ie ass > o we . petatel Par ok Ag de oe sadidataceentatany plecatole tet es acees teed eet HOC Srl treet tet pritstatanisrtarstreitiscaatonereceg Z8 ae ies . Eehieaiatat i: oy, Crt, 4 teh he # of fats rh fe f oA Korte uy i ete lrte pk H pa 3h a Aether pees < tte tf Tatetes Oh red tit fet th Ay te % = oe “~s ~ pigie wi EE Meek ts ut soe i Pk Puma eas SCRE (e Peto pe 4 eet oe MELE MPUERG NTT Pitertate fekete ess! ; Ue ode otata reteciesesest reatetaesietatae: eaetar rad rae Cett et ete tote ae ace Gees SEES ager bee these ke Sate ncetet te Xt ae ete, ae bee or ee Ort “4 aatihatatareh tas Eetes tke Gteteytt 7 a Perk, ft Ay Prbatienaevceneent ts pie itataestataeaoat: betty @ 3 ms Ot pir Gty ite WR bette he Wa AS a ae te SrOreeaasa shtiesstets “h B arate bet Be: Aree a spi THE JOURNAL Ol EXPERIMENTAL ZOOLOGY EDITED BY WiuuiamM FE. Castie Jacqugs LoEB Harvard University The Rockefeller Institute Epmunp B. WILSON Columbia University Epwin G. CoNKLIN Princeton University Tuomas H. Morgan CHARLES B. DAVENPORT : Ah: Columbia University Carnegie Institution GrorGEe H. PARKER HERBERT S. JENNINGS Brea nua ereity Johns Hopkins University RAYMOND PEARL FRANK R. LILLIE Maine Agricultural University of Chicago Experiment Station and Ross G. HARRISON, Yale University Managing Editor VOLUME 30° i JANUARY—MAY, 1920 THE WISTAR INSTITUTE OF ANATOMY AND BIOLOGY PHILADELPHIA, PA. i a oy 7 6 oo) Dae 1 Ld i e j 4 peUSS SALLE = \ Ey C4 \ Vee 9 a iW ; PAAR es Lia DL 5 ie AB Oe, » ‘a \ t (fds i : tif : t. ' , \ i, . ‘ } Mad ; ia) el Awe / * ‘Fey eye ioe, iS Ke € > Wy i “s ’ fw ; Wet i. ; ' na 7 * " . ‘ j ' don —_~" . . > orn : . = c os 6 Uy . Fhe i ny x a 4 rs ' CONTENTS No. 1. JANUARY Rosert W. Heaner. The relations between nuclear number, chromatin mass, cytoplasmic mass, and shell characteristics in four species of the genuspArcellaa) Morty=seven feunesiu.....- cei «sti tee 1 oes rte. il C. M. Jackson anp C. A. Stewart. The effects of inanition in the young upon the ultimate size of the body and of the various organs in the Ml binOyna te aeeiIve CHATS : 2. eens ae = ous s a cinch ss) ORO R ee ete 97 J. A. Dawson. An experimental study of an amicronucleate Oxytricha. II. The formation of double-animals or ‘twins.’ Twenty-two figures.. 129 H. Saxton Burr. The transplantation of the cerebral hemispheres of ply stoma us Nine, fie Tesaemerte ee ee otnce o'<5.> )-tdla.s Sie staal: « Se eiets sia uisels 159 No. 2. FEBRUARY H. V. Witson AND BLACKWELL MarKkHam. Asymmetrical regulation in anuran embryos with spina bifida defect. Five figures................. bal Bennet M. Auten. The results of earliest removal of the thymus glands IneRanaspiplens tadpolesumOneshieunresne ins cia eiaLran arene: 189 BENNETT M. AuLEN. The parathyroid glands of thyroidless Bufo larvae.... 201 L. V. Hernprunn. An experimental study of cell-division. I. The physi- cal conditions which determine the appearance of the spindle in sea- 10 URE) aN 0 Cet o9 a E's o cl uct ol EERE EArt pcre Gol stints Preemie OMe IRIs Ai 211 CaswELL Grapr. Amaroucium pellucidum (Leidy) form constellatum (Verrill). I. The activities and reactions of the tadpole larva. Four ILD TS) CA RRR co, 0 4 Sia Cea ee eae enone rs Songer toc ore uee Gothen 239 No. 3. APRIL CHARLES Howarp EpmMonpson. The reformation of the crystalline style in Miya‘ arenaria.aiter extraction. 'Thirty figyres..........5.0.6,..0.000: 259 Cuartes ZeLteny. A change in the bar gene of Drosophila melanogaster involving further decrease in facet number and increase in dominance. Nimestouimes tempter ene, 5 a ht ele pie Sonn UY te A tetas fk | ul 2 AL Ea ey 293 Haroup Cummins. The role of voice and coloration in spring migration and SOX TCC Grrammean MBREO UG HOR. ses ea le LRN Di ne alec oe ety as 325 Henry Laurens anp Henry D. Hooker, Jr. Studies on the relative physiological value of spectral lights. II. The sensibility of Volvox to wave-lengths of equal energy content. Two figures................. 345 ill | GDdbb oO lv CONTENTS Francis B. Sumner. Geographic variation and Mendelian inheritance. Seven: Heures, ooo css cc dle cies de «ete sin 0'0 loa cha eR ality Wad ols i fihais Ge'o.e'o'e ve 369 C. M. Cuttp. Studies on the dynamics of morphogenesis and inheritance in experimental reproduction. X. Head-frequency in Planaria doroto- cephala in relation to age, nutrition and motor activity................ 403 No. 4. MAY Epwin Carteton MacDowett. Bristle inheritance in Drosophila. III. Gorrélation. Might fieures: 2.) tt. «. <0.-. +e eens = Fes eres eres 419 Minna E. Jewretu. The effects of hydrogen ion concentration and oxygen content of water upon regeneration and metabolism of tadpoles. Twenty -fOurmZUhed e610 geen o ovis: ote one oS che < lee ake eee 461 ~~ Lak Resumen por el autor, R. W. Hegner. Universidad Johns Hopkins, Baltimore Las relaciones entre el numero de nticleos, la masa de cromatina, la masa citoplismica y los caracteres de la concha de cuatro especies del género Arcella. El presente trabajo contiene observaciones y experimentos sobre Arcella denta, A. polypora, A. vulgaris y A. discoides. Ejemplares de clones cuyo tamafio y numero de espinas era cono- cido se cortaron en dos pedazos. Los pedazos uninucleados de los padres binucleados produjeron descendientes uninucleados que tenfan aproximadamente la mitad del tamafio de los padres. Esto indica, por consiguiente, la relacion de una cantidad definida de citoplasma con un solo nicleo. Las masas decromatina fueron medidas en ejemplares procedentes de clones de A. dentata que diferfan en tamafo y en nimero de espinas; el autor ha com- probado que la cantidad de cromatina era menor en los ejemplares procedentes de los clones mas pequenos. Es decir, que la can- tidad de citoplasma varia directamente con la cantidad de cro- matina. A. polyporacontiene de tres a trece nucleos. El nimero de nticleos varia dentro de un clono, y el tamanho enun mismo clono est’ relacionado muy estrechamente con el ntimero de nuicleos. Los ejemplares con el mismo nimero de niticleos, pero pertenecientes a clones diferentes pueden diferir en tamano. Las medidas de las masas de cromatina demuestran que los ejem- plares del mismo tamajfio pertenecientes a clones diferentes tienen aproximadamente la misma cantidad de cromatina, aunque el numero de nticleos sea diferente. Esto indica que la cantidad de citoplasma no depende del nimero de niicleos, sino de la can- tidad total de cromatina. El autor discute los datos obtenidos con referencia a la teoria de la relacién nticleocitoplasimea, la teorfa de la relacién cromatocitoplasmica, las teorfas de las lineas puras y de la seleccion, y la hipétesis de los cromidios. Translation by José F. Nonidez Carnegie Institution of Washington AUTHOR’S ABSTRACT OF THIS PAPER ISSUED BY THE BIBLIOGRAPHIC SERVICE, OCTOBER 13 THE RELATIONS BETWEEN NUCLEAR NUMBER, CHROMATIN MASS, CYTOPLASMIC MASS, AND SHELL CHARACTERISTICS IN FOUR PIECES OF THE GENUS ARCELLA! ROBERT W. HEGNER School of Hygiene and Public Health, The Johns Hopkins University FORTY-SEVEN FIGURES CONTENTS 1. Introduction. . 1 a. The ois anie eR en 1 b. ore of rec ihe felon otaplscric Peden 1 c. Methods.. See ee ee + 2. Experiments on relics dentanee Re | a. Experiments on binucleate eee o Sule 150... 4 1. Introduction. . 4 2. Results of remenaeee Soh oF dibs anal cad ant of ne eytaplicas 4 3. Phe reproduction of uninucleate pieces:......:......-:......2+. 5 4. Empty shell formation and nuclear doubling. .................. 8 5. Correlations. . sa G pares. ND 6. Changes from athmdlen ys ie “ine ences amactitan Oe ee 14 7. Reproduction of a uninucleate piece formed without an opera- tion, line 159. . ant: 17 8. Microdisesenon exper ene a Pts cer 18 b. Experiments on uninucleate specimens so Emly 150... sp epaita ak c. Experiments on members of family 152.. St fone, oe ee Co d. Experiments on members of family 58........... see OO e. Summary of results of experiments on iN cella cee RE 3 eC 1 The observations and experiments described in this paper were begun in the zoological laboratory of the Johns Hopkins University on December 27, 1917; were continued at the laboratory of the Brooklyn Institute of Arts and Sciences at Cold Spring Harbor, Long Island, from June 17 to August 27, 1918, and were completed at the School of Hygiene and Public Health of the Johns Hopkins Uni- versity. The writer is indebted to the members of the Zoological Department of the Johns Hopkins University for many courtesies and to Dr. C. B. Davenport for the opportunity of carrying on work at Cold Spring Harbor. He is particu- larly grateful to Prof. H. 8. Jennings for his valuable counsel. Much of the sta- tistical work was done by Dr. Ruth Stocking Lynch, instructor in Protozoology in the School of Hygiene and Public Health. 1 2 ROBERT W. HEGNER 3. Observations and experiments on Arcella polypora...................... a. Origin of the specimens studied. . ay, ; b. Relations between nuclear naive sae Heeaeeen in Feild specimens. . On C. d. é 1. 2. 3 4 5. 6 7 Variations in diameter within families. . ee Relations between nuclear number and diareter: Ww cata ensilt farnilieg and between families. . . Relations between nuclear Seestaltjs eal dimeter SAN a eee ite ily. Family ap. 5.. See Correlation in sina: etessan oe ane anal ieemncdinte ae spring. . : oe Results of eiertiod Oye fees Ww As iets pan sail JHeneees . Relations between nuclear number and diameter a ae ae selection period.. ; Results of selection sor Ee ge Sad call diaeies pies groups Swithithe same mumber Of nuclei. sere. 2 ee) eee Changes in nuclear number and their results.. . Results of a further increase in nuclear Smee ne . Uninucleate and binucleate specimens oped icy aaa experiments. . f. Differences. between ere ee ith regard to eee num- ike ber and diameter.. ays Variations in niclean ques: Sil Sionasone in w ace specimens aaa within small families. . die . Differences between fanless ap. 5 Ain Gountlilos ¢ ap. 38, ap. . 39, ap. 69, and ap. 34.. Be te a. Families ap. 38, ap. 39 atl ap. 69. ee b. Family ap. 34, Comparison of nuclean eee and Gigs with Soule ap. g. Relation between diameter Be shell sie diamevce os acm in ne populations and within families. . h. Summary of results of sigeee Nene ond Susman on ic polypora. . sia and ssugnimantec on aN cells Rirecodes = . Pure line.. = ere b paises in memes onal in eeciees, eee c. The reproduction of uninucleate pieces. . d. Summary of results of observations snl ee shane on TArcella discoides.. . Observations and eecrmenesn on aM pelle eraleanie: te relations between chromatin mass and cell size. . Chromatin mass and cell size within two fetes A Areelia Gentsin b. Chromatin mass and cell size within two families of Arcella polypora c. The possible relation between the chromatin mass and the death of certain specimens of Arcella dentata.. d. Summary of the studies of chromatin mass ed dels SIZE... nse ae DISCUSSION. 