearly that the rapidity of growth of the tentacles decreases with increasing distance from the oral end, i.e., the tentacles not only require a longer time for the first stages of differentiation, but after their differentiation grow more slowly. In Series I a growth of 20-25 mm. occurs within fourteen days, in Series I] 10-15 mm. and in Series III 5-10 mm. in the same time. The time of first appearance of the tentacles in these three series differed by only about two days, yet the difference in the amount of growth after fourteen days is considerable; in Series IV and V, however, the rate of growth after the tentacles appear does not differ very greatly con that of Series III: in Series IV the ten- tacles appeared after about twelve days and grew 10 mm. in the next fourteen days, and in Series V they appeared after about nineteen days and grew 5 mm. in the next seven days. ‘These pieces from the aboral region of the body are always very irregular in time: in some other series the rapidity of growth in such pieces is less than half that in pieces like Series III. However, all the data which I have obtained on the subject indicate that there is a marked dif- ference in the rapidity of growth between distal and proximal regions, and that such difference is independent of gross anatom1- cal features in the different regions. The table shows that after twenty-six days the difference in the Factors of Form Regulation in Harenactis Attenuata 495 length of the tentacles in the different series is much less than after fourteen days: in Series I the tentacles have undergone reduction, in Series II slight growth has occurred in the pieces with shortest tentacles, and in Series III the tentacles have almost doubled their length between fourteen and twenty-six days. In Series IV and V most or all of the growth of the tentacles has occurred during this period. ‘The chief point of interest in this connection is that after twenty-six days there is no great difference in the length of the tentacles in Series I, I, and III, although the difference in size of pieces in the three Series, especially I and III is consider- able: moreover, in Series IV the length of the tentacles (10 mm.) is almost as great as the minimal length of tentacles in Series I (12 mm.), though the pieces of Series IV are less than half as long as those of Series I (Fig.2). In Series V, where the pieces include only the proximal fifth or fourth of the body, the tentacles are only half as long as those in Series IV. It should be noted in passing that in most cases such pieces as those of Series V simply close and do not form tentacles at all. Series V shows the greatest devel- opment of any case observed. Evidently the length of tentacles produced is not proportional to the size of the piece, though larger pieces do produce somewhat ‘longer tentacles. The smaller pieces, however, except in ex- treme proximal regions produce relatively longer tentacles than the larger. One other point remains to be considered in connection with these series, viz: the much more rapid reduction in length of the tentacles after they have reached their maximum length, in Series I than in the other series. Between fourteen and twenty-six days the tentacles in Series I decrease in some cases almost half their length (from 20-25 mm. to 12-20 mm.), while in the other series the decrease in length during four months is in most cases less than this (see Table I, last two columns). In another paper (Child ’og) I have called attention to the change of shape and behavior which Harenactis undergoes when forced to live without sand or mud in which to imbed itself. Decrease in the length of the tentacles is one of the features of this regulation, the atro- phied tips often being visible on the tentacles in such cases. In 496 C. M. Child the case of Series I of Table I the decrease in length of the ten- tacles between fourteen and twenty-six days is a part of this regu- latory process. The growth of these tentacles was so rapid that they attain a much greater length than the tentacles of the other series, consequently as the altered environment made itself felt, the tentacles in Series | were much more affected by it than those of the other series, which had grown more slowly. To sum up: my observations on the regional differences in oral restitution in Harenactis show that rapidity of formation and growth in length of the new tentacles decrease proximally in the body; that the amount of tentacle-restitution is not proportional to the size of the piece nor directly correlated with the region of the body, for smaller pieces produce relatively longer tentacles than larger, even though the level of restitution may be much farther proximal in the former than in the latter. Attention may be called in passing to the fact that in pieces of very different length, but with oral ends at the same level of the body, the rapidity and amount of tentacle-restitution are approx- imately the same, except when the differences in size are extreme, in which case the tentacles of the smaller piece do not attain quite the length of those of the longer piece. In Harenactis the results in small pieces are often complicated by the extrusion of enteric organs, so that it is difficult to obtain satisfactory data for pieces of very small size. In general the regional differences in rapidity and amount of tentacle-restitution and the relations between amount of restitution and size of piece are much the same in Harenactis as in Cerianthus (Child ’o3b): the later series of data confirms the earlier. Regional Differences in the Number of Tentacles In normal individuals of Harenactis the number of tentacles and mesenteries 1s twenty-four: the tentacles form a single row about the margin of the disc and arise as is usual over the inter-mesen- terial chambers. ‘The twenty-four mesenteries are arranged in twelve pairs, consisting of two cycles of six pairs each, the mem- bers of the two cycles alternating with each other (Torrey ’02). Factors of Form Regulation in Harenactis Attenuata 497 The mesenteries of the second cycle are smaller than those of the first, their retractor muscles are less developed and they bear gonads only occasionally. Both cycles, however, extend the whole length of the column, so that twenty-four mesenteries are present in any cross-section. “Thus no new mesenteries are formed in restitution from a distal cut surface at any level and the number of tentacles never exceeds and is very rarely less than twenty-four. Pieces from the proximal fifth or fourth of the body, however, when they produce tentacles at all produce a smaller number than twenty-four. Usually such pieces simply close after a consider- able time but produce no tentacles. In only four cases of this kind has the formation of tentacles been observed. One of these cases, a piece consisting as nearly as could be deter- mined, of that portion of the body proximal to the line e in Fig. 2, gave rise at first to ten tentacles and very soon after to two more, twelve in all (Fig. 23). The ten tentacles were first visible four- teen days after section and the two additional appeared within the following week. Of the twelve tentacles as shown in Fig. 23 eleven are about equal in length and one is much shorter. A month later the piece possessed fourteen tentacles, thirteen long and one short. Unfortunately the position and sequence of new tentacles was not determined. ‘Though the piece was kept for two months more no more tentacles were produced. A second case in which the piece included approximately the region proximal to the line f in Fig.2, gave rise after twenty days to twelve tentacles which attained a length of 5-6 mm. No addi- tional tentacles were produced by this piece and the tentacles present underwent gradual reduction during two months following, after which the piece began to degenerate. In two other cases in whi h the level of section was somewhere in the region of the line d, Fig. 2, the tentacles appeared after six- teen days. Their number was not determined at first, but ten days later it was found to be twenty-one in each case. A diagram of the disc of one of these cases is shown in Fig. 24: fifteen long and six short tentacles are present, and it is clear that the short ten- tacles belong to both cycles and are not definitely related as re- gards position to the siphonoglyphe. In the other case similar 498 C. M. Child irregularities are present, | kewise without definite relation tothe axes of symmetry. During two months following no additional tentacles were produced and then death occurred. These cases present a problem of considerable interest. In all cases twenty-four mesenteries were present in the pieces at the time of section: have some of these mesenteries degenerated or does the usual relation between tentacles and mesenteries or inter- mesenterial chambers not exist in these pieces? Microscopical examination would of course answer this question at once, but since I was desirous of determining whether these pieces were capable of producing more tentacles they were kept alive as long as pos- sible. Several attempts to obtain other pieces of the sort after —— their importance became evident failed, the pieces closing but pro- ducing no tentacles. I hope to return to the matter in the future and to decide this question. I believe, however, that the reasons for the reduced number of tentacles are not far to seek. In the first place, reduction in the number of tentacles may be brought about in more than one way: two different sorts of reduction undoubtedly occur in the pieces described. In the first piece (Fig. 23) and in the second, twelve tentacles arise—in the first piece two of them are later than the others—and there is no external indication of the other twelve. The tentacles present are regularly arranged and, except for one short tentacle in the first piece about equal in length. In conse- quence of the slight distension the body-wall was rather opaque Factors of Form Regulation in Harenactis Attenuata 499 and the number of mesenteries could not be determined with certainty by external examination. This form of reduction in the number of tentacles is due, I believe, to the small size of the individual formed from the pieces. In Harenactis, as in other actinians, more or less definite space- relations exist between the mesenteries. “This can only mean that the correlations between a mesentery and adjoining regions are of such a nature that other mesenteries are prevented from devel- oping within a certain distance of the already existing mesentery. Consequently if the size of a region in which mesenteries are present undergoes a real physiological decrease, the mesenteries are brought “too near” each other. In cases of this sort, where two organs come into physiological conflict, we commonly find, at least in the lower forms, either that one of them undergoes atrophy, resorption, or separation, or that both are more or less reduced. In the two cases under consideration these conditions are pres- ent. The closure of the pieces after section is very slow, conse- quently they remain collapsed for a long time. The body-wall and mesenteries of Harenactis, like those of other actinians with which I have worked, undergo more or less complete atrophy in the absence of the functional stimulus arising from distension of the body by fluid (Cf. Child ’o4b, ’o4c, ’o4d, ’oge, 05a, 08, ’0Q). There is then in such pieces a real physiological decrease in size. I think it probable that under these conditions the mesenteries of the second cycle, i. e., the six pairs of smaller, less highly devel- oped mesenteries, undergo atrophy or degeneration. This of course reduces the number of the inter-mesenterial chambers and consequently the number of tentacles by one-half. By this method of reduction in number every alternate tentacle disappears: appar- ently that is exactly what has occurred in the two pieces with twelve tentacles. One of these pieces produced two more ten- tacles after a considerable time: in this case it may be that the atrophy of some of the second cycle of mesenteries was not com- plete, or else that with the functional growth following renewed distension the reappearance of the mesenteries of the second cycle or of some of them becomes possible again. 