2": SNe ee Sa eck TRON ee Ere ain ns ne NUCLEOPLASMIC RELATIONS IN ARCELLA 3 1. INTRODUCTION a. The problem The investigations described in the following pages were undertaken to determine the relations between nuclear number, chromatin mass, cytoplasmic mass, and shell characteristics in certain species of the genus Arcella. As a rule, genetic research in animals is limited to the study of somatic characteristics alone or to the examination of germ cells that have been killed and pre- pared for microscopic observation. In many cases the germ cells of organisms that have been used in breeding experiments have been studied, but these germ cells have been obtained either from control specimens or from pedigree specimens that have been killed for the purpose. Arcella is peculiarly favorable for investigation because both nuclear and cytoplasmic charac- teristics can easily be seen drawn and measured, at the same time, in the living animal, and their relations can thus be established under the most favorable circumstances. b. Advantages of Arcella for nucleocytoplasmic studies In a previous paper (Hegner, 719), various characteristics that make Arcella a favorable organism for genetic investigations were pointed out. Among these are: 1) the power of multiplying vegetatively and rapidly; 2) the presence of definite measurable characters that are not modified by growth; 3) the semitranspar- ency of the shell which makes possible the examination of the contents, especially in the recently formed offspring; 4) the abil- ity to withstand severe operations, and, 5) the ease of cultiva- tion and examination. To this list should be added the fact that the nucleus is of the vesicular type with the chromatin, when in the resting stages, clumped into a spherical mass which may easily be drawn and measured. It is thus possible to study the relation between chromatin and cytoplasm, which offers a much more accurate means of comparison than when the entire nucleus with the nuclear sap is involved. 4 ROBERT W. HEGNER c. Methods The methods of rearing the specimens recorded in this paper are the same as those previously described (Hegner, 719). The operations were for the most part simple. The specimens were first drawn with a camera lucida; they were then cut in pieces with a small, sharp scalpel, and the positions of the cuts were indicated on the sketch. In one set of experiments the nucleus only was dissected out with the aid of a Barber microdissection apparatus. Parts of specimens or those from which nuclei had been removed were cultivated as were the entire animals. 2. EXPERIMENTS ON ARCELLA DENTATA a. Experiments on binucleate members of family 150 1. Introduction. The progenitor of this family (fig. 1) was taken from a pond on the campus of the Johns Hopkins Uni- versitv at Homewood, Baltimore, on December 27, 1917. Its spines consisted of almost indistinguishable ridges and could not be counted. In diameter it measured 34 units of 4.3u each. Four immediate offspring were obtained from this specimen, all of which exhibited well-defined spines, showing that the absence of spines in the parent was probably due to some envi- ronmental factor. The fourth offspring was represented by an empty shell; the other three possessed 13, 14, and 14 spines, respectively; one was 34 units and the other two were 35 units in diameter (fig. 2). 2. Results of removing part of the shell and part of the cytoplasm. The first experiment was designed to answer the following ques- tions. If part of the shell is removed, is a new part regenerated? What influence on a specimen and its descendants has the re- moval of part of the cytoplasm and part of the chromidia? The first offspring was operated on as indicated in figure 3. Part of the shell and some of the cytoplasm were removed. No regen- eration of the shell occurred. The first offspring produced by this specimen after the operation was smaller than the parent (15-31 in fig. 2), this decrease in size being due probably to the ~ NUCLEOPLASMIC RELATIONS IN ARCELLA o removal of part of the parental cytoplasm. The first offspring of this specimen, however, was as large as the original parent and the mass relations between nuclei and cytoplasm were thus regained. It may be noted that the protoplasm removed con- fa) 70 6 Fig. 1 Arcella dentata. Outline of the shell of the progenitor of family 150. The inner circle represents the mouth opening. X 207. Fig. 3 Arcella dentata. Outline of the shell of specimen 150.1. The cross line indicates the portion of the shell and cytoplasm removed. The small circles represent the two nuclei. X 207. Fig. 5 Arcella dentata. Specimen 150.2. The cross line indicates where it was cut in two. The upper, smaller portion was the progenitor of line 150.2a; the tower, larger portion, of line 150.2b. > 207. Fig. 6 Arcella dentata. Specimen 150.2bl. The first offspring of the larger portion of specimen 150.2 shown in figure 5. XX 207. tained a portion of the chromidial net which is so conspicuous in Arcella (fig. 4), but there is no evidence that its removal affected the characteristics of the organism. 3. The reproduction of uninucleate pieces. The next problem was to determine the effects on the organism of removing one 6 ROBERT W. HEGNER nucleus and half of the cytoplasm. The second offspring was cut into two slightly unequal parts, each containing one nucleus (fig. 5). The smaller part was labeled 150.2a and the larger part 150.2b. Both of these parts survived and reproduced. Their immediate offspring were slightly irregular in shape (fig. 6) but exhibited spines that could easily be counted. The effects of injuries to the parental shell upon the shape of the off- spring are clearly indicated in figures 7 and 8. The immediate eee 11225 E 2 Fig.2 Arcela dentata. Pedigree of family 150 showing the number of spines and diameter of the shell of the progenitor of the family and of a few of the progeny. Each vertical series of numbers represents a generation. The number preceding the dash is the number of spines and the number suceeeding the dash is the diameter of the shell in units of 4.34 each. Fig. 4 Arcella dentata. Nuclei, cytoplasm, eytoplasmic attachments to the inside of the shell, and the chromidial net are shown in this figure as they appear in a stained specimen. %X 310. progeny (B) of the half specimen (A) is irregular in shape, but the normal shape (C) is usually regained in the next generation. The condition of the shell of the parent evidently has only a very slight influence upon the shape of the shell of the offspring. Besides being slightly irregular in shape, these offspring were much smaller than the original, entire parent, and possessed only one nucleus each. Furthermore, the offspring of the smaller part (150.2a) were smaller than those of the larger (150.2b) as is shown in the pedigree in figure 9. This result is similar to that NUCLEOPLASMIC RELATIONS IN ARCELLA 7 described in the first experiment, in so far as it indicates that the size of the immediate progeny is, at least in part, dependent upon the amount of cytoplasm in the parent. By the fourth generation (table 1) the specimens in line 150.2a had reached a mean spine number approximately equal to that of line 150.2b and were rapidly approaching the latter in diameter. This sug- (3 @ Srciac 8 Figs. 7 and 8 Arcella dentata. Specimens of family 150. A is one-half of a binucleate specimen; B, its first offspring, and C, the first offspring of the suc- ceeding generation. The normal shape is usually regained in the second genera- tion. X 207. gests that the nuclei and cytoplasm in the two lines were quali- tatively alike and that their interaction was such as to lead gradually to the production of specimens in which an equilibrium between the nuclear and cytoplasm masses was regained, thus resulting in specimens similar as regards spine number and diameter. 8 ROBERT W. HEGNER 4. Empty shell formation and nuclear doubling. At first it appeared that a smaller, uninucleate line of Arcellas had been established by the cutting experiments, but soon a most inter- esting phenomenon occurred at the time of division in the case of certain of these uninucleate specimens; this was the forma- eee 9=24 10-26 9-25 sae oe 10-25 10-25 150.2e 8-23 10=24 9-23 8=26 10-26 10-26 10-26 11-25 11-26 9-25 11-25 11-26 (asowe} 9-26 a Be a 12-27 10-26 pe: 10-27 11-26 10-26 11-27 150.2b 8-25 10-26 10-26 10-26 12-27 12-27 hi. ea 8-27 10=26 Fig. 9 Arcella dentata. Part of the pedigree of the two lines, 150.2a and 150.2b resulting from the bisection of specimen 150.2. TABLE 1 Arcella dentata. Table showing differences in mean spine number and in mean diameter for the first four generations of lines 150.2a and 150.2b, exclusive of empty shells and the progeny of binucleate specimens. The unit of measurement AS Leo p yh GENERATION 1 GENERATION 2 GENERATION 3 GENERATION 4 LINE ae Mean me Mean ee Mean a Mean es i i : i iame nnmibes: diameter AES diameter nbn diameter anes d ter 24.86 | 10.80 | 25.00 26.43 | 10.33 | 26.40 150.2a 8.67 23.33 | 9.12 24.37 9 150.2b 9.50 25.75 | 9.91 26.18 | 10. 1 Eo NUCLEOPLASMIC RELATIONS IN ARCELLA 9 tion of empty shells and the regaining of the binucleate con- dition. Empty shells were formed at irregular intervals, and in every case the parents, on examination, were found, after empty shells were thrown off, to possess two nuclei instead of a single nucleus. The next offspring produced by these binucleate parents also possessed two nuclei and were always larger and had more spines than their parents or the empty shells (fig. 10). The offspring of these binucleate progeny were 11-26 E __ 15-52 (2) 12-29(2) 13-32 (2) : ee a p> 15-25(2) 13-33(2) 1-27(1 (150.20.2.1.1) we BEDI TW ¥ aa Fig. 1 Bufo embryo. Dorsal view. pn, posterior end of neural plate; tb, tail buds. X 18. Fig. 2 Transverse section of embryo shown in figure 1, through the blasto- pore. arch, archenteron; de, dorsal entoderm; Ib/, left blastopore lip; lmf, left medullary fold; nc.m, gastral mesoderm; p.m, peristomal mesoderm; rbl, right blastopore lip; rmf, right medullary fold. X 40. It is clear that in this embryo a full set, not a half set, of axial structures had begun to develop on the left side of the blastopore. In this differentiation of the axia! body, the right lip had no share. CHOROPHILUS LARVA WITH ASYMMETRICAL REGULATION The frog larva next to be described (figs. 8, 4, and 5) represents fairly well a later stage of the embryo shown in figure 1, suppos- ing that the asymmetrical regulation begun in the latter had gone on. This embryo appeared in a batch of normally develop- 174. H. V. WILSON AND BLACKWELL MARKHAM ing eggs of Chorophilus feriarum which had been artificially in- seminated February 15th. It was early singled out by Mr. T. E. Rondthaler as one in which the blastopore did not close, and was kept under more or less