500 C. M. Child In the other two cases described only three tentacles were entirely absent, but a number of others in irregularly arranged groups about the margin of the disc appeared later and were for a long time much smaller than the remainder (Fig.24). In these cases no definite relations exist between the mesenterial arrange- ment and the regions where tentacles are lost or reduced in size. In my work on Cerianthus I showed that local inhibition of tentacle-formation might occur as the result of folds in the body- wall (Child ’o4b, pp. 281-284). If these folds persist for any considerable length of time atrophy of the region occurs. ‘This is especially noticeable in Cerianthus zstuarit (Child ’08) where complete degeneration and disintegration of the body-wall in folds and compressed regions often occurs. The reason for this local atrophy in folded or wrinkled regions lies in the fact that even in “collapsed”’ pieces, i.e., in pieces with an artificial opening into the enteron which prevents normal dis- tension, some slight degree of distension exists so long as the pieces are undisturbed (Child ’o4b, ’o4c, etc.). In Cerianthus the wound becomes plugged by the slime secretion, in Harenactis by slime and in certain regions by the cut ends of muscles and mesen- teries. Consequently some degree of distension arises, aS can easily be shown experimentally, but the pressure cannot attain anything like that in normal animals for as soon as it exceeds a certain small amount water escapes through the wound. Such slight distension is not sufficient to remove the folds and wrinkles from the body-wall, and these regions are consequently not sub- jected even to the slight degree of tension existing elsewhere, therefore they undergo more reduction or atrophy than the other regions. Experimental demonstration of these facts is not in the least difficult. I have observed and established experimentally these conditions in a very large number of cases, especially in Cerianthus zstuaril. The two pieces of Harenactis are simply further examples of local inhibition of tentacle-formation in consequence of local atrophy. Since closure is retarded in these pieces they remain collapsed for a much longer time than is usual in more distal pieces, and during the period of collapse the body becomes variously Factors of Form Regulation in Harenactis Attenuata 501 folded and greatly reduced in size. ‘There is no doubt that a ereater or less degree of atrophy occurs in the most compressed regions: if the period of collapse is sufficiently long the local atro- phy may bring about complete disappearance of certain mesen- teries. When distension occurs and tentacles develop the ten- tacles corresponding to these mesenteries will be absent, or if the mesenteries reappear after the normal functional conditions are reéstablished, these tentacles may develop later, and perhaps remain of small size for a longer or shorter time. In short, these cases of local absence or retardation of development of tentacles in short proximal pieces are the result of local reduction or atrophy of body-wall and mesenteries which in turn is the result of com- pression or of absence of tension in wrinkled or folded regions of collapsed or almost collapsed pieces. As regards regional localization of these phenomena in the body, irregular reduction in number and size of tentacles may occur at any level of the body if the proper conditions arise, but in pieces from the more distal regions closure and distension usually occur so rapidly, if they occur at all, that there is no time for atrophy such as occurs in these proximal pieces. Regular reduction to half the number of tentacles has been observed only in the two proximal pieces above described (fg. 23), but, as was shown in preceding sections, the course of restitution in small pieces from other regions of the body is so highly modified by the presence of the cesophagus in the more distal regions and the large mass of enteric organs in other regions that it is impos- sible to obtain any accurate data as to the physiological capacities of these regions. It is probable, however, that the mesenteries, and especially those of the second cycle disappear more readily in the proximal regions of the body than elsewhere, because they are much smaller and less highly differentiated there than elsewhere. IV GENERAL CONSIDERATIONS It has been pointed out above that the contraction following the wound is not adaptive in character; apparently it is merely the direct result of the stimulus of the wound, and it occurs in essen- 502 C. M. Child tially the same manner whatever parts are concerned. It is not even directly related to the closure of the wound, for under cer- tain conditions it renders closure impossible. The process of closure of the wound by new tissue is ‘likewise not related to any “purpose” such as the return to the normal form but is determined like any other physico-chemical process by con- stitution and conditions. In many cases the process of wound closure renders absolutely impossible any “return” and leads inevitably to death. Whether closure of the wound shall occur in such manner as to render possible continued existence and restitution of the part lost depends very largely upon the anatom- ical structure of the region involved, i.e., the arrangement of mesen- teries, the cesophagus, the enteric organs, etc. It is also depend- ent on the character of the wound contraction, for in many cases this contraction establishes physical conditions which render closure impossible. Physiologically the character of the wound reaction complex is apparently much the same in different regions of the body, though its morphological results may be very different, according to the anatomical relations of parts. ‘The only regional physio- logical difference which I have been able to discover is a slight decrease in rapidity of the growth of new tissue with increasing distance from the oral end. The closure of the wound, at least a provisional closure or plug- ging, is a necessary condition for the occurrence of anything like normal restitution, for in the absence of distension, atrophy of the body-wall and other parts occurs instead of growth. ‘The growth of new tissue in general occurs only under a certain degree of mechanical tension: restitution proper consists therefore, alae entirely of localized differentiation in a continuous sheet of tissue rather than of outgrowth from a cut surface. The regional differences in oral restitution are apparently chiefly difference of quantity. The rapidity of restitution de- creases with increasing distance from the oral end of the body. This difference is apparently physiological in character and independent of the gross anatomical relations of parts. It is, I believe, an expression of a physiological characteristic common Factors of Form Regulation in Harenactis Attenuata 503 to at least many forms with well marked polarity. In such forms the anterior or oral end and the regions adjoining it are commonly the regions of most rapid or most intense reaction to external con- ditions, and are usually more frequently or more continuously affected by these conditions and their changes. In short the anterior or oral end is commonly dominant functionally in the body (Child ’08a). Regions adjoining this pole share to a greater or less extent in its activity, and their physiological character is more or less completely determined by their correlations with it. In the more complex forms these correlations may be very definite in character and localization, and structural localization may be correspondingly definite, but in simple forms like the actinians the degree of correlation between the anterior or oral region and other parts is apparently more or less nearly proportional to the dis- tance between them, i.e., the greater the distance between the oral end and a given region the less the physiological similarity between them. If the body of such a form, e.g., Harenactis, be cut into pieces the oral end of each piece is, so far as “oral proc- esses”’ are concerned, the dominant region. Since visible morpho- logical differentiation is to be regarded as the expression of the functional processes in the system, the development of the morpho- logical structures characteristic of an oral end may be expected to occur in the most oral region of the piece, if anywhere. Such development will occur, provided the region is sufficiently similar to the original oral region in its physiological capacities or becomes sufficiently similar in consequence of its new position and correla- ations as the most oral region of the body. No exact limit can be established for the occurrence or non-occurrence of restitution in a given case: we can only say that if the region in question is or becomes so far similar physiologically to the part removed, that it can take the place of the latter to a certain extent in the system, it will develop a morphological structure approaching that of the part removed. The completeness of the restitution will depend upon the completeness with which the substituted region takes the place of the old in the system. In Harenactis all levels of the body except the extreme aboral region are capable of substitution for the original end in sufhcient 504 C. M. Child degree to give rise to the characteristic structures. But the rapid- ity of morphological restitution decreases as the distance of the level of restitution from the original oral end increases. As I have attempted to show above, this means simply that physio- logical likeness to the oral end decreases with increasing distance from that end. This change in physiological likeness with difference in level is not uniform: in the oral half or two-thirds of the body it is not very great, but further aborally the rapidity of oral restitution decreases rapidly until it becomes zero, 1.e., in the extreme aboral region the distal end of the piece is incapable of physiological sub- stitution for the original oral end in sufficient degree to produce visible morphological results. That such substitution is not a purposive or adaptive act but merely a necessary consequence of the constitution of the system and particularly of the correlations of its parts, I have pointed out elsewhere (Child ’o8a). In my discussion of polarity in Tubularia I called attention to the fact that polarity might appear in quantitative and qualitative regional differences as mall as in the oral-aboral or anterior-pos- terior differences which are commonly termed polar differences (Child ° o7. [hese regional differences in the rapidity of restitu- tion in Harenactis are as a matter of fact one expression of the physiological differences along the axis which we commonly group together under head of polarity. But discussion along this line is postponed until after the presentation of further data. SUMMARY 1. The wound-reactions and the general course of restitution in Harenactis do not differ widely from those processes in Cerian- thus. The contraction following the wound is certainly not purely muscular: it is apparently characteristic of the tissues in general and is probably, at least in part, the consequence of certain phys- ical properties of the tissues. The result of this contraction de- pends upon incidental factors, at least in large measure: under certain conditions the contraction approximates the margins of Factors of Form Regulation in Harenactis Attenuata 505 the wound and renders possible its closure by new tissue, but under other conditions closure becomes physically impossible, although contraction occurred in the usual manner. 2 Growth of new tissue occurs only under a certain degree of mechanical tension, therefore, no appreciable degree of growth occurs on a single free cut surface. Closure of wounds by new tissue occurs only when two cut surfaces, or two parts of a cut surface are closely approximated to each other or partly in con- tact. The conditions determining the extension over the opening of the thin membrane of new tissue are such as to indicate that the physical factor of capillarity plays an important role in the process. 3. Union may occur between any two cut surfaces which hap- ¢ pen to come into contact, without any relation to the “normal” form. Under the usual conditions the gross anatomical feature of the region concerned in a particular case, e.g., the oesophagus, the mesenteries, the retractor muscles, etc., determine whether and how closure shall occur under certain conditions, e.g., in pieces from the cesophageal region, these anatomical factors deter- mine that closure shall occur in such a manner that continued existence becomes ‘mpossible. 4 As in Cerianthus, the tentacles are not outgrowths from a wound, but localized differentiations in a continuous sheet of tissue. 