continuous observation. It elon- gated and became somewhat flattened dorsoventrally, the yolk plug occupying a median dorsal position in the posterior part of the body. Immediately behind the yolk plug two tail buds, one left, one right, appeared. They both early acquired a spiral Fig. 3. Chorophilus larva. Dorsal view. dlt, base of left tail; brt, base of right tail; tr, tip of right tail; ya, exposed yolk area. X 25. Fig. 4 Ventrolateral view, from the left, of larva shown in figure 3. le, left eye; ln, left nostril; m, mouth; élt, tip of left tail; rt, tip of right tail; ya, exposed yolk area. X 25. curvature in different planes. External gills, which are small in this species, developed. With their disappearance and the steady increase in size of the head and trunk region, the yolk plug grad- ually shifted over toward the left side of the body. One tail, the left, was thus carried into a position which was distinctly ventral (fig. 4). The other tail, the right, which gradually be- came much the stouter, took up a position which was distinctly dorsal, its basal part directly in line with the median longitudinal axis of the trunk (fig. 3). REGULATION IN ANURAN EMBRYOS 175 By this time the larva showed well-developed eyes, mouth, and nostrils, and was about 2 mm. long with a greatest breadth of 1 mm. It did not die naturally, but was preserved, March Sth. At that time it seemed to be in good health with a fair prospect of living on. Sections showed that internal gills and the opercular cavity were well developed. The right tail had relatively large dorsal and ventral fns in its posterior part, and at its tip the left tail likewise had fins. The exposed yolk area lig. 5 Transverse section of larva shown in figures 3 and 4, through anterior end of yolk mass. arch,archenteron; alt., tubes representing an alimentary canal in process of formation; c, coelom; l, liver; bl, left blastopore lip; Im, left myo- tome; /s, lymph sinus; nc, notochord; nt, neural tube; rbl, right blastopore lip; rm, right myotome; rpr., right pronephros; ya, exposed yolk area. X 67. measured in length about one-half the total length of the body. Internally the yolk mass was found to occupy slightly more than this ratio. A study of the internal anatomy as made out in a series of transverse sections gave the following facts: The brain, eyes, auditory sacs, nostrils, mouth, pharynx, gills (internal), opercu- lar cavity, and heart showed nothing obviously abnormal. But in the trunk and tail region there were numerous results that had followed from the failure of the blastopore to close. We 176 H. V. WILSON AND BLACKWELL MARKHAM take these up in order, as affecting the alimentary canal, spinal cord, notochord, myotomes, and pronephros. Alimentary canal. The anterior end and the pharyngeal re- gion of the alimentary canal have the usual anatomy, appearing as parts of a wide, straight passage. This comes to an end at the anterior limit of the yolk mass. The latter in this region is found to be excavated by numerous tubular channels, round which the cells are arranged in epithelial fashion. A few other similar tubes with less yolk in the cells lie at the surface of the mass (fig. 5). It is possible to trace some interconnections be- tween these, and the tubes are perhaps all interconnected, rep- resenting an alimentary canal in the making. A remnant of the original archenteron persisting as a shallow slit-like cavity, which extends well into the yolk and connects with the exterior round the right lip of the blastopore, is present in this region (fig. 5, arch). It connects with, at any rate some of, the tubular chan- nels above referred to. Possibly it forms the terminal, anal, part of the alimentary canal which is in process of differentiation. The posterior and much the greater part of the yolk mass re- mains as solid and compact and undifferentiated as in a gastrula stage. It is clear that in this larva an alimentary canal was in process of formation by a method different from the normal. The de- parture is probably adaptive to the continued presence of the yolk mass. In the figurative language of vitalism, the embryo makes an effort to form an alimentary canal, although the cus- tomary road to that end is not open. Moving slightly away from vitalism, we seem to see the destined end of the ontogeny work- ing backward as a cause, a philosophic idea which certain embry- ological writers (ef. Jenkinson’s admirable book, ’09, p. 20) have in recent years shown a willingness to adopt into our family of concepts concerning the processes at work in ontogeny. To in- voke such retroactive influences may, as Jenkinson says, in the end prove necessary, but on the other hand it may not. The fact is that while the idea of individual adaptation or regulation in ontogenetic processes is now familiar to us, the detailed facts (comparative and experimental) of any particular set of cases REGULATION IN ANURAN EMBRYOS Viva have only been touched upon in a preliminary, reconnoitering sort of way. Brilliant as the results are, they need to be inter- woven with rich collections of special knowledge, which shall combine the data of experimental organogeny with tabulations of the various organogenetic methods practiced by sets of indi- viduals, and races, of a species and by related species. "T'wenty- five years ago, one of the writers mapped out in a descriptive paper (Wilson, ’94) this plan as affecting the study of the com- parative embryology of sponges. The idea was to learn some- thing, through comparison, of the kinds of changes that may be made in the ontogenetic processes of a race and the laws govern- ing their appearance. The comparative study remains as im- portant to-day as it was then, and the field is almost as virgin. In a word, before accepting this conclusion that the end sets in activity the means, we need to know in any particular case much more both about it and allied cases. Ontogeny we picture as an intricate series of events, each event setting free stimuli which bring about other events. lf, then, a certain event does not occur, the embryo in respect to this point may, for all we know, fall back to a lower level, to a more generalized ancestral condi- tion, in which it makes responses such as it would have made in former times. Or it is possible that steps in ontogeny not directly dependent on the occurrence of the inhibited event may continue to be made, one after another, and thus we may come out with an embryo in which only a part is missing, in which case the missing part might conceivably be supplied as if it had developed, had been lost, and now were being regenerated. All this, of course, is conjecture, but it is at least physiological. Liver. The liver appears as a large cellular mass excavated by abundant sinuses which divide it up into cords. It is inti- mately connected, throughout most of its extent, with the yolk mass, shading off into the latter in places. Spinal cord. The spinal cord in the trunk region, as far back as the base of the right tail, is symmetrical, one side having the same amount and arrangement of incipient gray and white mat- ter as the other. At the base of the right tail, the spinal cord is continued into a tube with a much wider lumen. The wall of 178 H. V. WILSON AND BLACKWELL MARKHAM this caudal continuation of the cord is markedly thicker on its morphological right side than on the opposite side. On the for- mer side it consists of a high lining epithelium, outside which lie some rounded cells and then a little white matter, whereas on its left side the wall consists only of a layer of low epithelium. The tube narrows to one which still presents a relatively large lumen, but the wall of which is nearly uniform in thickness, consisting of a simple columnar epithelium. It extends nearly to the tip of the tail. Summing up for the neural tube, it may be said that the neural tube of head and trunk is continued directly into the right tail. Elsewhere symmetrical, in the right tail it gives an indication of being a half-cord. In the left tail and along the left lip of the blastopore there is no sign of a neural tube. Notochord. The eylindrical notochord of the posterior head region and trunk is continued directly into the right tail with no diminution in size. It extends the whole length of the tail, re- maining cylindrical, but decreasing gradually in diameter in the usual way. In the left tail and along the left blastophore lip there 1s no sign of a notochord. Myotomes. ‘The myotomes of the trunk are as in figure 5. There are two series, but those on the right side, rm, are strongly developed, those on the left side, lm, very small. The series of strongly developed myotomes on the right side of the body is continued into the right side of the right tail, and the series of small myotomes on the left side of the body into a similar series on the left side of this tail. This disproportion in size, between the two series of myotomes in the tail, is as great or greater than in figure 5. The series of right-hand myotomes is continued almost to the tip of the tail. The series on the opposite side comes to an end shortly behind the basal part of the tail. In respect to the myotomes, then, the basal part of this tail is sim- ilar to the trunk (fig. 5), but the greater part is more distinctly a half-tail in that it lacks myotomes on one side. Doubtless these were developing from before backward, when the larva was killed. REGULATION IN ANURAN EMBRYOS 179 The other, originally left, tail has a distinctly developed series of small myotomes on one side, the side turned away from the yolk mass (originally the left side). This extends from about the tip of the tail to its base and forward as a very small stripe for some distance, close to the morphological left lip of the blasto- pore, fading away anteriorly. It is evident, then, that the left lip of the blastopore had made steps toward organization. But the-regulatory processes going on along the other lip made these steps futile. Pronephros. The right pronephros, figure 5, r.pr., 1s well de- veloped and the right Wolffian duct extends backward through the trunk. What appears to be the left Wolffian duct is present near the left blastopore edge, about the middle region of the body, extending through a number of sections. The anterior end of this tube could not be traced. Thus whatever steps have been taken to form a left pronephros have not gone beyond the merest start. In the embryo just described the axial organs (neural tube, notochord, myotomes) have not, it is clear, been built up by the fusion in the middorsal line of two halves. Instead, an asym- metrical method has been followed, the fundamental feature in which is the utilization of only one blastopore lip. The steps in this process are briefly as follows: 1) The exposed yolk area is shifted over toward the left side. 2) The right blastopore lip and the right tail bud are thus brought in line with the median axis of the head end. 3) The right lip organizes a backward con- tinuation of neural tube, notochord, and the right series of myo- tomes, all continued into the right tail bud, which grows much larger than the left one. 4) After having established axial or- gans in the trunk region, the right blastopore lip grows for some little distance over the exposed yolk, thus coming to lie to the left of the median body plane. A series of myotomes is then or- ganized along the left side of the notochord. 