5 The rapidity of wound closure and of oral restitution de- creases with increasing distance of the level from the original oral end. ‘This decrease is not uniform, being small in amount in the oral two-thirds of the body and much greater in the aboral third. In small pieces from the extreme aboral end oral restitution does not occur, but closure of the wound may take place sooner or later, the reaction being slow. 6 In cases where oral restitution does occur in small pieces from the proximal region of the body the number of tentacles is sometimes twelve instead of twenty-four. This development of only half the usual number of tentacles is probably a result of dis- appearance of the secondary cycle of mesenteries in these very small pieces, i.e., a phenomenon of size rather than of region, 506 C. M. Child though mesenterial atrophy or resorption probably occurs more readily in the proximal region than elsewhere. BIBLIOGRAPHY Cuitp, C. M—Form Regulation in Cerianthus. ’o3a—I. The Typical Course of Regeneration. Biol. Bull., vol. v, no. 5, 1903. ’o3b—II. The Effect of Position, Size and Other Factors upon Regenera- tion. Biol. Bull., vol. v, no. 6; vol. vi, no. 1, 1903. ‘°o4a—III. The Initiation of Regeneration. Biol. Bull., vol. vi, no. 2, 1904. °o4b—IV. The Role of Water-Pressure in Regeneration. Biol. Bull., vol. vi, no. 6, 1904. ’o4c—V. The Role of Water-Pressure in Regeneration: Further Experi- ments. Biol. Bull., vol, vii, no. 3, 1904. °o4d—VI. Certain Special Cases of Regulation and their Relation to Internal Pressure. Biol. Bull., vol. vii, no. 4, 1904. ‘o4e—VII. Tentacle-Reduction and Other Experiments. Biol. Bull., vol. vii, no. 6, 1904. ’o5a—VIII. Supplementary Partial Discs and Heteromorphic Tentacles. Biol. Bull., vol. viii, no. 2, 1905. ’o5b—IX. Regulation, Form, and Proportion. Biol. Bull., vol. viii, no. 5, 1905. °o7—An Analysis of Form-Regulation in Tubularia. VI. The Signifi- cance of Certain Modifications of Regulation; Polarity and Form-Regulation in General. Arch. f. Entwickelungsmech. Bd-aaxiy, EH. 21007. °08—The Physiological Basis of Restitution of Lost Parts. Journ. Exp. Zo6l., vol. v, no. 4, 1908. . 08b—Form Regulation in Cerianthus Aéstuarii. Biol. Bull., vol. xv, no. I, 1908. Regulation of Harenactis attenuata in Altered Environment. Biol. Bull., vol. 00, no.oo Moszxkowsk1, M. ’07—Die Ersatzreaktionen bei Actinien (Actinia e@quina und Actinoloba dianthus). Arch. f. Entwickelungsmech. Bd. xxiv, |g BEN Wee i(0\ 0) Torrey, H. B. ’02—Papers from the Harriman Alaska Expedition. xxx. Ane- mones, with Discussion of Variation in Metridium. Proceed- ings of the Washington Acad. of Sci., vol. iv, 1902. fine ERRECIS OF CENTRIFUGAL FORCE UPON THE EGGS OF SOME CHRYSOMELID BEETLES? BY R. W. HEGNER Wirn Twenty-Four Ficures CONTENTS Tt, “Tbetaefs Rhsr0) 0p Seen CORT OR SnIa oT tite Deena anna Mea NOC aG NOAA OG OSS 507 lee ateriallam dime th ods tye rctes sia accusers eo sis e/a = ie oisvars a ickels Sieveyetate sssie-sieeverore eel eteteroistersraiacsrais ois 508 I The orientation of the eggs of Calligrapha bigsbyana ..........6..00.--cceenrceseenecees 510 IV The effects of gravity upon the development of the eggs of insects ...........++eeeeeeeeeeee 511 V_ Abrief account of the normal embryonic development of Calligrapha bigsbyana............ 512 VI The structure of the egg of Calligrapha bigsbyana at the time of deposition ...............-- 516 VIL The effects of centrifugal force upon eggs centrifuged after deposition...........+...+..06- 517 VIII The effects of centrifugal force upon eggs laid by centrifuged beetles...............-20.0005 538 il LBxQoeicnneaisankdla Colplsd N/E 5 cpopaocepaeonoDoo Up Bo oucoMsaudoDUDoGOdDoNSaG0DOC 538 2 Experiments with Leptinotarsa decemlineata ..........-.20:eceeeeceeceerncersuces 539 IX Review of the effects of centrifugal force upon developing eggs.........-.+seeeeeeeeeeeeee 540 Tene ldestrib ution obthe eee COMtEMESs/\aicisle sie lclelerene)oleitte te eer stated ye ialetelensietetataletel oh st=t=r 540 2 The restitution of the egg substances after centrifuging. ...........-eeeeee eee ences 546 ae bheapelottherepp whem cemtrsu pedir. cic) -!tslclele ei eteler=)-ieelelelrlok-lolel+)e]a]o!e(el-felels=)-)=ici-F-i= 547 / “ayereteoy: Glaielloyynvaslt Soap odsocnnboud sondconeow0couEaS anode gacossacdcououddse 547 5 Eggs centrifuged before deposition. ...........- 00+. 2cc cece reece tere eee reccescenee 548 25 QUENTETR Ga coq cocoooe aduooJObes sdoduooa apo bod OO DoAaadusoDbdonop HON GS ODosHoOd IDC 548 SkID DACA aco SERA Ce OU OU BORO G OAC OD AyD OSG OU GH DAOU OOM OO OIC COCA 550 I INTRODUCTION It has been found that insect eggs are definitely oriented within the ovaries and the exact position of the future embryo seems to be determined already at this early period. This fact has led many embryologists to believe that the eggs of insects are very highly organized. If this is true a redistribution of the contents of the egg would have a profound effect upon the development of the embryo. In order to obtain a rearrangement of material a centrifugal machine was used successfully, as is shown by the experiments described in Part VII of this paper. 1 Contributions from the Zodlogical Laboratory of the University of Michigan. No. 125. Tue JourNnaL or ExPEeRIMENTAL ZOOLOGY, VOL. VI, NOs 4- 508 R. W. Hegner During the course of the study of the germ-cells in some chryso- melid beetles? a disc-shaped mass of granules (Fig. 9, g. c. d) was discovered in the freshly laid eggs suspended in the peripheral layer of cytoplasm at the posterior end. I have called this struc- ture the “pole-disc.” ‘These granules are taken up by the germ- cells in the course of their migration and apparently determine the character of these cells; on this account I have called them “germ- cell determinants” (Hegner ’o8 b). It was hoped by means of centrifugal force to scatter the granules of the pole-disc and obtain an embryo either without germ-cells or with germ-cells in various parts of the body. It was also thought possible that the pole-disc might move as a whole and, becoming massed in some other region of the egg, might influence at this point cells which would ordinarily become body-cells. As will be seen later some data were secured but not enough to warrant any definite conclusions. So far as I have been able to learn from the literature no experi- ments with centrifugal force upon the eggs of insects have ever been performed successfully and in only one case has any arthropod ege been tested in a centrifugal machine (Lyon ’o7). Lyon merely says: “The ovarian eggs of the common garden spider could be separated by one minute’s centrifugalizing into two layers” (p. 169). The experiments described below were begun at the University of Wisconsin in the spring of 1908 and were continued at the Marine Biological Laboratory at Woods Holl, Mass., where I occupied a room subscribed for by the Wistar Institute of Anatomy and Biology. ‘The material was further studied at the Zoological Laboratory of the University of Michigan. II MATERIAL AND METHODS During the course of this work eggs of the following beetles were used for experiments: Calligrapha multipunctata, C. bigsbyana, C. lunata, Leptinotarsa decemlineata and Lema trilineata. ‘The posterior ends of these eggs are fastened to the leaf on which they 2 The Origin and Early History of the Germ-Cells in Some Chrysomelid Beetles. Accepted for publication by the Journal of Morphology. Centrifugal Force upon Beetles’ Eggs 509 are laid. In the case of Calligrapha multipunctata and C. bigs- byana the eggs can be definitely oriented as is explained in the next part of this paper. A number of beetles were kept in the laboratory and the eggs were marked on the anterior-ventral sur- face with a small spot of waterproof india ink. ‘The exact time of deposition was recorded in all cases. The eggs are not always in the same stage of development at the time of laying, but all those in one batch are approximately in the same condition. When the eggs had developed to the desired point they were placed in small indentations in a block of parafhn. ‘The entire block containing the eggs was then lowered to the bottom of a glass tube of an ordinary water-power centrifugal machine. The eggs were then 15 cm. from the axis of rotation. “The number of revolutions per minute was not accurately determined, but was probably between 1500 and 2000, although in some cases (those described in experi- ments C. M. 1 and L. D. 1 and 2) a slower rate of speed was used (360 revolutions per minute). The eggs when taken from the centrifugal machine were left in the cavities in the paraffin block with the heavy end down until they were fixed. In previous work I found a modification of Petrunkewitsch’s fluid the best for killing and fixing the eggs. [his was used entirely for the centrifuged material, although control eggs were fixed in a number of the common mixtures. Eggs were stained in toto with Mayer’s hzmalum acidulated with 2 per cent of glacial acetic acid, or with alum cochineal. Sections were stained on the slide principally with hemalum followed by Bordeaux red. One difficulty in doing experimental work with the eggs of Calli- grapha is that only a few are laid at one time (eight is the average number) and, as the conditions of the experiments frequently are responsible for the destruction of some of these, no series contains very many successive stages. There are also causes for trouble in making preparations. In some it stances the eges stuck fast to the chorion at the outer end, where the contents had been strongly driven against it; the chorion could not be removed from these without injury to that part of the ege. After eggs have been centrifuged they are more difficult to section than before because the large deutoplasmic spheres collect 510 R. W. Hegner at the end away from the axis of rotation and a breaking out is frequent in this region. III THE ORIENTATION OF THE EGGS OF CALLIGRAPHA BIGSBYANA It has been known for more than twenty years that the eggs of insects are definitely oriented within the ovaries of the adults. Hallez in 1886, finding this to be true of the ova of Hydrophilus and Locusta, expressed the fact in his “ Loi de l’orientation de l’embryon chez les Insectes’’ as follows. ‘‘La cellule-oeuf posséde la méme orientation que l’organisme maternel qui l’a produit: elle a un pole cephalique et un pole caudal, un coté droit et un coté gauche, 2. Fic.1 A diagrammatic drawing of C. bigsbyana clinging to the under side of a willow leaf and showing the orientation of the egg in the ovarian tubule and after deposition. Fic. 2 Four eggs of C. bigsbyana laid intworows. a. = anterior. d. = dorsal. /. = left. p. = posterior. r. = right. x. = anterior ventral surface where the spot of India ink was placed as a guide for orienting the eggs during the experiments. une face dorsale et une face ventrale; et ces différentes faces de la cellule-oeuf coincident aux faces correspondantes de l’embryon.”’ No difficulty is experienced in distinguishing the anterior from the posterior end of the eggs of Chrysomelid beetles as it is always the posterior end which first emerges from the vagina. This end is fastened to the leaf on which the egg is laid and subsequently becomes the posterior end of the embryo, regardless of the posi- tion of the leaf. In only two species (Calligrapha multipunctata and C. bigsbyana) of the many Chrysomelid beetles examined could the right and left sides of the egg be accurately determined. The egg laying of these insects is as follows: ‘‘The beetle selects Centrifugal Force upon Beetles’ Eggs 511 a leaf and clings to its under surface. ‘The tip of the abdomen moves ehyfieaicolly up and down about fifteen times at intervals of a little less than one second. ‘This results in the exudation of a drop of viscid, colorless fluid about one-third the transverse diameter of the egg. ‘he egg is forced out a moment later and carries with it this drop of fluid by means of which it is fastened to the leaf. When the egg reaches the leaf it is pushed back away from the beetle (Fig. 1), which then moves to one side and again begins the rhythmical movements which precede the laying of another ege. In this way eggs are laid in a double row as shown in the accompanying figure (Fig. 2), but frequently three or more may be laid in one row. ‘The intervals between the layings of the individual eggs average one minute and twenty seconds” (Hegner 08 a). ‘Two to nineteen eggs are laid at one time, the average number being eight. Fig. 1 indicates the orientation of the egg of C. bigsbyana lying in the ovary and also the final position after it has been laid. IV THE EFFECTS OF GRAVITY UPON THE DEVELOPMENT OF THE EGGS OF INSECTS That the position of the insect egg after laying has no influence upon the development of the embryo was proved by Wheeler (1889) in the case of Blatta. This author kept capsules from fourteen to twenty days in the following positions: “t Resting with the lateral faces perpendicular and crista uppermost. “2 Resting on the crista with the lateral faces perpendicular. «2 Resting on the left lateral face. “4 Resting perpendicularly on the anterior end. «5 Resting perpendicularly on the posterior end. “In all these cases the eggs developed normally, without the slightest indication of displacement in position or alteration of shape in the embryo; whether they were forced to develop with their heads pointing up or down.” ‘The conclusion reached was that “the force of gravitation has no perceptible effect on the development of the eggs of Blatta ‘i 512 R. W. Hegner Wheeler also proved that the antero-posterior differentiation of the embryo of Leptinotarsa decemlineata is not affected by changes in the position of the egg after laying, but is predetermined in the ovary. During the course of my work with Calligrapha, eggs were taken as soon as laid and placed inevery possible position. “lhe embryos were found to be in no way affected by the orientation of the egg with respect to gravity. The only exception to the rule that gravity has no influence upon the development of insects’ eggs seems to be that of Hydrophilus aterrimus reported by Megusar (’06). The eggs of this water beetle are laid in a boat-like cocoon which is kept in an upright position in the water by means of a peculiar mast. Megusar found that if these cocoons were inverted, thus also inverting the eggs within, the development of the eggs was retarded and a deformity in the embryos resulted. ‘The small number of larvz that hatched lived for only a short time. V A BRIEF ACCOUNT OF THE NORMAL EMBRYONIC DEVELOPMENT OF CALLIGRAPHA BIGSBYANA Eggs that have just been laid contain polar bodies in various phases of formation; these are given off into a thickening of the “Keimhautblastem” at a point slightly anterior to the median transverse axis of the egg. The female pronucleus lies in an amoeboid accumulation of cytoplasm among the yolk-globules. It moves inward and conjugates with the male pronucleus at a point level with the polar bodies. Here the first cleavage divisions take place. As cleavage progresses a separation of the nuclei into two sections occurs. ‘The nuclei of one group form a more or less regular layer equidistant from the periphery; these pre- blastodermic nuclei (Fig. 17, pb/. m) move outward and fuse with the “Keimhautblastem.”’ Cell walls now appear for the first time and a blastoderm is formed of a single layer of regularly arranged cells. The nuclei of the other group (vitellophags), remain behind scattered throughout the yolk (Fig. 17, vt). Eight of the nuclei that reach the posterior end of the egg do not remain a5 Centrifugal Force upon Beetles’ Eggs = 32 +3-- *ppoj-qrea = {7 ‘aAoo1d yemuaA ‘pueq-ui1es = gS *syjao-wi4ie3 yerprowtd =98'¢ —-aovyins [esrop ayy uo dn Aem-yey ppoj-[te3 ay} SuIMoYs eurAqssiq “> Jo oATqu) uy i *yuotr8as 0} SutuurIseq ysn{ euekqsstq *>D Jo oAquia uy *2A00I8 [PIJUSA JY} puk pULq WIId ay} SUTMOYS plo snoY OM}-AjITY] LULAGSSI *- Jo 38a Uv JO adPJANS [PIIUBA JY} JO MITA “pus ro1a3sod ay3 ye dnosd e wr103 sjfao-wed peIprowid ayy, ‘wortsodap saqye sinoy 1noj-AJUaM} vULAssIq *>D Jo 33a uv Jo MarA aovyING 9 S +; 48 9 OI S-org F917 €-o1g 514 R. W. Hegner in the peripheral layer, but collect about them a number of granules (germ-cell determinants, Fig. 17, g c d) which they encounter in this region and continue their migration until they are entirely separated from the blastoderm. ‘These are the primordial germ- cells (Fig. 3, p.gc). The first change noticed in the blastodermisa crowding together of the cells on the ventral surface of the egg. This results in the formation of a broad longitudinal band of closely ageregated cells, the ventral plate. ‘The edges of this plate are Fic. 7 Surface view of embryo described in Series C.B. 2, c. The embryo has begun to broaden and shorten. Fic. 8 An embryo of C. bigsbyana in which the tail-fold is coincident with the posterior end of the egg. c.ap.== cephalic appendage. p.= posterior. t.ap.= thoracic appendage. 7.f. = tail-fold. thrown up into two folds; these spread out in the posterior region extending to the end of the egg where they pass around the pri- mordial germ-cells and meet on the dorsal surface. ‘The ventral plate now decreases both in length and in breadth and a longi- tudinal concavity, the ventral groove, appears. ‘The germ-band can now be recognized; it covers the entire ventral surface of the egg except a wedge-shaped area anterior to the groove (Fig. 4, gb). Centrifugal Force upon Beetles’ Eggs 515 The germ-band becomes narrower as development advances; its posterior end pushes around that end of the egg and up on the dorsal surface. The lateral folds gradually cover over the ventral groove and the amnioserosal fold grows forward from the posterior Fic. 9 Longitudinal section through an egg of C. bigsbyana four hours after deposition. ¥ gc.d. = germ-cell determinants. gn.= germ-nuclei copulating. Ahb/. = ‘‘Keimhautblastem.” p. =§pos- terior. v.m. = vitelline membrane. y. = yolk. end to meet the anterior fold (Fig. 5). The segmentation of the germ-band and the lengthening of the entire embryo now pro- gresses rapidly. The cephalic extremity extends almost to the 516 R. W. Hegner anterior end of the egg and the tail-fold extends a little more than half-way up on the dorsal surface (Fig. 6). The tail-fold now begins to recede as the embryo shortens and broadens (Fig. 7) and in a short time coincides with the posterior end of the egg. The embryo now grows laterally around the yolk (Fig. 8), its various parts being situated approximately in the positions they occupy at the end of about six days when it hatches as a larva. VI THE STRUCTURE OF THE EGG OF CALLIGRAPHA BIGSBYANA AT THE TIME OF DEPOSITION At the time of laying the eggs of Calligrapha bigsbyana are not always in the same stage of development, although usually polar body formation is taking place. The egg figured (Fig. 9) was fixed four hours after deposition. ‘The polar bodies have already been produced in this egg and the male and female nuclei are in the act of conjugation. ‘The egg consists of a large central mass of yolk and a comparatively thin peripheral layer of cytoplasm, the ““Keimhautblastem” of Weismann. The interdeutoplasmic spaces are filled with cytoplasm which is connected with the “Keimhautblastem” by delicate strands of the same material. The enormous amount of yolk contained in the eggs of these insects makes the identification of other substances extremely difficult. The yolk-globules range in size from large deutoplasmic spheres to small granules, and, as the dissolution of some of them is continually taking place, one is unable to determine where yolk ends and cytoplasm begins. The only accumulations of cyto- plasm large enough for examination are those surrounding the nuclei ae the oll mass, and the peripheral layer, the ‘‘ Keim- hautblastem.” No differences in composition or staining qualities were observed between the cytoplasm of these two regions. The ‘““Keimhautblastem” consists of a fluid ground substance in which are suspended very fine granules. It is a homogeneous layer of cytoplasm everywhere except at the posterior end of the egg. At this point there is a disc-shaped mass of larger granules imbedded within the inner portion of it. These granules stain deeply with haematoxylin. They are easily seen not only in sections but also i Centrifugal Force upon Beetles’ Eggs 517 in eggs that have been properly stained in toto. Because of their ultimate fate, as explained in the introduction, I have called these granules the germ-cell determinants (Fig. 9, g c. d). VII THE EFFECTS OF CENTRIFUGAL FORCE UPON EGGS CENTRI- FUGED AFTER DEPOSITION Table I presents in concise form the main points in the series of experiments which have been selected for detailed description. Besides the thirteen series noted here there are also two tables (XI and XII) which give the results of a number of other series of experiments which were not considered of sufficient importance to describe at length. TABLE I List of the experiments described in detail | | | Age when centri-| Length of time | . ; | | Orientation Name Number of series | fuged | centrifuged C. bigsbyana | C.B.4. ° | 1§ min. to 4 hrs.) post. end in ss CB | ° | 4 hours ant. end in . C.B. 10 ° 6 hours side in : C.B.9 ° 6 hours post. end in Y. : C.B.2 14 hours 1 hour Kg " C.B. 5 21 hours 2 hours ‘ C. multipunctata C.M.1 fo) 16 hours | - C. lunata Cilia 1 hour 12 hours is e CEs g hours 12 hours ant. end in L. decemlineata EDS x 2 hours 5 min. to 2% hrs| post. end in : D3 ° | 5 days | ‘ i | L.D. 2 ° | 7 days * Lema trilineata ys bg t | ce) | 2 hours H Series C.B. 4—Table II The eggs of this series were centrifuged as soon as laid. They were held in place with their posterior ends towards the axis of rotation. “Two eggs were removed at the intervals indicated in Table II; one of these two was fixed immediately, the other was allowed to develop. If the latter did not hatch within a period several days longer than the normal hatching time it was fixed. 518 R. W. Hegner TABLE II Calligrapha bigsbyana—Series C.B. 4 | Interval between) Number of |Age when cen-Length of timeend of experi- J : experiment trifuged | centrifuged ment and fixa- Onestsaas | Renae | tion C.B. 4,4 Control | C.B. 4,5 c | I5 minutes fe) C.B. 4,¢ |. ° 30 minutes ° C.B. 4,d ° | 1 hour ° C.B.4,e ° 2 hours ° Posterior end C.B.4.f | ° 4 hours ° | toward axis of CB.4,2 | ° 15 minutes 6 days rotation Normal larva CByaik || fo) 30 minutes 6 days | Normal larva CiBe4y 2 | ° 1 hour 10 days Did not hatch CBuayf. || ° 2 hours 10 days Did not hatch C.B.4,k ° 4hours | todays | | Did not hatch The progressive effect of a centrifugal force upon the distribution of the contents of the egg is shown by these experiments. They also furnish data concerning the amount of disturbance neces- sary to prevent the hatching of a normal larva. C.B. 4,a. Sections of the fresh control egg of this series show a condition similar to that illustrated in Fig. 9. C.B. 4, b. An egg centrifuged for fifteen minutes is very slightly affected. The “Keimhautblastem” has apparently not been changed at all. The yolk shows a partial redistribution; the larger, heavier globules have begun to move toward the outer end of the egg, z.e., the end away from the axis of rotation, and the inner portion of the yolk mass is almost entirely composed of the smaller globules. ‘The pole-disc occupies its normal position at the posterior end of the egg; all about it are small, irregular vesic- ular spaces which are no doubt caused by the accumulation of the lighter fats in this region. No polar bodies could be discovered in the sections of this egg, but no significance can be attached to this fact as they cannot always be found in normal eggs. C.B 4, c. The effects of centrifugal force applied to this egg for thirty minutes are similar to those just recorded for C. B. 4, 6.5 the changes however are more pronounced. We findthat there are Centrifugal Force upon Beetles’ Eggs 519 more of the large yolk-globules near the outer end and less of them at the other pole. ‘There is also a slight thickening of the “ Keim- hautblastem” at the sides of the egg near the innerend. ‘The pole-disc is present in its usual position, but it is surrounded by a greater number and larger, irregular vesicles than in C.B. a, 0: C.B. 4, d. An egg taken from the centrifugal machine at the end of an hour is definitely stratified, two distinct layers being visible. ‘There is a small cap of orange-colored material situated at the extreme inner end, while the rest of the egg representing the other layer has changed in color because of the redistribution of the yolk. The intense yellow color of the outer end is dug to the invasion of a vast number of large deutoplasmic spheres into that region. No definite layers can be distinguished in this large por- tion, since the change in color from bright yellow at the outer end to pale yellow at the inner end is gradual. A longitudinal section. through this egg is shown in Fig. to. Most of the large yolk- globules lie in the outer region; the interdeutoplasmic spaces are entirely free from the cytoplasm which usually fills them. The “Keimhautblastem” has been forced almost entirely away from the outer end and from the periphery of the outer third of the egg and has added its mass to that of the inner region. At the extreme inner end one large, bud-like protrusion (one-third the short diameter of the egg) and several smaller ones have formed. ‘They are covered externally by a thin layer of “ Keimhautblastem” and are composed of a great number of vesicles. A similar vesicular portion was noted in the two eggs described above (C.B. 4, b and c), but it has in this egg reached such proportions that we shall hereafter call it the vesicular zone. ‘This is the material ts appeared bright orange in color in the living centrifuged eg Only one Be cleus pauld be discovered in the entire egg. This Ts situated at one side near the inner end, as shown in Fig. 10, 7. The pole-disc has moved from its position at the end and has traveled en masse away from the axis of rotation. It has carried that portion of the “ Keimhautblastem” in which it is suspended along with it, producing a distinct depression at oneside of the inner end of the egg. ‘The sections containing the pole-disc fell outside 520 R. W. Hegner of the vesicular zone so that a figure has been introduced to show the change in position of this structure (Fig. 11, g c. d). C.B. 4, e. A third layer makes its appearance if a fresh egg is centrifuged for two hours. Before fixation this appears as a cs EN] SP As 2 w—-—- ---— ------V. 14, Fic. 10 Longitudinal section through egg C.B. 4, d. Fic. 11. Longitudinal section through the posterior end of egg C.D. 4, d, showing the effects of a centrifugal force applied for one hour upon the position of the pole-disc (gc.d.). Fic. 12 Longitudinal section through the posterior end of egg C.B. 4, e, showing the effects of a centrifugal force applied for two hours upon the position of the pole-disc. Fic. 13 Longitudinal section through the posterior end of egg C.B. 4, f, showing the effects of a centrifugal applied for four hours upon the position of the pole-disc. gc.d. = germ-cell determinants (pole-disc). khb/. = ‘‘Keimhautblastem.” #.