5) The left lip of the blastopore and the left tail bud make no more than the first steps toward organization. The behavior and the anatomy of the larva at the time of preservation indicated that the process of establishing the normal 180 H. V. WILSON AND BLACKWELL MARKHAM structure would have gone farther had the larva been allowed to live. The yolk mass was apparently being covered in through further extension over it of the blastopore lips. The mass itself had made some headway toward transformation into a part of an alimentary canal. The left series of myotomes would prob- ably have extended itself to the tip of the tail (right tail), -and would have continued to grow in size. With the continued dorsal extension of the coelom on the left side of the body (fig. 5), the conditions for the formation of a left pronephros, to match that already developed on the right side, would have been more nearly realized. Absorption of the left, small tail would have completed the steps by which, in spite of its failure to go through certain early phases of the normal cycle of changes, the embryo might in the end have acquired the type structure. We hope at some later date to have actual observations to report on the final stages in the restoration of such asymmetrically developing lar- vae, if indeed they do succeed in going through the entire proc- ess. It is to be expected that here, as elsewhere, in respect to the actual details, individuals will vary. Cases essentially similar to the above have been recorded by Lereboullet._ (’63) for the teleosts. In the embryo shown in his figures 30 and 31, pl. III, one of the half-bodies degenerates, while the other develops into a whole body. This latter comes to lie in direct line with the head end. We thus get what is not far from a normal embryo, a lobe-like projection from one side and a split tail alone remaining to indicate the original du- plicity. This conclusion, which Lereboullet draws, is made prac- tically certain by the several embryos which he describes and which individually illustrate different stages in the process. In- deed, in the case of another embryo like his figure 31, Lereboullet was able to observe from day to day the gradual degeneration of one of the half-bodies. In another closely similar case, only briefly described, no. 55, loe. cit., p. 224, it would seem that only one blastopore lip became organized as a half-body, but Lere- boullet is not very explicit here. The frog embryos studied by O. Hertwig (92) do not, as he points out, all conform to the symmetrical type of development. REGULATION IN ANURAN EMBRYOS 181 For in some of them the organs develop in very unequal degree in the two lateral blastopore lips. This asymmetry is extreme in the embryo pictured by Hertwig in figures 15 and 16 of pl. XVI and figure 27 of pl. XVIII. In this individual development had gone so far that eyes and suckers had been formed. Never- theless, nearly the whole of the dorsal surface was occupied by an exposed yolk area. The special point of interest is that one of the lateral blastopore lips was found to be not organized at all. The opposite lip, on the contrary, contained a small neural tube and notochord, which anteriorly were continued into the corresponding structures of the very short head end and _ poste- riorly into a well-developed tail (half-tail or tailbud). Hertwig regards this interesting embryo merely as one in which the axial half-structures developed on one side and not on the other. He does not view or discuss it as illustrating a step in the direction of forming a whole trunk out of one blastospore lip. And indeed it may not illustrate such a step, although it obviously presents a close analogy to the Chorophilus embryo described above. Endres and Walter (’95) described a frog embryo in which possibly a process was going on similar to that in our asymmetri- eal embryos. They, loc. cit., p. 41 and passim, find with Roux that half-embryos, developing from one of the first two blasto- meres, restore more or less completely through ‘postgeneration’ the missing organs of the other side. In the case referred to (Hemi-embryo sinister: F, loc. cit., pp. 43 to 48) there is a large, dorsal, exposed yolk area. One blastopore lip is well organized and passes backward from the head end into a well-developed tail. The other, the ‘post-generated’ lip, is very imperfectly organized. Ill The common or typical regulatory method of establishing the type structure employed by embryos in which the closure of the blastopore has been extensively interfered with, is generally thought to be that in which both lips organize and come together symmetrically in the midline (group 3, p. 172). And there seems to be no doubt that this method is employed sometimes, although observations on its occurrence in an actual embryo are scanty. 182 H. V. WILSON AND BLACKWELL MARKHAM Lereboullet records some direct evidence for the teleosts. In case no. 56 (loc. cit., p. 225, figs. 32, 33, and 34) he watched the actual development of the embryo from day to day. Anteriorly and posteriorly some fusion of the two half-bodies apparently took place, although as late as the seventh day the half-bodies were still separate throughout an extensive part of the trunk, and no further fusion took place during the remaining five days of the embryo’s life. It is, however, probable that the now common view, which Lereboullet advanced, applies to this case and that under favorable circumstances the two half-bodies would have completely fused. The truth of this idea is made almost certain by cases like Lereboullet’s no. 57 (loc. cit., p. 226), in which the half-bodies have been brought so close together as to be practi- cally in contact, and his no. 59 (loc. cit., fig. 35), in which at the posterior end of the single trunk there is found a small dorsally situated aperture, interpreted as the remnant of the blastopore. O. Hertwig, in his well-known paper (92) on frog embryos exhibiting the spina-bifida defect, describes many such embryos in anatomical detail. He arranges them in a series which he interprets as representing the actual ontogeny. His series be- gins with the so-called ‘ring embryos,’ in which the exposed yolk area is very large and the part of the axial body lying in front of the blastopore lip very small—so small, indeed, that it includes no more or scarcely more than the anterior cerebral commissure. In such embryos he concludes with Roux (’88) that the lateral lips of the blastopore become organized into half-bodies, which gradually meet in the midline. And the individual frog embryos which he describes and arranges in a series represent, he thinks, successive stages in the actual ontogeny of such a ‘ring embryo.’ It is far from certain that all of the abnormalities, so arranged, really belong in a single ontogenetic series. Nevertheless, the data, recorded with admirable precision, doubtless justify the conclusion that sometimes the lateral lips do organize and come together in frogs, as Lereboullet had already claimed was the case in fishes. But Hertwig has no observations on the actual occurrence of this process in one and the same embryo. REGULATION IN ANURAN EMBRYOS 183 In fact, direct observations on the occurrence of the symmet- rical fusion of half-bodies are, as we have said, few and scanty. They lack detail. Nevertheless, Roux (’88, p. 443) has actually watched the gradual approximation of the lateral lips of the blas- topore in frog embryos exhibiting the spina-bifida defect (desig- nated by him, consistently with his theory, asyntaxia medul- laris), until the blastopore had quite closed. This symmetrical mode of development of spina-bifida em- bryos is looked on by many embryologists (cf. especially Roux, "88, pp. 448-444, and O. Hertwig, loc. cit.) as not fundamentally different from the normal embryology. The difference, accord- ing to these writers, is that, whereas in normal development the lateral blastopore lips come together before they are organized, in the spina-bifida embryos their fusion is so delayed that they become organized first. Thus the organization of the blastopore lips in a spina-bifida embryo is not looked on as an abnormal reg- ulation that starts up when something checks the activity of the morphogenetic process by which the body is usually lengthened. It is, on the contrary, a process that is perfectly normai in itself, only out of place in time. This is the point of view of the con- crescence theory of vertebrate ontogeny. Undoubtedly, symmetrical spina-bifida embryos strongly sug- gest the idea of concrescence—so strongly, indeed, that they are thought of by some as almost demonstrating the truth of its occurrence in normal development. But this reasoning is ob- viously vicious, since it begs the question as to whether organi- zation of the lateral lip is a normal process (although out of place in time) or a radically abnormal one. Such radically abnormal regulatory processes of course occur. Driesch and many others have made us far more familiar with them than were the older embryologists. And in establishing or restoring the type struc- ture, their disregard of the customary mode of attaining that end is well known. Following this line of thought, we may entertain the idea that the organization of the blastopore lip in the embryos in question is a thoroughly abnormal process, as far away, al- though in a different direction, from the normal mode of forming the axial body as is the process, for instance, in those rare teleost 184 H. V. WILSON AND BLACKWELL MARKHAM monsters, in which the blastoderm does not grow round the yolk, but remains small, while an embryonic body is developed extending completely through it from posterior to anterior edge (Kopsch, ’04, p. 95, Taf. X, Fig. 118). If the organization of the lateral blastopore lip is an unusual, abnormal process, spina-bifida embryos become no less interest- ing than they have been hitherto, since they tell us plainly that here is tissue which may be activated to develop, in very atypical fashion, into certain of the specific structures. From this point of view, such embryos may be thought of as arising in the follow- ing fashion. The axial body which normally forms in front of the blastopore lip is short. It grows in length especially at the posterior end which is carried backward, the growth being in part due to the incorporation of tissue belonging to the blasto- pore lip. The indifferent mass of tissue at the posterior end of this embryonic body is constantly being organized and added to the organs in front of it, some to neural tube, some to notochord, etc. Frankly abandoning deterministic theories which would see in the tail bud of a vertebrate embryo not indifferent tissue, but neural, notochordal, and other kinds of cells, we ask what brings it about that certain cells go to one, other cells to another organ? Plainly O. Hertwig and the thinkers of that school are right; it is position that determines the fate of the cells. Those behind the notochord become notochord. Those behind the neural tube become neural tube. What underlies this phenom- enon? The answer seems clear: the already differentiated or- gan, notochord, e.g., exerts a controlling influence on the con- tiguous, indifferent cells behind it and makes them into its like.* Now when the backward growth of the blastopore lip (to be construed as part of the general closure of the blastopore), and hence of the axial embryonic body, is prevented, what is the sit- 1 Roux has discussed this kind of influence (‘‘eine eigenthiimliche ordnende und gestaltende Wirkung,”’ 1888, p. 505 and passim) in the case of half-embryos of the frog produced by killing, or nearly killing, one of the first two blastomeres. In the formation, ‘postgeneration,’ of germ layers in the operated half, the or- ganizing influence extends outward from the edge of each, already formed, germ layer of the normal half. This progressive differentiation of relatively indiffer- ent stationary material, where the ‘‘differentiative stimulus passes from the al- REGULATION IN ANURAN EMBRYOS 185 uation? The embryo finds itself unable to elongate directly, and yet continuous at its posterior end with two stripes of indifferent tissue. These are accordingly, under the general influence of the regulatory tendency, organized through the action of the struc- tures in front, precisely as the growing tail bud is organized in the normal embryo. The difference is that in the spina-bifida embryo the entire stripe is there from the start, while the cord of cells, to which the backwardly growing tail bud gives rise, is formed and metamorphosed into its several derivatives gradually. There are still other ways in which we may conceive of spina- bifida embryos as being formed. Kopsch (’96, ’99, ’04) and H. V. Wilson (00), for instance, have held that it is not necessary to regard them as due to organization in situ of the blastopore lips, but that they may be produced as the result of a progressive splitting of the axial body. This explanation may well apply to some cases. It is plain that spina-bifida embryos of themselves will not tell us whether they have been built up by normal or abnormal proc- esses. If the blastopore lip organizes, we may claim with one school that it is a normal process, with the other that the lip has been activated perhaps in the manner described above. If it does not organize, as may happen, we may adduce this as proof that there is no normal impulse in it toward organization, or with the other school that the normal impulse has been inhib- ited. In the midst of this debating one does not forget, however, that the proximate, so-called practical, thing to do is to learn how to control the differentiation of the lip, to call out in it, or pre- vent perhaps, the formation of a neural tube, notochord, and somites. ready differentiated cells’’ into the indifferent mass, falls in Roux’s category of ‘dependent differentiation’ (‘abhingige Differenzirung’). The directive force Roux designates ‘an assimilative and differentiative action,’ (loc. cit., p. 509). To it is applicable the term ‘morphological assimilation’ (Roux, ’12, p. 28), which might be used, pending the coinage of an appropriate Greek word. Other embryologists had already postulated the existence of such an influence in nor- mal development, but Roux, I believe, for the first time gave precision to the idea. THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 30, No. 2 186 H. V. WILSON AND BLACKWELL MARKHAM Whether in ordinary development there is a virtual concres- cence of lateral blastopore lips in the midline, and eventual organ- ization of these, is a question which is most definitely answered not by a study of monsters, but by analytical studies on the nor- mal embryo. Some years ago one of the present writers (Wilson, 00, ’01) published data which together with the work of O. Schultze (ref. in Wilson, loc. cit.), Assheton (ref. in Wilson, loc. cit.), Kopsch (00 and earlier papers), Eycleshymer (’02 and ear- lier papers), Ikeda (’02), seem to make the concrescence theory untenable for the amphibian egg, since they necessitate the con- clusion that a considerable part of the dorsal axial body is formed in situ, viz., in front of the dorsal blastopore lip where it first forms. That the posterior part of the dorsolateral wall of the embryo (gastrula) is produced by the backward growth of the corresponding part of the blastopore lip is admitted by everyone, but it is plainly arbitrary to construe this as a modified form of concrescence. All parts of the blastopore lip grow backward, dorsal, lateral, and ventral. The difference in the distance cov- ered by dorsal and ventral lips in the frog is, to be sure, consider- able, but this inequality is readily understood as a part of the general asymmetry in gastrulation—an asymmetry due to the acquisition and distribution of yolk in the egg, as Balfour long ago pointed out. Likewise the data adduced for fishes by Morgan (’95), Kopsch (Kopsch’s splendid study published 1904 contains references to his earlier papers), and Sumner (’00, 703) make it practically im- possible to believe that concrescence normally occurs in these forms. , . That the concrescence theory may contain truth as a phylo- genetic theory, in other words, that concrescence may have actu- ally occurred in the evolution of the distinctly bilateral metazoa from coelenterate-like forms, remains of course possible. And with richer and more precise knowledge, such questions doubt- less will be taken up again in the future. Lereboullet is often cited as an adherent—in fact, as the first promulgator—of the concrescence theory. But this is to read into his paper (’63) an interpretation that is not, I believe, war- REGULATION IN ANURAN EMBRYOS 187 ranted. He recognizes that the fish body is first marked out as a projection reaching forward from the blastodermic rim. This ‘tubercle,’ in normal development as in double-embryo formation where two tubercles are formed, elongates to form a linear body (‘bandelette embryonnaire’). Lereboullet speaks of the pro- duction of the original tubercle as due to a sort of ‘vegetation’ from the blastodermic rim (a description that is clearly not war- ranted). As to the processes involved in its elongation, he is not explicit. But it would seem that he means that the tubercle elongates by its own growth, possibly incorporating the neigh- boring tissue of the blastodermic rim (which no one would deny). By contrast, in his fourth kind of anomaly, what we call to-day the spina-bifida embryo, the original tubercle does not grow and elongate, but the lateral blastodermic edges organize and par- ticipate directly, as such, in forming the body. In fact, in his search for the growth processes that lead to his various anom- alies, he gives the impression of having distinctly in mind the power of the embryo to proceed to the type structure in more ways than one. Whereas, some of the later embryologists, like O. Hertwig, assume for the moment at least that ontogeny always exhibits the same series of events, anomalies being due to retard- ation or acceleration in the occurrence of particular events. While undoubtedly many anomalies are produced in this way, there are as certainly others the causation of which is much more subtle and complex. SUMMARY Bufo and Chorophilus embryos occur in which, when blastopore closure is inhibited, an asymmetrical regulatory process comes into activity. Instead of the two lateral blastopore lips fusing in the midline, the blastopore is shifted over toward one side, and from a single lip a backward extension of the axial organs is pro- duced. Such a tadpole was reared to a stage in which external gills had been absorbed, internal gills and opercular cavity formed. University of North Carolina Chapel Hill, North Carolina 188 H. V. WILSON AND BLACKWELL MARKHAM LITERATURE CITED Enpres, H., anp Watter, H. E. 1895 Anstichversuche an Eiern von Rana fusca. Erster Theil. Archiv f. Entw.-mech., Bd. 2, Heft 1. Enpres, H. 1896 Anstichversuche an Eiern von Rana fusca. Zweiter Theil. Ibid., Bd. 2, Heft 4. EyYcLEsHYMER, A. C. 1902 The formation of the embryo of Necturus, with re- marks on the theory of concrescence. Anat. Anzeiger, Bd. 21. Hertwic, O. 1892 Urmund und Spina bifida. Archiv f. mikr. Anat., Bd. 39. Ikepa, Saxusrro. 1902 Contributions to the embryology of amphibia: The mode of blastopore closure and the position of the embryonic body. Journ. Coll. Science, Imp. Univ. Tokyo, vol. 17, part 2. JENKINSON, J. W. 1909 Experimental embryology. Oxford. Kopsco, Fr. 1896 Exper. Unters. i. d. Keimhautrand d. Salmoniden. Ver- hdlg. d. Anat. Gesellsch. zu Berlin. 1899 Die Organisation d. Hemididymi, etc. Intern. Monatsschr. f. Anat. u. Physiologie, Bd. 16. 1900 Uber das Verhiiltniss d. embryonalen Axen zu d. drei ersten Furchungsebenen beim Frosch. Ibid., Bd. 17. 1904 Unters. ti. Gastrulation u. Embryobildung bei den Chordaten. I. Die morphologische Bedeutung d. Keimhautrandes u. die Embryo- bildung bei der Forelle. Thieme, Leipzig. LEREBOULLET, A. 1863 Recherches sur les monstruosités du Brochet observées dans l’oeuf et sur leur mode de production. Ann. des Sci. Naturelles, 4me série, Zoologie, T. 20. Morean, T. H. 1895 The formation of the fish embryo. Jour. Morph., vol. 10. Roux, WitHELm 1888 Uber die kiinstliche Hervorbringung ‘halber’ embryo- nen durch Zerstérung einer d. beiden ersten Furchungszellen, etc. Virchow’s Archiv, Bd. 114 (Roux, Gesammelte Abhandlungen ii. Entw.- mech., Bd. 2, to which page references are made). 1912 Terminologie der Entw.-mechanik. Leipzig. Sumner, F. B. 1900 Kupffer’s vesicle and its relation to gastrulation and con- crescence. Mem. New York Acad. Sci., vol. 2, part 2. 1903 A study of early fish development. Experimental and morpho- logical. Archiv f. Entw.-mech., Bd. 17. Witson, H. V. 1894 Observations on the gemmule and egg development of marine sponges. Jour. Morph., vol. 9. 1900 Formation of the blastopore in the frog egg. Anat. Anzeiger, Bd. 18. 