= nucleus. p. = posterior. pt. = pathway made by the outward movement of the pole-disc. v. = ventral. v.z.= vesicular zone. y. = yolk. Fic.14 ‘Transverse section through egg C.B. Io, c. Centrifugal Force upon Beetles’ Eggs | 521 colorless cap at the outer end. ‘The sections show it to consist of a small mass of gray material which is heavier than the large yolk-globules. I shall call this the gray cap. It is composed of very small granules, does not stain like yolk nor as intensely as the cytoplasm of the ‘“Keimhautblastem.” The pole-disc has moved forward during the second hour the egg was centrifuged, and now lies anterior to its original position about one-fourth of the total length of the egg. In its progress it has pushed its way forcibly through the yolk mass, leaving a long, narrow, open path- way behind it (Fig. 12, pt). No nuclei were found in the sec- tions. C.B. 4, 7. This egg was centrifuged for four hours and then fixed. A surface view of the egg stained in toto revealed a large central, colorless bud at the posterior (inner) end surrounded by a number (at least seven) of smaller buds. These are produced by wrinkles or folds in the surface of this region, due either to poor fixation or to a decrease in turgidity at the inner end. ‘The entire egg seems to have been shortened slightly antero-posteriorly by the continued application of centrifugal force. The longitudinal sec- tions made of this egg are not perfect, certain portions of the outer end being lost because of the accumulation of large yolk-globules (yolk which is not imbedded in cytoplasm is liable to break on the knife in cutting). I cannot be positive, therefore, of the presence of a gray cap in this egg. ‘There is little doubt, however, that this structure was not absent in this instance, since the other eggs similarly treated possess a gray cap. ‘The vesicular zone has increased in size during the last two hours this egg was centrifuged, and has been folded into larger bud-like prominences than were noted in the last egg described (C.B. 4, e). The pole- -disc has made further progress in its journey away from the inner end. It has now reached a point about one-third of the total length of the egg anterior to its original position (Fig. 13, gc. d). The open pathway which was observed behind it in C.B. 4, e, has become closed and the ‘‘ Keimhautblastem” that was pulled in with it has passed back and taken part in the vesicular layer. C.B. 4, g. A normal larva hatched from an egg centrifuged for fifteen minutes with the posterior end towards the axis of rota- §22, i HP; Hegner tion. ‘The hatching period of approximately six days is the nor- mal one for eggs of this species. C.B. 4, h. ‘This egg, which was centrifuged for thirty minutes, also developed normally, the larva hatching in six days. C.B. 4,1, 7 and k. None of these eggs continued its develop- ment farther than the early cleavage stages. TABLE III Calligrapha bigsbyana—Series C.B. 3 | Interval | Numberof |Age when cen-Length of timebetween end of \ , | experiment trifuged | centrifuged experiment and| epee ite Be fixation | C.B.3,a Control | C.B. 3,6 ° 4 hours ° | Anterior end CiBaa,c ° 4 hours 37 hours- toward axis of | Did not develop C.B.3,d fo) 4hours | 6rhours | rotation | a CB. 33e ° | 4 hours 96 hours 3 CBs); ° 4 hours 63 days % Series C.B. 3—Table III Series C.B. 3 will serve to show the effects of a centrifugal force applied for four hours to fresh eggs with their anterior ends toward the axis of rotation. C.B. 3, a. The control egg was normal and in a stage slightly younger than that shown in Fig. 9. C.B. 3, b. When taken from the centrifugal machine at the end of four hours this egg appeared stratified ina manner exactly like that of C.B. 4, e. Longitudinal sections show a gray cap at the heavy outer end, a middle zone of yolk and an inner light vesicular zone. The distribution of the “Keimhautblastem”’ is also similar to that in an egg centrifuged with the opposite (pos- terior) end toward the axis of rotation, 7.¢., it has moved toward the lighter end of the egg. The inner pole is creased and folded as inC.B. 4, e. One nucleus is present near the vesicular zone. ‘The pole-disc remained at the posterior end of the egg near one side; OS ee ee ee ee Centrifugal Force upon Beetles’ Eggs 523 it fell outside of those sections containing parts of the gray cap so that I was unable to determine whether it is of greater or less specific gravity than the later substance. GiB. 3;t5d, eand j.1- These eggs did. not develop: very-far, although the youngest (3, c) contained a number of nuclei in the course of disintegration. Sections of the other eggs (3, d, e and 7) show a further dissolution of the nuclei, the vacuolation of the “ Keimhautlbastem”’ and other evidences of catabolism. TABLE IV Calligrapha bigsbyana—Series C.B. 10 Interval | Numberof Age when cen-Length of timebetween end of | P ; experiment trifuged centrifuged experiment and) Pe ageleu st | oo | fixation | C.B. 10, a Control C.B. 10, b ° 4 hours ° Right side in* CGB: 10),.¢ fo) 4 hours ° Ventral side in C.B. 10, d ° 4 hours 36hours | Right side in C.B. 10, e ° 4 hours 48hours | Right sidein | C.B. 10, f ° 4 hours 60 hours Ventral side in C.B. 10, g ° | 4 hours g days Right side in fc) C.B. 10, h 4 hours 9 days Ventral side in * This means that the right side of the egg was placed toward the axis of rotation. Series C.B. 10—Table IV These experiments were undertaken to detemine if the position of the embryo upon the egg can be changed by altering the distri- bution of the cytoplasm. Four of the eggs were oriented in the centrifugal machine so that their right sides were toward the axis of rotation, the other three with their ventral surfaces toward the center. C.B. 10, a. The control egg was in an early cleavage stage. C.B. ro, b and c. No differences could be discovered between an ege centrifuged with its right side turned inward and one with its ventral surface in the same direction, either before or after 524 R. W. Hegner fixation. ‘The sections also show a similar arrangement of mate- rials. Fig. 14 represents a transverse section through C.B. 10, c. There is a light vesicular zone at the side which was turned toward the axis of rotation; this is folded into bud-like prominences just as we found to be the case in C.B. 4, e, and others. The yolk- globules are distributed in the usual manner, the large ones being on the heavy side. The “Keimhautblastem” has moved away from the side of greater specific gravity and toward the lighter side. No gray cap could be found. It is probable that the material which produces this zone has all been thrown to the outer side, but the area is too great to allow of any perceptible accumula- tion. . C.B. 10, e,f7, g and h. No one of these eggs developed beyond an early cleavage stage. The nuclei then disintegrated and the amceboid masses of cytoplasm in which they lay became vacuo- lated as did also the ‘“‘ Keimhautblastem.” TABLE V Calligrapha bigsbyana—Series C.B. 9 | | Interval Number of Age when cen-Length of time between end of : é experiment | trifuged centrifuged experiment and) Deas Reale | fixation C.B. 9,4 | Control CBj9;0n ° 6hours ° C.B. 9, ¢ | ° 6 hours | 41 hours Posterior end C.B. 9, d ° 6 hours | 65 hours toward axis of C.B. 9, e ° 6hours | 8g hours rotation C.B. 9, f ° 6hours | 10 days Series C.B. g—Table V Although these eggs were taken as soon as laid, sections through C.B. 9, a show that they were in a rather advanced cleavage stage when the experiments were begun. They represent a condition intermediate between those of series C.B. 2 (Fig. 17) and C.B. 4 (Fig. 9), resembling in structure an egg ten hours old. Centrifugal Force upon Beetles’ Eggs 525 C.B. 9, b. Three zones are recognizable in this egg corre- sponding to those already described in egg C.B. 4, e and, although centrifuged for six hours, no noticeable difference is discernible in the distribution of material in this egg and one of nearly the same age which was centrifuged for only one hour (C.B. 2, 5). The nuclei of many of the vitellophags are distorted or disintegrating. The granules of the pole-disc have, as in normal eggs, become imbedded in the cytoplasm of the primordial germ cells; the latter occupy their usual position at this stage between the vitelline mem- brane and the blastoderm at the posterior pole. C.B. 9,c. ‘Two eggs were fixed forty-one hours after they were taken from the centrifugal machine. One of these did not develop, its nuclei disintegrating and the “Keimhautblastem” becoming vacuolated; the other carried an embryo with a distinct ventral groove (Fig. 15). Superficially this embryo resembles that of a normally developed egg of this age except that it does not reach as far anteriorly on the ventral surface, but extends farther around the posterior end and up on the dorsal surface (compare Figs. 4 and 15). It is evident that under the influence of centrifugal force the nuclei and “ Keimhautblastem” have become massed in the posterior half of the egg, where development has continued. This egg if it had been allowed to develop would no doubt have produced an embryo resembling that described under C.M. 1, 6 (Fig. 21). Sections of this egg show a rearrangement of the yolk- globules, a condition being reached similar to that illustrated in Fig. 9. The gray cap and vesicular zone are still present. C.B 9, d. One of two eggs preserved sixty-five hours after being taken from the centrifugal machine did not develop; the other produced a shapeless mass of tissue, no definite organs being distinguishable. Fig. 16 is a diagram of a sagittal section through this egg. The gray cap and vesicular zone are still present, the former at one side of the outer end of the egg, the latter just dorsal to the embryonic tissue. R. W. Hegner 526 ‘sginuvis 3xfok = S€ sxok = *€ *yx Suuatzo ur parnfur sem 33a aay ovjINs [eUAA IOLAJUe = “yx ‘adRyINS [es -Iop uo pueq-uria3 Jo pua ro11aysod = *w ‘auoz IvpNIISVA = “za “sBeydoyayA = ‘7a *aAooIS [eJUGA = “Fa ‘sorayds yfoA Tes = *s-K+s *s]]99-W1I93 [eIp round = -98-¢ “tajonu orumsapoyseyqaid = *uyq¢ ‘ro1sajsod = *¢ ‘saroyds yok a8aey = ssc mayseTqinequney,, = "7q7y ‘syueUTULIajp ]]20-WHI93 = "ps ‘deo Avis = 93 = spueq-ues = qs -anssy ouoAIquia = “7a “wepoyseyq = "7q ‘“IOUAjUe =‘ ‘Lr “Big Jo advys ayy ur 389 ue Jo syUa]UOD ay} JO woRNNqrystp ay} uodn anoy uO 410j paydde ao10y peSnyryyUId e& Jo sjaya oy Surmoys “q ‘z *g’D 389 Ysnosy) UONIes [eUIPNUOT gi ‘OIg ‘uontsodap 10342 smoy wsa}inoJ dstp-ajod pur jpnu ,“wayse[qineyWIoy,, sy} Jo UONNGINsIp [euOU oY) SuIMoys ‘y ‘z *g*d 33a YSnoryy UONIas [euIpNysuOT LI ‘org *pua ro1193sod ay} ye anssyy Jo sseur ssajadeys ev Surmoys ‘p “6 *g'D 33a YsNoIG) UONIes [eUIpNyISuOT g1 ‘OI b3ry yymoredmog +9 6"g7a 83a Jo aovjinsyeruaA S191 ‘Ol ‘Gl at Centrifugal Force upon Beetles’ Eggs 527 TABLE VI | / Interval Numberof (8° yes See a Me between end of : pee eeien | trifuged centrifuged eee aad Orientation | Remarks fixation C2B.2; a control Posterior end C.B. 2,5 14 hours 1 hour ° toward axis of CBi2, ¢ 14 hours 1 hour | 48hours | rotation normal embryo C.B.2,d 14 hours 1 hour | 6 days | normal larva Series C.B. 2—Table VI The eggs used in these experiments were laid at 7 p.m. on July 19. One egg was fixed at the end of fourteen hours; the others were at the same time placed in the centrifugal machine with their posterior ends toward the axis of rotation and subjected to the usual number of revolutions for one hour. C.B. 2, a. Fig. 17 shows the “‘Keimhautblastem,”’ the pole- disc and the distribution of the nuclei in the control egg, aged four- teen hours. ‘The yolk is not included in this figure, as its distribu- tion is similar to that of the freshly laid egg (Fig. 9). The two groups of nuclei, those which form a more or less regular layer near the periphery and will fuse with the “‘ Keimhautblas- tem”’ in a few hours producing the blastoderm (pd/. n), and the vitellophags (vt) scattered about in the yolk, are quite clearly marked at this stage. When taken from the centrifugal machine a colorless layer of material was observed at the outer end of the egg; this is the gray cap occupying a position similar to that noted under C.B. 4, e. The color of the egg was deep yellow posterior to the gray cap and gradually faded out toward the inner end until near that pole it was almost colorless. A bright-yellow cap, the vesicular layer, occupied the extreme inner end. A sagittal sec- tion of this egg is shown in Fig. 18. At the anterior end is the heaviest substance in the egg, the gray cap. Just posterior to this we find the largest deutoplasmic spheres which gave to the living egg its bright-yellow color. The spaces among these are 528 R. W. Hegner free from cytoplasm. ‘The yolk-globules become smaller and smaller posteriorly until they cease altogether in the middle region, where smaller and lighter yolk granules take their place. At the posterior end there are many irregular vacuoles caused by the accumulation of fat in this region. During the hour the egg was under the influence of centrifugal force the preblastodermic nuclei (Fig. 17 pb/. n) migrated outward until they fused with the “Keimhautblastem” forming the blastoderm. ‘The “ Keimhaut- blastem”’ in the mean time flowed away from the anterior end of the egg, adding this portion to that posterior to it and producing a blastoderm in the latter region decidedly thicker than usual. The nuclei in the blastoderm seem to have been influenced by the centrifugal force; those near the central region have apparently been drawn out of their normal spherical shape and are now oval. As the inner pole is approached the nuclei become less and less oval until at theextreme end theyare sphericalas normally. ‘The vitellophags have migrated toward the axis of rotation and the outer end is free from them altogether, while a greater number than usual are present near the posterior end. ‘The centrifugal force used has apparently had no effect upon the position of the nuclei of the vitellophags in relation to the mass of cytoplasm which surrounds them, as in every case the nucleus is in or near the center. The direction of division of these vitellophags, how- ever, seems to have been influenced for we find in almost every instance that the daughter cells produced by a recent division lie one posterior to the other, z.e., in the direction of the centrifugal force. The germ-cell determinants have found their way as usual into the primordial germ-cells at the extreme posterior end of the egg (Fig. 18, p. gc). C.B. 2, c. A normal embryo (Fig. 7) was produced by this egg, which was fixed forty-eight hours after being centrifuged. Not the slightest difference could be discovered between an in toto preparation of this egg and a normally developed egg of the same age (63 hours). Sagittal sections show that the yolk has under- gone segmentation and that the yolk-spheres and yolk-granules are equally distributed throughout the entire yolk mass. ‘The germ- cells have migrated from the posterior amniotic cavity through the Centrifugal Force upon Beetles’ Eggs 529 pole-cell canal and into the embryo and lie near the end of the tail- fold. ‘The gray cap has not entirely disappeared, but what [ take to be a remnant of it is situated at the dorsal anterior surface. C’.B. 2,d. A normal larva hatched from this egg in the average length of time required for eggs of this beetle when developed under natural conditions. TABLE VII Calligrapha bigsbyana —Series C.B. 5 | Interval Numberof Age when cen-\Length of timejbetween end of . : experiment | __ trifuged centrifuged experiment and SOAR PES fixation | C.B. 5,4 | control | | Posterior end | (Calbh fal | 21 hours 2 hours ° toward axis of Cab ge |) 21 hours 2 hours 27 hours rotation normal embryo @B45,d | 21hours 2hours | 6 days normal larva ! Series C.B. 5—Table VII This series of experiments was performed in order to discover if centrifugal force would have any appreciable effect on an egg in which the blastoderm has already been formed, and if so whether or not the egg would at this late stage continue to develop and eventually produce a larva. C.B. 5,a. The eggs of C. bigsbyana at the age of twenty-one hours have usually reached a stage in which a blastoderm of a single layer of cells completely covers the central yolk mass. Scattered about irregularly among the yolk-globules are numerous vitellophags. At the posterior pole are a number of cells lying in a closely packed group between the vitelline membrane and the egg (Fig. 3, pgc.); these are the primordial germ-cells (pole-cells) which a few hours earlier migrated through that part of the pos- terior end occupied by the pole-disc, taking the granules of which this is composed along with them. C.B. 5, b. The application of centrifugal force for two hours has very little effect upon an egg twenty-one hours old as seen in 530 R. W. Hegner surface view. he surface at the inner end is creased and folded just as was found to be the case with younger eggs (C.B. 4, e). Longitudinal sections through this egg present a distribution of material similar to that with which we are already familiar. A gray cap is present at the outer end (Fig. 19, g. c); the largest deutoplasmic spheres are adjacent to the gray cap, and there is a gradual decrease in the size of the yolk-globules until near the inner end where these are lacking altogether giving way to the vesicular zone. Most of the vittellophags have passed into the Fic. 19 Longitudinal section through egg C.B. 5, b., showing the effects of a centrifugal force applied for two hours upon an egg covered by a blastoderm. Explanation of letters same as Figs. 15-18. vt.a, = vitellophags which have fused with the blastoderm. inner half of the egg; a number of them seem to have fused with the “‘Keimhautblastem”’ in the equatorial region. As was the case with the “‘ Keimhautblastem”’ in the younger eggs (C.B. 4, e) the superficial layer (blastoderm) has become thinner at the outer heavy end until it is barely visible at certain points; its mass has been added to that toward the inner end. ‘The primordial germ- cells occupy their normal position at the posterior end of the egg. C.B. 5, c. An egg in the condition just described was allowed ——_— = Centrifugal Force upon Beetles’ Eggs 531 to develop for twenty-seven hours and then preserved. [xter- nally the embryo it carried appeared to be normal in every respect. It was in a slightly younger stage than that of C.B. 2, c, shownin Fig. 7. Upon sectioning it was found that the vesicular zone had disappeared entirely, that the yolk had segmented and both this and the vitellophags had regained their normal distribution, but that there still remained a small amount of the heavy gray cap. The embryonic tissue seems to have sustained no ill effects from the centrifugal force. C.B. 5,d. A normal larva hatched from the remaining egg of this series in the average period, six days. Senes (C.M. 1. Two freshly laid eggs of Calligrapha multipunctata placed with their posterior ends inward were centrifuged for sixteen hours at a rate much slower than that applied to the eggs in most of the other experiments. At the end of this period three perfectly distinct zones could be recognized by their colors. “Vhe nearly uniform pale-orange color of the normal egg had given way at the inner end to bright orange; at the opposite pole was a whitish cap, while the comparatively large central zone faded gradually from bright yellow at its outer end to pale yellow where it joined the inner orange stratum. C.M. 1, a. An in toto preparation of one of these eggs which was fixed immediately after being taken from the centrifugal machine shows that the three zones do not differ in color only, but are composed of three different substances. Sections of this egg show a stratification similar to that already described for C. bigs- byana (C.B. 4, e). The stage of development, however, is unlike that of any egg so far examined. Fig. 20 shows the nuclei aggre- gated in the inner portion of the egg. The “Keimhautblastem” at the sides of the egg and surrounding the folded vesicular zone contains many nuclei producing a kind of blastoderm. ‘The vitellophags have accumulated in the inner portion of the central zone. Many of these are either dividing by amitosis or seem to have recently completed such a division. R. W. Hegner 532 *ypod =f ‘*9u0Z Av[NIISVA = “20 ‘ssvydojpaqtA St *[eajueA ee “HINYpOUIOys = iS ‘umapoyaoid yd ‘ok1quia jo pua roysaysod = ‘ad = 101103s0d = -d ‘deo vx3 = 93 anssy st0kiqma = ‘79 "[esIop = “Pp “Waporseyq = “14 ‘a3epuadde = “dv ‘okiquia JO pa IOWeyUe = “37 *yjod aya Jo apisyno padoyaaap eavy yoy sofiquia SurMoys "J'Q""T Seeg Wor} e932 JO SMTA aDvJANS apIg £7 pue 77 “SOY “pua ro1zaysod ay} ye HOA ay} Jo apisyno ofrquia JIeMp & Jo yuaudoTaAap ey Surmoys “q *['W'D 399 Jo Mats vovjns [eryey 12 “OM ‘v ‘LW’ 332 Ysnosryy Woes JeUIpNsuoT OF “ONT "SS 1G 0G Centrifugal Force upon Beetles’ Eggs 533 C.M. 1,b. Fig. 21 represents a surface view of the right side of an egg like that just described which was allowed to develop for nine days. The embryo has continued to develop at the inner (posterior) end. Its orientation is normal except that the entire embryo has shifted its position posteriorly so that the posterior end instead of being coincident with the posterior end of the egg is now part way up on the dorsal surface. A small mass of embry- onic tissue is imbedded in the large mass of yolk. Normally this yolk would be surrounded by the embryo and become. included within the mid-intestine; in this case a dwarf embryo has developed without growing around the nutritive material. TABLE VIII Calligrapha lunata—Series C.L., a Interval | Number of [Age when cen-Length of time between end of ; : | : ; : ; | Onientation | Remarks experiment | trifuged centrifuged experiment and | fixation Ciivayt 1 hour | 12 hours ° Posterior end C.L. a, 2 . — 55 hours toward axis of Cay 3 ns | 79 hours rotation { 24 days | hatched in 6 days C.L. a. 4 “ \ | 18 day larva | | series ©. LL. a—Vable VIII The effects of centrifugal force upon the eggs of C. lunata are shown by this series of experiments. ‘The results, as may be seen from a comparision of the above table and the following descrip- tions, differ only in minor details from those recorded for eggs of C. bigsbyana similarly treated. C.L. a, 1. This egg was stratified by the centrifugal force into three layers, a gray cap at the outer heavy end, a middle yolk zone with large deutoplasmic spheres at the outer end gradually decreasing in size toward the inner pole and a light vesicular layer at the extreme inner end. Longitudinal sections resemble those of C. M. 1 shown in Fig. 20. There are a number of nuclei present scattered about among the yolk-granules near the inner end of the middle zone. Each nucleus is approximately in the 534. R. W. Hegner center of the amceboid mass of cytoplasm in which it lies embedded; the whole apparently has moved en masse toward the lighter end of the egg. The pole-disc is situated between the vesicular layer and the middle zone; it is probable that its change of position is due, not to any movement of the granules, but to the accumulation of lighter fats posterior to it. C.L. a, 2. The only redistribution of material that has taken place since this egg was taken from the centrifugal machine is a movement of the “Keimhautblastem,” resulting in several large accumulations at the periphery in the middle region. ‘The nuclei have disintegrated and the “Keimhautblastem” has the vacuolated appearance indicative of its early dissolution. No larva could possibly have developed from this egg. C.L. a, 3. Sections of this egg show a continuation of the cata- bolic processes mentioned in C. L. a, 2. C.L. a, 4. The only egg which was not fixed before the end of the hatching period seems to have developed normally, as it produced a normal larva. I can account for this only on the assumption that the eggs of this series were differently affected by the centrifugal force and that C.L. a, 2 and C.L. a, 3 were too severely injured to continue their development while C.L. a, 4 was able to readjust itself to the new conditions imposed by the change in the position of the egg contents. A perfect series of sagittal sections was made through this larva; they showed no irregularices in the size, position or structure of the internal organs. ‘The reproductive organs (female) are in their proper positions. TABLE IX Calligrapha lunata—Series C.L.I. Interval T | . | ae of | Age when cen-| Length of time ee end of Gueniaeae Rewurke experiment | _ trifuged centrifuged experiment | and fixation Callas control Anterior end (CHES 9 hours 12 hours ° toward axis of Chile # | : 24 hours rotation Cian rs | te ' 48 hours Calne fs 4 days hatched - Centrifugal Force upon Beetles’ Eggs 535 Series C.L. 1—Table IX The effects of centrifugal force upon the eggs of C. lunata when oriented with their anterior ends toward the center are shown by these experiments. C.L. 1, a. The control egg of this series proved to be in a stage similar to that already described for C. B. 9, a. C.L. 1, 6. After being centrifuged for twelve hours this egg showed the customary three strata. Longitudinal sections resem- ble those of C. M. 1 (Fig. 20); they differ from them only in the absence of a well-defined blastoderm in the inner region. C.L. 1, c. During the twenty-four hours since this egg was taken from the centrifugal machine the yolk has had time to redistribute itself to some extent and many of the larger globules are present at the lighter end. Development has proceeded and the inner half of the egg is one large syncytium in the center of which is the vesicular zone containing a few nuclei. C.L. 1, d. Sections of this egg may be compared with that of C.B. 9, d, shown in Fig. 16. There is a mass of tissue at the inner end which is thrown up into folds, but no definite structures are distinguishable in it. C.L. 1, e. The only egg that was allowed to develop through- out the entire hatching period produced a larva at the end of six days. This larva is apparently normal. It was preserved when three days old. TABLE X Leptinotarsa decemlineata—Series L. D. 1 | Interval | | Borer of | Age eee cen- |Length of time | between end of OrentaGon | Resacke experiment| _trifuged centrifuged experiment and fixation IFAD Bae | control L-D:1,2 2 hours 5 minutes Posterior end hatched in 6 days IG AD BIg si 10 minutes toward axis of $ Dil, 4 . 