1901 Closure of the blastopore in the normally placed frog egg. Ibid., Bd. 20. Resumen por el autor, Bennet M. Allen. Universidad de Kansas. Los resultados de la extirpacion temprana de las glandulas timo de la larva de Rana pipiens. El autor ha extirpado cortando, los esbozos de las glandulas timo de renacuajos de Rana de 8 a 8.5 mm de longitud, después de la cual los animales operados sanaron raépidamente a pesar de la grave operacién doble que fué necesario practicar. Se han notado los siguientes hechos: 1) Las glandulas timo, desde su primera aparicién, no ejercen influencia ni sobre el desar- rollo ni sobre el progreso de la metamorfosis. 2) No son indis- pensables para la vida del animal en ningtin estado del desar- rollo, ni su extracion parece causar una deficiencia marcada en el metabolismo general del cuerpo. 3) La extirpacién de las glandulas timo no modifica en modo alguno ni la cualidad ni la marcha y grado de desarrollo de las gonadas. 4) Un estudio bastante incompleto de las glandulas tiroides de ranas desprovis- tas de timo no ha podido demostrar su anormalidad en ningun sentido. Se puede pues afirmar que las glindulas timo no ejercen influencia sobre las glindulas tiroides. 5) No se ha podido comprobar la existencia de modificaciones de los rasgos externos e internos como resultado de la ausencia de las glandulas timo. Translation by José F. Nonidez Carnegie Institution of Washington AUTHOR’S ABSTRACT OF THIS PAPER ISSUED BY THE BIBLIOGRAPHIC SERVICE, DECEMBER 29 THE RESULTS OF EARLIEST REMOVAL OF THE THYMUS GLANDS IN RANA PIPIENS TADPOLES BENNET M. ALLEN Department of Zoology, University of Kansas ONE FIGURE It is a matter of common knowledge that we are much in the dark regarding the function of the thymus gland. Results of extirpation and of thymus feeding have been very conflicting. Gudernatsch (712, °14) claimed that administration of thymus glands as food to Rana tadpoles retards metamorphosis and at the same time stimulates growth. It has seemed to the writer that Gudernatsch killed his tadpoles prematurely—that he should have carried out his experiment at least a month longer than he did. Romeis (’14) gained rather indifferent results upon this question. Swingle (17), working in this laboratory, repeated these experiments with uniformly negative results. He used fresh thymus glands of young calves and found that his tadpoles thus fed all underwent metamorphosis at the usual rate and in typical fashion. Uhlenhuth (718), in earlier researches, found that feedmg mammalian thymus to larvae of salamanders caused actual increase in growth in some while others showed slight re- tardation. He concluded that this was due not to qualitative, but to quantitative factors in the food. Later experiments where the diet was exclusively made up of thymus-gland mate- rial showed a decided resultant retardation in growth and differ- entiation. He concluded that the thymus gland is deficient in certain substances essential to growth. In a series of interesting experiments upon salamander larvae he produces tetanus by means of thymus feeding. He concludes 189 190 BENNET M. ALLEN that this is due to the presence in the thymus gland of a tetany- producing substance. He fails to produce tetany by feeding thymus glands to frog tadpoles, and concludes that here the in- jurious influences of the thymus gland are counteracted by the parathyroid glands that here develop much earlier than in the salamanders. The feeding experiments of Paton and Goodall (714) upon young guinea-pigs and rats are also negative. These authors give an extensive discussion of the literature cf the subject. There is a good discussion of this field in general in a recent paper by E. R. Hosking (18). Many experiments have been made upon the extirpation of the thymus gland. j per cent to 1 per cent. Of these concentrations, 1 per cent of the reagent produced coagulation, but concentrations of } per cent to 74, per cent had the opposite effect and produced reversal of the normal gelation. These concentrations prevent segmentation, but the inhibition is only tem- porary, and the eggs segment (although often somewhat irregularly) upon return to sea-water. In the above description many of the less important details of of the experiment were omitted. The following experiment is re- ported more fully: July 22nd. Amyl alcohol. At 11:44 a.m., eggs fertilized five min- utes previously were put into vials containing various concentrations of isoamyl alcohol (isobutyl carbinol). Six vials were used. A con- tained 2 per cent of the alcohol, B 1.33 percent, C 1 per cent, D 0.67 per cent, EH 0.33 per cent, and F 0.17 per cent. At 11:51 a.m., eggs in A and normal eggs were centrifuged simul- taneously, the handle of the centrifuge being turned 50 times in 29 sec- onds. The A eggs were evidently coagulated, for they showed no strat- ification; the normal eggs had the hyaline zone barely indicated. At 12:023 p.m., eggs in B and normal eggs were centrifuged simultaneously at the same speed as in the previous test. They, too, gave no evidence of stratification. At 12:125 p.m., the eggs in D and normal eggs were 226 L. V. HEILBRUNN centrifuged, the handle being turned 50 times in 30 seconds. The D eggs had a hyaline zone extending one-third or more of the distance along the stratification axis. In these eggs the gray cap was appar- ently lacking. In the control normal eggs the original gelation was beginning its normal reversal, and the hyaline zone extended about one-fourth of the stratification axis. At 12:235 p.m. eggs in C and nor- mal eggs were centrifuged simultaneously, the handle being turned 50 times in 30 seconds. Of the C eggs almost all were cytolyzed and showed no stratification. A few, however, showed a hyaline zone one- half to two-thirds of the distance along the axis of stratification. Such a marked liquefaction is often the preliminary of coagulation. In the normal eggs at this time, segmentation was beginning, and the hyaline zone was poorly shown. The eggs in A, B, C, D, did not segment. Those in # had numerous small cells cut off from the margin. This is an appearance typically found in all cases where the concentration of the anesthetic is not quite sufficient to stop segmentation completely. In F, the cleavage was much more nearly normal. These observations were made at about 1:30 p.m. At 12:36 to 12:37 p.m. some eggs from each of the dishes A to F had been transferred to Stender dishes containing fresh sea-water. Those from A were put into a, those from B into b, etc. These dishes were then examined at 1:45 p.m. No segmentation occurred in a, }, c, and in d only abnormal evidences of the segmentation process were found. But in e and f normal segmentation occurred, and motile blastulae were later found in these dishes. July 24th. Acetonitrile (methyl cyanide). At 11:353 a.M. eggs were fertilized. Ten minutes later they were placed in vials containing solu- tions of acetonitrile in sea-water. Vial A contained 5 per cent, B 4 per cent, C 24 per cent, D 2 per cent, H 1 per cent, F % per cent, G 4 per cent. At 11:52 a.m., eggs in B and normal eggs were centrifuged simultaneously the handle being turned 50 times in 28 seconds. The B eggs had a hyaline zone one-third to one-half the extent of the egg axis, whereas the normal controls showed barely a trace of hyaline zone in a few cases. At 11:59 a.m., eggs in A and normal eggs were centri- fuged at the same speed as in the previous test. A few of the A eggs showed a hyaline zone extending along one-third of the egg axis, but in most cases the eggs were coagulated and showed no stratification. The control of normal eggs exhibited no stratification, except in a few cases which showed trace of a hyaline zone. At 12:05 to 12:074 p.m. some eggs were transferred out of A, B, C, D, E, F, G, into normal sea-water in Stender dishes a, b, ¢, d, e, f, 9, respectively. At 2:15 to 2:30 p.m. eggs in A to G and a to g were ex- amined. In A and B, no segmentation of any kind had occurred. In C and D, there was no normal segmentation, but in many of the eggs small cell-like masses had apparently been cut off from the cell periph- ery. In EL, 10 per cent of the eggs had segmented in more or less normal fashion, others had segmented abnormally. In F and G, nor- mal segmentation had occurred generally. In a, 20 per cent of the CELL DIVISION—-SPINDLE IN SEA-URCHIN EGGS DT. eggs had segmented, some of these had stopped at the two-cell stage, but others had gone on. In b, 98 per cent had segmented, and almost all were normal. In c to g normal segmentation of course occurred. At about 7 p.M., a to g were examined for blastulae. In a, about 2 per cent of the eggs had developed motile blastulae. In 6b to g, prac- tically all the eggs had developed to motile forms, only immature eggs remaining on the bottom of the dish. The eggs in A to G were then examined. No motile blastulae were found in A to #, in F and G, mo- tile blastulae were abundant. July 28th. Ethyl nitrate. Eggs were fertilized at 11:50 a.m. and at 12 m they were placed in vials A to F. Vial A contained 0.5 per cent ethyl nitrate in sea-water, B 0.4 per cent, C 0.3 per cent, D 0.2 per cent, E 0.1 per cent, and F 0.05 per cent. At 12:063 p.m., eggs in C and normal fertilized control eggs were cen- trifuged simultaneously, the handle being turned 50 times in 30 seconds. The eggs in C had a hyaline zone extending through half the egg, in the normal eggs the hyaline zone was not distinct, but was indicated in a fourth of the egg. At 12:19 p.m., eggs in A and normal eggs were cen- trifuged simultaneously, the handle being turned 50 times in 30 seconds. The eggs in A showed a hyaline zone extending halfway. In the nor- mal eggs, the cytoplasmic gelation had reversed, and the hyaline zone extended one-third of the distance along the axis of stratification. At 12:30 p.M., some eggs were removed from each of the vials A to FP and transferred to normal sea-water in Stender dishes a to f, respec- tively. In 6 to f normal segmentation occurred and motile blastulae were produced. In a, some eggs segmented normally, others abnor- mally; a few motile blastulae resulted. Of the eggs which remained in A to F, those in A, B, C, did not segment, those in D, E, F, seg- mented abnormally. No blastulae were produced in A to F. It is scarcely necessary to mention other experiments of the same sort which were performed with the various substances pre- viously enumerated. All fourteen substances, in suitable con- centration, prevent the appearance of the mitotic figure without otherwise injuring the egg. Those concentrations which act in this way are the very ones which inhibit gelation and preserve the fluid condition of the cytoplasm. Perhaps the action of these substances depends upon their lipoid-solvent action. This action does not appear to be exerted on the plasma-membrane, as many students of anesthesia have thought, for the vitelline membrane, which I have shown (’15) to be the plasma