20 minutes rotation | a L.D.1, 5 ‘ 45 minutes | Dal, 6 4 14 hours | | - 1 DIG) us 2% hours | | . 536 R. W. Hegner Series L.D. /.—Table X The above table (Table X) shows the results of a graded series of experiments upon eggs of Leptinotarsa decemlineata centri- fuged from five minutes to two hours and one half. ‘These eggs, including the control (L. D./, 1) all hatched at the same time, showing that the amount of centrifugal force has no perceptible influence upon the rate of development. Series L.D. 1 L.D.1. A number of fresh eggs of the potato beetle, Leptino- tarsa decemlineata, were centrifuged at a low rate of speed (360 revolutions per minute) for five days. ‘They were oriented with their posterior ends toward the axis of rotation. ‘The resulting embryos (Figs. 22 and 23), which of course would not have hatched, are very similar in appearance to that described under C. M. 1, 6. The heavy substances in these eggs are apparently non-essential for the development of the embryo, being made up principally of nutritive yolk. When deprived of this material a dwarf embryo is produced at the inner end of the egg. Series L.D. 2 Another batch of potato beetles’ eggs were centrifuged at the same rate of speed for seven days. Dwarf embryos developed at the inner light end in every case. No sections were made of these embryos. Series L. T. 1 A number of eggs of Lema trilineata were centrifuged with their posterior ends turned inward. In all cases the stratification induced resembles that of the eggs of C. bigsbyana similarly treated. Table XI presents the data obtained from a number of experi- ments which have been selected from fifteen series of the eggs of Calligrapha lunata. Eight of these centrifuged eggs produced larve in the normal hatching period; of these, four were centrifuged with their posterior poles toward the axis of rotation, three with their anterior ends toward the center and one with its side turned Centrifugal Force upon Beetles’ Eggs 537 TABLE XI Calligrapha lunata | Interval | Number of | Age when cen-| Length of time between end of Gieikaces Rts experiment | trifuged centrifuged | experiment and fixation =r Aae SRG SSS ee Cid, 3 ° 23 hours | post. pole in | hatched in 6 days COE PHS co) thour —8days < ee oes | speed CLxe,2 4 hour 3 hours 7 days x did not hatch C.L. a, 4. 1 hour 12 hours s hatched in 6 days Caliy732 | 1 hour | 13 hours 7 days K | did not hatch Cle7; fi Shours 13 hours I1rdays | ~ C.L.4,3. | 24hours | 2hours | S | hatched in 6 days Cx; 5 47 hours 6 hours 7 days | 2 | did not hatch C.L.s,2 | sohours 2 hours 3 hatched in 6 days Cig 2) || ° 2 hours 7 days ant. pole in “did not hatch Cnn ,2, 3 hour | 4 hour CS hatched in 6 days C.L.i,e | 9 hours 12 hours - | : C.L.i,2 -| 14hours 2 hours 7 days . did not hatch CEyk, 2) |) t4ihours 2hours — @ hatched in 6 days | | [hatched in 6 days C.L.b, 2 1 hour 20 minutes on one side ; | | high speed TABLE XII Leptinotarsa decemlineata | Interval Number of | Age when cen- Length of time between end of Oneteae ul Resaie experiment trifuged centrifuged | experiment | | and fixation PD cy 5 3 hour 2 hours 7 days post. polein | did not hatch L.D.c, 6 rf n | 10 days * : L.D.e, 4 4 hour 4 hours 6 days ‘ : TD ¢, 5 “ a 8 days 2 L.D. d, 5 1 hour 13 hours e hatched in 6 days L.D. }, 5 2} hours 1 hour - f L.D.k, 5 | 24 hours 2 hours = $ \L.D. n, 6 24 hours 3 hour ~ : eDaeS. | ° 3 hours 7 days ant. pole in did not hatch L.D. i, 2 ° S mi hatched in 6 days Dts 7, || 24 hours 4 hour > | ‘ 538 R. W. Hegner inward. The age when the eggs were centrifuged ranges from freshly laid to fifty hours. ‘The length of time centrifuged ranges from twenty minutes to twelve hours. It is obvious that there is no definite total amount of centrifugal force which will prevent the hatching of the egg. ‘The orientation of the egg is apparently of no importance. The data given in Table XII have been selected from eleven series of experiments upon the eggs of Leptinotarsa decemlineata. There are too few items in this list to warrant any general con- clusions, but the experiments tend to show that an older egg has greater chances of producing a larva after being centrifuged than does one experimented upon a short time after deposition. Both eggs oriented with the posterior end toward the axis of rotation and those with the anterior end toward the center gave rise to normal larvee. VIIL THE EFFECTS OF CENTRIFUGAL FORCE UPON EGGS LAID BY CENTRIFUGED BEETLES. I Experiments with C. bigsbyana Series C.B. 12 A female C. bigsbyana was centrifuged at the usual rate of speed for two hours and fifteen minutes with her posterior end toward the axis of rotation. When taken from the machine she seemed to suffer no ill effects but proceeded to walk about and feed as usual. Three days later, July 24, five eggs were laid; two of these were fixed at once and the other three allowed to develop. The former showed no outward signs of any disturbances due to cen- trifugal force. Sections also failed to disclose any rearrangement of materials. “The eggs that were left to develop were fixed at the end of eight days. A superficial view of one of these is shown in Fig. 24; a shapeless mass of tissue lies imbedded within the disinte- grated yolk mass. Senes CoB. 13 The same beetle as that of Series C. B. 12 laid a second batch of five eggs two hours after the first five were deposited. “Two of Centrifugal Force upon Beetles’ Eggs 539 these which were fixed immediately showed no effect of centrifugal force; the other three hatched in six days. 2 Experiments with Leptinotarsa decemlineata series LD. 7 At 3:30 p.m. on July 17 a female L. decemlineata was centri- fuged for one hour with her posterior end toward the center. Fic. 24 Surface view of egg C.B. 12 laid by acentrifuged beetle. ap. = appendage. e.t.= em- bryonic tissue. p. = posterior. s. = space between two yolk masses. y. = yolk. One egg was laid an hour after being centrifuged and others were laid at irregular intervals until g:30 the next morning, when the total number reached seventy. The first egg laid, as well as all of the others in the series, showed a stratification produced by centri- fugal force. Only two layers, however, resulted, no gray cap being discovered in the sections. The vesicular zone is not as large as in older eggs centrifuged outside of the body of the mother for a 540 R. W. Hegner similar length of time, but the yolk-globules have a distribution almost exactly like that induced in the latter. Many of the eggs were allowed to develop; all of these hatched in six days. Series L.D. m The same beetle as that of Series L.D. 7 laid a batch of eggs at 7 p.m. July 19, 7.e., fifty-one hours after she was centrifuged or thirty-three and one-half hours after the first lot were deposited. No effects of centrifugal force could be discovered in sections of these eggs. Normal larve hatched from those eggs which were not preserved. Series L.D. o A third batch of eggs were laid by the beetle of L.D. f at 1 p.m. July 20. The preserved eggs showed no effects of centrifugal force; the others hatched in six days. Series L. D. ¢ A fourth lot of eggs were laid by the same beetle as in L.D. 7 on July 22. These agreed in every respect with those described in Series L.D. m. IX REVIEW OF THE EFFECTS OF CENTRIFUGAL FORCE UPON DEVELOPING EGGS I Distribution of the Egg Contents The most noticeable result obtained by centrifugal force is the redistribution of the materials contained in the egg because of the differences in their specific gravities. A number of cases have been reported of eggs whose contents are normally visibly different and localized in particular regions. For example, Boveri (’01, a, p. 145, Fig. 1, and ’o1,5, Taf. 48 and 49, Figs. 6-22) found three hori- zontal zones present in both unfertilized and fertilized eggs of Strongylocentrotus lividus. These zones could still be recognized in young blastulz. Wilson (’o4, p. 68) states that, “The Denta- lium egg shows from the beginning three horizontal zones, an equatorial pigment-zone and two white polar areas. Each of the Centrifugal Force upon Beetles’ Eggs 541 polar areas includes a specially modified protoplasmic area proba- bly comparable to a polar ring.”’ Conklin (05, a, p. 211) says of the Ascidian egg (Cynthia): “All the principal organs of the larva in their definite positions and proportions are here marked out in the 2-cell stage by distinct kinds of protoplasm,” and again on p. 216 this author states that “the substances of the ectoderm, mesoderm and endoderm are recognizable in the unsegmented egg.” In another place (Conklin ’o5, 5, p. 220) we find the state- ment that, “Three of these substances are clearly distinguishable in the ovarian egg and [| do not doubt that even at this stage they are differentiated for particular ends.” Many other eggs that do not exhibit a normal stratification and are apparently homogeneous throughout take on a zone-like appearance under the influence of a strong centrifugal force. Morgan (06) found that when the unsegmented eggs of Rana sylvatica are revolved 1600 times per minute for seven minutes the pigment and yolk are driven to the top of the egg, leaving a clear polar field. Similar results were obtained in toads’ eggs in three minutes. Lyon (’06, ’07) was able to induce four layers in the egg of the sea-urchin, Arbacia. ‘Two layers were obtained in eggs of the starfish, Asterias (Lyon, ’07). The annelid, Che- topterus, exhibited three layers. “The same author also centri- fuged the eggs of the Ascidian, Cynthia, the Gephyrean, Phascolo- soma, and the common garden spider; the eggs of Cynthia and Phascolosoma were stratified into three layers, those of the spider into two. Lillie (06) found that not only in the unsegmented eggs of Chetopterus but also in the two, four and eight celled stages three zones appeared in each cell when placed under the influence of centrifugal force. ‘The contents of the egg of the mollusk, Cumingia, may be separated into three zones (Morgan, 08). The eggs of the rotifer Hydatina senta were centrifuged by Whitney (’o9) while still within the mother. Three distinct layers resulted: a pink zone, a clear middle zone and a gray zone. The eggs of Calligrapha bigsbyana are when laid of a nearly uniform pale-yellow color. When subjected to a strong centri- fugal force for a sufficient length of time three zones are distin- 542 R. W. Hegner guishable: (1) a bright-orange light zone at the inner end (the vesicular zone, Fig. 20, v. z), (2) a comparatively large central mass composed of yolk globules which are largest at the outer heavy end, gradually becoming smaller until indistinguishable from cytoplasm at the inner end, and (3) a colorless layer (the gray cap, Fig. 20, g. c) at the extreme heavy end. ‘These three zones are produced when the eggs are oriented either with their posterior (C.B. 4, e) or their anterior (C.B. 3, 5) ends toward the axis of rotation. When placed with their sides toward the center only two layers are induced, the vesicular zone and the yolk zone. Three layers may be obtained in fresh eggs (C.B. 4, ¢) in eggs which have reached a late cleavage stage (C.B. 2, b, Fig. 18) and in eggs which are covered by a blastoderm (C.B. 5, a, Fig. 19). The gray cap. The material of the gray cap is the heaviest of the egg contents. It is composed of very fine granules whose positions before being driven to the heavy end of the eggs could not be determined A fresh egg when centrifuged for one hour does not exhibit this layer (C.B. 4, d, Fig. 10). At the end of two hours, however, a distinct gray cap is present (C.B. 4, ¢). Eggs in late cleavage stages require a lesser amount of centrifugal force in order to produce this structure (C.B. 2, 6, Fig. 18). We conclude from this that either the gray cap material is liberated during develop- ment and the egg fourteen hours old (C.B. 2, 4) contains a greater quantity of it, or else some condition of the yolk mass at this age allows it to pass more rapidly toward the heavier end. Longitu- dinal sections through egg C.B. 2, c (Fig. 7) show that although the embryo has developed normally the material of the gray cap is still at the heavy end where it was driven by the centrifugal force. A like condition also exists in a slightly younger egg (C.B. 5, c). It is evident that the gray cap substance is not necessary for the normal development of the embryo. The vesicular zone. The light fats which probably produce the vesicular zone at the inner end of the egg collect very quickly under the influence of centrifugal force. An egg centrifuged for only fifteen minutes (C.B. 4, 6) has a small number of vesicular spaces near the pole-disc. Continued application of centrifugal force results ina greater number of these vesicles until at the end of Centrifugal Force upon Beetles’ Eggs 543 one hour a very distinct zone may be recognized (C.B. 4, d, Fig. 10). The surface of the egg in this region is in every case wrinkled and folded as though the volume had decreased at this end and the firm layer of “Keimhautblastem” had become pulled in (C.B. Bee ee: 103 B.4, 7, Fig. 13; C.Bs 10,7. Pig. ig; C.B. 1g, 30 ai ciliakstopsanm yameiinc ve sae 20 AuCilial sto pirtgerlteats tere teraersy a eratelste 25 AxCiWay SEO Dasterersteieetiaky tee ree 35 Gecilia Ist Pant ecctmer: verve tates 30 Seika stops iosiscts o-steisa ese antes 20 Grciliastopee meer SHOU CoG Eee 30 Gicilia stop: tes cies caso teere 30 7 Acilta, StO Pi. sic esielsiaeuste leis craieome ae 40 HERG ash oSocudD hbo nde qoeb cc 60 SyCiliarsto piss cssiarachstssterotasecseieveretes 20 Siteiliaysto potest rer ca teeters 35 Oucilia sto presmrerititerctserscr ences 105 Qiciliaxstop wis. accmeveele cw anesies 60 TO Ciliiat Stops. on dodudesin Sopakonmee 135 Average resistance = 172.5 Immunity of Lower Organisms to Ethyl Alcohol 589 Stentors of this type with a normal resistance averaging 15305 seconds show in a 0.5 per cent solution a gain in resistance which though slight is nevertheless constant and typical. In a 1 per cent medium the resistance is increased to 229.5 seconds. Al- though the experiment was selected as one giving low results for the type, nevertheless it denotes for the animals of the 1 per cent solution a substantial gain over animals both of the control and of the 0.5 per cent medium. But the point of most interest is that the two media, though differing but slightly in strength, show a corresponding difference in degree of immunity. From this it is seen that subjection to two media of different strengths for a definite time gives corre- sponding differences in the increase of immunity. Time as a Factor in the Degree of Immunity Produced. We have now established two points. ‘These are that at the end of four or five hours immunity normally begins and that at the end of the fourth day this is definite and considerable. The progress from the one to the other may now be noted. The difference in degree of adjustment due to the same medium at the end of a few hours and again at the end of a few days, shows that immunity increases as the period in the acclimatizing fluid is lengthened. ‘This is true, however, for only a limited time. In an acclimatizing medium of average strength the adjustment which as we saw comes on at the end of the first few hours has by the end of the first day become sufficiently clear not to be mistaken. As an example of this and its subsequent history a test series may be given, which gave the highest immunity found for the type. In this the animals had at the end of the first day a resistance of 166 seconds. ‘The resistance had on the third day increased to 249 seconds. On the fourth day a notable rise was shown in which the killing time for the acclimatized organisms reached a maximum of 334 seconds, the control at the same time showing a resistance of 160 seconds. By the fourth day in this and subsequent experiments it was found that immunity had reached a high degree of constancy. For this reason, in the following experiments on this type, the fourth day has been selected as the time at which tests requiring a maximum degree of immunity were made. 590 F. Frank Daniel The progiess of immunity from the first to the fourth day, while showing various fluctuations, increases with a considerable degree of constancy. ‘This may be shown in the following experiment, which is in general typical for others of the same series. In this the animals were reared in a I per cent medium and tested to 6 per cent alcohol on the first, second, third and fourth days of their acclimatization. Experiment X RESISTANCE OF STENTORS OF TYPE F To 6 PER CENT ALCOHOL AFTER DIFFERENT PERIODS IN I PER CENT ALCOHOL. A = ACCLIMATIZED ANIMALS. C = CONTROLS First Day Second Day Third Day Fourth Day (ee aS A Cc A Cc A Cc A c Seconds Seconds Seconds Seconds Seconds Seconds Seconds Seconds 220 eS 150 65 210 225 330 275 170 80 420 120 215 go 370 80 145 Hee 75 fee 435 75 aE, uo 170 =—-120 375 145 175 155 490 115 125 120 205 150 210 140 270 140 150 155 215 165 380 195 225 100 135 130 290 125 410 195 150 120 18 £0) 105 IIo 205 190 165 410 130 235 180 100 7o 330 175 300 135 180 80 300 180 190 240 240 145 Average resistance = 164 126.5 224 132.8 27 At) ORS 299.5 143 ‘The two extremes—the first and fourth days—are in this experi- ment especially typical. On the third day—as was seldom the case—almost as high a degree of immunity is represented as on the fourth. On this. day, as well as on the fourth day of the test example, the control also showed a considerable increase. In both of these cases, however, the degree of increase for the control does not approximate that attained by the acclimatized animals. It is thus seen that the same animals kept in a given percentage of acclimatizing medium give different degrees of immunity at different periods of time. - The variability which was seen to be a prominent feature in beginning immunity is nowhere seen more strikingly than in a study of the maximum degree of immunity. Immunity of Lower Organisms to Ethyl Alcohol 591 Type F illustrates this in a clear way. Animals of the same strain kept under similar external conditions and for a like period of time, varied at different times in their reactions to the same concentration. ‘Thus type F, kept in a 1 per cent medium for four days, gave at different times average resistance periods to 6 per cent alcohol of 229.5, 301, 299.5 and 334 seconds—a differ- ence of more than 100 seconds between the two extremes, though their controls, 153.5 and 160, were practically equal at the dif- ferent times. But not alone did type F show varying degrees of resistance at different times when tested to the 6 per cent alcohol, but it mani- fested to a more remarkable degree the same variability when tested to a stronger concentration. Type E as we may recall, while it showed unusually high resistance to a 6 per cent killing fluid, showed low resistance when tested to 8 per cent. Type F, on the contrary, though low to 6 per cent alcohol, sometimes gave a resistance to 8 per cent which was but little short of that given to a 6 per cent solution. This may be shown in the following experiment: Experiment Xla RESISTANCE OF STENTORS OF TYPE F To 8 PER CENT ALCOHOL, AFTER LIVING 4 DAYS IN I PER CENT ALCOHOL A Four Days in 1 Per Cent C Control Seconds Seconds EX PeMMUNCiANStOPeisiaie.ci-i-ie a lsie\< fee = sacccoosanac 80 DAC ANStODis. sctemetnae Meee a a5 Dy Gita ySEO Peraeisteloveler ele erate eee 50 Q ciliaystO pees Jerre. scutes eels s 220 Qhciliaysto preeietiast err eee 65 AiCiaystOpeacepreiseeloniiste Ons 45 AY Ciliansto Decent eee ee 55 GACiliabste Paessenpene nemo te Te) EM GHIVANSEO Petefetcteletseeiekerer orate 50 Gyciliaysto pyarcenenie cement aee 150 Grciliatisto peers ear eaten ee 60 7 CUNY SiOPp oboe census seme de one 210 Ff GME! Sti coosandacasoonens occ 180 Bi ciliaysto pepnrieis sess, ois eons ere 165 Scciliaesto pemeeeeee ae eee 50 Onciliaustopeyatecicce eee ce ene 240 Well Gif dcgoocardoodaose ac0e 135 LOVciiatsLO pee sect erae e 190 HO CiliastO Perle eee 80 156.5 80.5 From Experiment XII it is seen that the increase is much like that given in type F of Stentor. Although in Spirostomum there is a greater range of variability yet the striking feature in the many cases observed was that although the normal animals varied in their resistance from time to time the ratio between acclimatized and unacclimatized (control) animals remained practically the same. Immunity of Lower Organisms to Ethyl Alcohol 597 From experiments in an 8 per cent solution, another point of extreme interest was observed. Organisms often showed the first signs of injury within the usual time—one to two minutes—and disintegration followed in the usual way. Upon reaching mid-body, however, this was suddenly stopped. At the point of injury a round plug of proto- plasm formed, filling up the wound. Thereupon the cilia resumed a backward stroke and the body moved forward in a normal fashion. This phenomenon was observed again and again as the method by which the organisms often prolonged life for considerable periods of time. In the same way that type E of Stentor was tested on the second, fifth and seventh days, we may test Spirostomum to see the general indications of resistance. These results, shown in a condensed way in the following table, are different from those obtained in type E of Stentor. Experiment XII] RESISTANCE OF SPIROSTOMA TO 8 PER CENT ALCOHOL AFTER LIVINGIN I PER CENT ALCOHOL. A = LIVING IN I PER CENT ALCOHOL. C = CONTROL Second Day Fifth Day Seventh Day A c A Cc A Cc - Seconds Seconds Seconds Seconds Seconds Seconds 45 ye) 195 5 2) 45 205 150 240 220 260 55 HS) 45 75 12 195 Wes 55 he) ifs 9° ge 4° go 65 215 18 fe) 55 50 200 150 255 5° go 50 120 105 JO 7o 190 85 95 65 255 60 60 5° 225 go 205 Jo 190 60 95 go 7° 45 105 120 Average resistance ....124 104 165 83.5 168.5 65.5 In the foregoing series only a slight increase for the acclimatized animals is shown on the second day. It will be noted, however, that the normal resistance (104 seconds) is high. In other cases 598 i ¥. Frank Daniel a better increase was shown at the end of the first day than is here given on the second. Probably the most typical results for the series are those on the fifth day. On the seventh a slightly higher average obtains for A, but C is less resistant. Had it been typical, however, there would still be a ratio slightly above two to one. Another point may be noted in passing. ‘The control animals, unlike those of Stentor, instead of increasing in resistance, showed a gradual decrease. A further study was made in Spirostomum of the immunity shown at considerably later periods of time. The greatest difficulty encountered was in keeping the acclim- atizing medium A at its usual strength. In order to do this two plans were tried. In one the culture was changed every few days; in the other it was kept in ground glass vessels which were sealed and set at constant temperature until the time of experiment. The latter method was adopted as giving better results. An experiment under these conditions follows in which the organisms were tested after eleven days in the acclimatizing and control media. Ex periment Xi. RESISTANCE OF SPIROSTOMUM TO 8 PER CENT ALCOH OL, AFTER LIVING II DAYS IN I PER CENT ALCOHOL A 11 Days in 1 Per Cent Alcohol C Control Seconds Seconds IDeqohe 1 ll Oya senooo ge snacawoo 3205 190 Exp:, leilia stops.css ee st ee eee Decal sto Pw /actactale steric rene cto a= 50 2 CilfaySto pice. tier sient Cae eS QuciliagstOp oem rytee eteaes 85 NEWER SC ac acopagt nose gd rennet 25 /WEGIVEY O}3), cboob odes so ddeobo pac 95 4 Cilia stOps.. ss ecn ahd ence cs ee RectliagstOpectatieip ea aciei emr 95 Siciliaistopsac. nee se ee (GUE) Nese ot sconoepasbe 390 Giciliastopin.2. sce S uC ATStO Dicer-aneeei te alsa ene 45 Filia StO Pane see.cigs sore ot Os rier eS Sheiliacstopeeesemeaee eta c eee: 285 Sicilia isto pee ceome eae ei aeRO GHEUNEE TO) Wen qogodes cuss QueU ToT 150 Qiciliastopaees eee aes et n TOV ay StO Ps eee eres ieee es 70 LON ia StOp