membrane of the unfertilized egg, is, as far as I can determine, morphologically unaffected. Similarly, the hyaline layer, which becomes the plasma membrane soon after THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 30, wo. 2 228 L. V. HEILBRUNN fertilization, shows no signs of alteration. On the other hand, the lipoids of the egg interior are oftentimes visibly changed. This is easily seen after the egg is centrifuged, for the cytoplasmic lipoids then become aggregated at one pole of the egg, forming there a small polar accumulation known as the gray cap.. After the egg had been treated with one of the substances used in the above experiments, it was often noted that the gray cap appeared pale and indistinct. Sometimes the gray cap was apparently absent. Thus these substances which prevent gelation possibly produce their effect by acting on the lipoids of the egg. Even before Wilson had shown that ether prevented the ap- pearance of asters and spindle, O. Hertwig had made the observa- tion that this effect could be produced by low temperatures (—2° to —3°) without otherwise injuring the egg. My views, therefore, demanded that such low temperatures have a liquefy- ing effect on the cytoplasm. This was fully borne out by ex- periment. June 24th. At 4:36 p.m., eggs fertilized sixteen minutes previously were exposed to a temperature of —3°. Fifteen minutes later (at 4:51 P.M.), the eggs were removed from the cold, and after an interval of two minutes they were centrifuged simultaneously with control eggs, also fertilized at 4:36 p.m., but not exposed to cold. The handle was turned 40 times in 30 seconds. On examination, the untreated control eggs showed no stratification whatever, whereas the eggs exposed to cold showed the various stratification zones plainly. Not only does cold exert an antigelatinizing action on fertil- ized eggs, but it has a similar liquefying effect upon the cytoplasm of unfertilized eggs. June 24th. Some unfertilized eggs were exposed to a temperature of —3°, and after ten minutes they were centrifuged simultaneously with normal eggs, the centrifuge handle being turned 21 times in 15 seconds. Both sets of eggs showed stratification. In the normal eggs, however, the granular zone was not distinct from the pigment zone, whereas in the cold-treated eggs the pigment granules had migrated more completely, thus effecting a separation between granular and pigment zones. The question now arises as to how this antigelatinizing effect of cold is produced. Obviously it cannot act act as a lipoid sol- vent. The idea suggested itself, however, that cold might pro- CELL DIVISION—-SPINDLE IN SEA-URCHIN EGGS 229 duce the same effect that fat solvents do, but in quite a different way. Possibly low temperatures tend to precipitate the fat globules out of the cytoplasmic emulsion. Such a precipitation might produce an effect comparable to that of the lipoid solvents, for either precipitation or solution would tend to remove the lipoids from their emulsified state. If cold and lipoid solvents both produce their effect by acting on the lipoids of the cell, it is evident that these effects, instead of being complementary, would be antagonistic. Actual experiment .demonstrated an antagonism between cold and ether. Eggs treated both with cold and ether showed less antigelatinizing effect than when treated with cold alone or with ether alone. EFFECT OF INCREASED GELATION The preceding discussion was concerned with the effects of various antigelatinizing agents upon the cytoplasm. It might be interesting to mention briefly some other experiments with sub- stances which tend to intensify the normal gelation. Many authors have investigated the effect of hypertonic solutions on dividing eggs. Loeb (’92) and Norman (’96) found that if the hypertonic solution was sufficiently strong, cleavage was stopped. Oftentimes, nuclear division without cytoplasmic division resulted. The effect of hypertonic solutions was investigated by the cen- trifuge method, and in all cages an intensified gelation could be demonstrated. | Thys. is apparently especially marked in the cor- tex of the egg, and it is probable that such a cortical gelation is the main factor which inhibits division of the cell, even when the nucleus is still able to divide mitotically. July 5th. Eggs were fertilized at 10:484 a.m. At 11:08 a.m. some of these eggs were placed in Stender dishes AaB Cy Distt: A contained 40 cc. sea-water plus 2 cc. 24 N NaCl B contained 40 ce. sea-water plus 4 cc. 25 N NaCl C contained 40 ce. sea-water plus 6 ec. 25 N NaCl D contained 40 ce. sea-water plus 8 ec. 25 N NaCl E contained 40 ce. sea-water plus 10 cc. 25 N NaCl 230 L. V. HEILBRUNN At 11:17 a.m., eggs from C were centrifuged simultaneously with a control lot of untreated fertilized eggs, the handle being turned 50 times in 29 seconds. Whereas the normal eggs had the hyaline zone well indicated, the eggs from C showed not a trace of stratification. At 11:254 a.m., eggs from B were compared with normal control eggs. The centrifuge handle was turned 50 times in 29 seconds. The normal eggs showed a hyaline zone extending a third of the distance along the axis of stratification. The B eggs showed not a trace of a hyaline zone. At 11:364 a.m., eggs from A were compared with normal eggs. The centrifuge handle was turned 50 times in 28 seconds. The normal eggs showed a prominent hyaline zone, extending at least a third of the distance along the axis of stratification. In the A eggs, the hyaline zone was barely indicated. At 11:53 a.m., the eggs from D and from F were centrifuged, the handle being turned 50 times in 30 seconds. Neither eggs from D nor those from £ showed any trace of stratification. Thus it is evident that the addition of hy pertonic NaCl to sea-water, has the effect of intensifying the gelation of the egg cytoplasm. At 12m, the control of untreated fertilized eggs contained eggs in the two-celled stage, but there was no segmentation in A, B, C, D, E. At2-p.m., these dishes were again examined. At this time cleavage was occurring in A. In B there was nuclear division without cytoplas- mic division. In C, D, and E there was neither nuclear nor cytoplasmic division. At 2:24 p.m., eggs from B and from C were centrifuged, the handle being turned 50 times in 28 seconds. In the C eggs the hyaline zone was indicated in some eggs, but not very clearly. In the B eggs the hyaline zone was prominent, extending through about one-third of the egg. But it was not very transparent, for a cortical zone of granules covered it. This indicates a cortical gelation. It will be remembered that the B eggs are those in which nuclear division without cytoplasmic division was found to take place. The effect of potassium cyanide is worth recording. Even in very dilute concentrations, the cyanide inhibits cell division. But, curiously enough, the early phases of mitosis are able to continue in such concentrations. Likewise, the final stages of the process can go on. The explanation that I would offer is a very simple one. Potassium cyanide intensifies gelation and the normal gelation is rendered irreversible. This can be shown true by viscosity measurements.! 1 Moreover, it is directly in line with my previous observation that cyanide prevents swelling of the gel which forms the vitelline membrane of the egg. CELL DIVISION—SPINDLE IN SEA-URCHIN EGGS Dol In order to determine the concentration of KCN necessary to suppress division, I made a 0.004 per cent solution? in sea-water by diluting a 1 per cent solution (in distilled water) with 249 parts of sea-water. This 0.004 per cent solution was successively diluted with equal parts of sea-water until nine solutions were obtained, each half the concentration of the preceding one. Eggs were subjected to all these solutions, and it was found that a concentration of 0.000625 per cent sufficed to check segmenta- tion. In concentrations of 0.0000313 per cent and 0.0000156 per cent, segmentation proceeded to about the four-cell stage, but then went no further. The following experiment shows that the early stages of mito- sis can occur in concentrations much above those which inhibit the entire process: June 25th. Eggs were fertilized at 5:185 p.m. In these eggs the spindle first began to be visible twenty minutes later (at 5:39 to 5:40 p.M.). At 5:23 p.m., 5:27 p.M., 5:32 P.M., 5:38 P.M., some of the fer- tilized eggs were transferred into Stender dishes A, 6, C, D, respec- tively. Each of these dishes contained 0.0025 per cent KCN, prepared by diluting 1 per cent KCN (in distilled water) with 399 parts of sea- water. The eggs in A were observed at 5:47 p.m. Instead of showing a small nucleus, they showed a large pale spot with a vague border. This spot was often elongated, and probably represented an abnormal spindle. When the eggs in B were examined at 5:58 p.m., they showed a pre-spindle plainly, and they appeared much like the normal eggs. None of the eggs in A, B, C, D, proceeded to develop any further than the spindle stage, and observation at 10:40 p.m. showed them all with spindles, but unsegmented. It might be thought that the above experiment owes its ex- planation to the fact that the cyanide penetrates the eggs slowly, and that only after a time is its influence felt. However the eggs in D appeared to be checked almost immediately. Moreover, a later experiment showed that this interpretation could not be the correct one. In this case the fertilized eggs were put into a 0.004 per cent solution of KCN in sea-water five minutes after fer- tilization. The eggs in the KCN solution were kept in a test- 2 Of course the actual concentration of KCN is not referred to, as there is a reaction between KCN and the salts of sea-water. 232 L. V. HEILBRUNN tube tightly sealed with a rubber stopper, and the test-tube was then exposed to a temperature which varied from 10° to 12°. No spindle appeared while the eggs were in the cold, but when the test-tube was removed from the cold after an exposure of two hours and thirty-eight minutes and warmed with the hand, then spindles soon made their appearance. Thirteen minutes later they could be observed plainly. During their two and one- half hours’ stay in the cold, the eggs must have been thoroughly penetrated by the KCN solution, and yet as soon as they were placed in a warmer temperature, development proceeded as far as the spindle stage. Of course no segmentation occurred in these eggs. The following experiment shows that in a 0.005 per cent solu- tion of KCN, the final stages of mitosis can proceed: August 23rd. Some eggs were fertilized at 8:23 p.m. and they first began to segment forty-two minutes later. At 30, 33, 35, 38, 40, 42 minutes after fertilization these eggs were removed to stoppered test- tubes A, B, C, D, E, F, respectively, each of which contained 0.004 per cent KCN in sea-water. Counts of the segmenting eggs showed in F, of; in E, 845; in D, #8; in C, 2%; m B, 24; in A, 2; normal con- trol eggs, ;%4,. In these fractions the numerator represents the num- ber of segmenting eggs counted, the denominator, the total number of eggs observed. Thus it is apparent that for the last twelve minutes the cleavage process is able to continue in the presence of 0.004 per cent In order to explain this curious action of KCN on one particu- lar stage of mitosis, I have already suggested that the cyanide renders irreversible the normal gelation. Even before the above experiments were performed, I had evidence supporting this view. June 25th. Some eggs fertilized at 2:54 p.m. were at 3:14 p.m. sub- jected to the action of 0.005 per cent and 0.0025 per cent KCN. At 3:34 p.m. eggs in 0.005 per cent KCN were centrifuged simultaneously with normal eggs, the handle of the centrifuge being turned 45 times in 30 seconds. The eggs exposed to the cyanide showed no stratification whatsoever, whereas the normal eggs, as expected, showed a very dis- tinct stratification. A later test of the eggs in 0.0025 per cent KCN gave similar results. Thus the KCN prevents the normal reversal of gelation. CELL DIVISION—-SPINDLE IN SEA-URCHIN EGGS 233 The results with KCN lend additional support to the views already expressed on the relation of the mitotic process to the colloidal changes in the cytoplasm. Cyanide acts by intensifying gelation. Hence, as is to be expected, it does not (in moderate concentration) prevent the early stages of mitosis, and develop- ment proceeds as far as the spindle stage. But that particular stage of mitosis which is associated with a reversal of gelation cannot take place in the presence of the cyanide. Chloretone acts somewhat like KCN. A 0.08 per cent solu- tion checks segmentation, although it does not markedly injure the eggs. Such a solution intensifies the normal gelation. THE NATURE OF THE NORMAL MITOTIC GELATION Karlier in this paper I have attempted to show that the ap- pearance of the mitotic figure is necessarily preceded by a cy- toplasmic gelation. Such a gelation can be artificially produced in unfertilized eggs by various reagents, the best of which ap- parently is a hypertonic solution. The question now arises as to how the gelation occurs normally. The fact that its artificial production appears to be best imitated by hypertonic solutions leads to the suggestion that, similarly in the developing egg, gela- tion and consequent formation of astral rays and spindle is initiated by the abstraction of water from the cytoplasm by the growing pronuclei. Some of the early students of artificial par- thenogenesis thought that the essential step in the initiation of development was the abstraction of water from the cytoplasm. If the normal gelation is produced by an excessive salt concen- tration, then it should be possible to show that gelation produced artificially by hypertonic solutions behaves like the normal gela- tion. To a certain extent this has been done. When the cyto- plasm of the unfertilized egg is gelatinized by hypertonic solutions, such a gelation can be reversed by ether. On the other hand, ether has no effect in reversing or antagonizing the gelatinizing (or coagulating effect) of acids or of distilled water. Hence of these three types of gelation, that produced by hypertonic solu- tions behaves most nearly like the normal. 234 L. V. HEILBRUNN June 28th. At 9:38 a.m., 8 cc. of 24 N NaCl were added to 50 ce. of sea-water containing unfertilized eggs. At 10:39 a.m., some of these eggs were removed to a solution containing 3 per cent ether. This solution was made up by adding to 434 cc. of sea-water, 5 cc. of 25 N NaCl, and 13 cc. of ether. At 10:494, the cytoplasmic viscosity of the eggs in both solutions was compared by a simultaneous centrifuge test, by which it was found that the eggs in the solution containing ether had a more fluid cyto- plasm. The centrifuge was turned 45 times in 28 seconds. After this treatment, the eggs in the hypertonic solution without ether showed no sign of stratification, whereas the eggs which had for ten minutes been exposed to ether (although still in a hypertonic solution) showed the beginnings of stratification. In them, the pigment granules had shifted somewhat, the gray cap and hyaline zone were beginning to appear. That ether exerts an antagonistic action toward hypertonic solutions was shown in still another way. As has been previ- ously pointed out, 2} per cent ether reverses the normal gelation and thus prevents the appearance of the mitotic figure. How- ever, when 23 M NaCl was added, although the concentration of ether remained the same, the ether was no longer able to repress the formation of spindles and asters. In several experiments I tried to discover if ether would pre- vent the gelation of the egg cytoplasm by distilled water or by acid. Rather than a decrease, the addition of ether apparently produced a slight increase in the gelatinizing power of dilute acid solutions. The following experiment serves as a sample: July 20th. At 5:20 p.m., unfertilized eggs were placed into Stender dish A, which contained 50 ce. of sea-water plus 1.3 cc. 75 HCl, and also into Stender dish B, which contained 50 cc. of 2 per cent ether dissolved in sea-water + 1.3 ce. %, HCl. At 5:25 p.m. the cytoplasmic viscosity of the eggs in A and B was simultaneously tested with the centrifuge, the handle being turned 35 times in 30 seconds. Many of the eggs from A were injured. The intact eggs showed a hyaline zone extending one- fourth of the distance along the axis of stratification. The eggs from B were coagulated and showed not a trace of stratification. At 5:35 p.m., eggs from A and B were again centrifuged, the handle being turned 35 times in 28 seconds. This time both sets of eggs were coagulated, and in neither case was any stratification visible. The following experiment, although perhaps not conclusive, indicates that ether does not prevent gelation of egg cytoplasm by distilled water: CELL DIVISION—-SPINDLE IN SEA-URCHIN EGGS 235 June 29th. At 11:38 a.M., some unfertilized eggs were dropped into distilled water, and one minutelater they were transferred out of the distilled water into A, which contained pure sea-water, and B, which contained 23 per cent ether in sea-water. At 11:50, the eggs in A and B were centrifuged simultaneously, the handle being turned 50 times in 30 seconds. When the centrifuged eges from A and B were compared, both lots appeared the same. In both cases most of the eggs showed stratification, with a wide and dis- tinct hyaline zone. In both cases, a considerable number of the eggs were cytolyzed and showed not a trace of stratification. If the normal gelation is due to an abstraction of water, then it should be possible to show an antagonism between cold and hypertonic solutions, whch would be comparable to the effect of cold on the normal gelation. So far, my results in this direction have not been completely successful. Although I have been able to show that cold retards the gelatinizing effect of hyper- tonic solutions on the unfertilized egg, I have not yet demon- strated that cold can cause a reversal of gelation when once this has been produced by hypertonic solutions. But only a single experiment has been tried, and perhaps further observation will also show this to be true. The following experiments shows that cold tends to prevent gelation of the cytoplasm by a hypertonic salt solution: August 30th. A hypertonic solution was prepared by adding 8 cc. of 23 n NaCl to 50 cc. of sea-water. This solution was then divided into two portions, of which A remained at room temperature, and B was kept at a temperature which varied from —1.3° to +1°. At 10:58 A.M., unfertilized eggs were placed in A. At 11:08 a.m., some of the eggs in A were removed to B. At 12:31 p.m., eggs in A and B were cen- trifuged simultaneously, the handle being turned 50 times in 30 seconds. The eggs in B showed a gray cap and a hyaline zone extending about one-third of the distance along the axis of stratification. The eggs in A showed not a trace of stratification. If the gelation which occurs normally is due to water abstrac- tion, it might be also expected that when the egg is made to take up water, this gelation could be reversed. This is in fact true, and hypotonic solutions effectually cause a reversal of gelation in the fertilized egg. This was shown clearly by centrifuge tests. Because of this antigelatinizing effect, hypotonic solutions act like ether and prevent segmentation without otherwise injuring the egg. 236 L. V. HEILBRUNN June 28th. Eggs fertilized at 11:31 a.m., were at 11:464 a.m.. dropped into Stender dish B, which contained 40 cc. sea-water plus 10 cce.dis- tilled water. The untreated fertilized eggs remained in Stender dish A. At 11:513 a.m., eggsin A and B were centrifuged simultaneously, the handle being turned 45 times in 26 seconds. The eggs in A showed just the beginnings of a hyaline zone. The eggs in B were markedly more stratified, they showed gray cap and hyaline zone plainly. In this experiment the sea-water was not sufficiently dilute to prevent segmentation. In another experiment it was found that a solution made up of equal parts of sea-water and distilled water was the most favorable for the reversible prevention of cleavage. In such a solution, eggs remained unsegmented, and yet after a three-hour exposure, they were able to resume their development when returned to normal sea- water. SUMMARY 1. During the period between fertilization and the first cleay- age of the sea-urchin egg, the viscosity of the cytoplasm rises until it reaches a maximum, then it decreases again. 2. Similar viscosity changes occur in, relation to the second cleavage. 3. The changes in viscosity are very marked and indicate the occurrence of.a gelation in the cytoplasm. 4, This gel-formation reaches its height just before the spindle appears. Later the cytoplasm becomes more fluid again. 5. That gelation is not secondary, but is a predetermining fac- tor in spindle or aster formation, is indicated by the fact that when gelation is suppressed, the mitotic figure does not form, although the eggs may be otherwise uninjured. 6. Such suppression of gel formation was produced by four- teen different substances, all lipoid solvents. 7. It can also be produced by cold. 8. Although they produce the same effect, the action of cold and of lipoid solvents is mutually antagonistic. 9. The effect of hypertonic solutions on dividing eggs can be interpreted on the basis of the fact that they increase the cyto- plasmic viscosity. Potassium cyanide and chloretone also act in this way. 1 DS 10. The cytoplasmic gelation which occurs in relation to mito- sis is apparently due to an abstraction of water, for— CELL DIVISION—-SPINDLE IN SEA-URCHIN EGGS 237 a. It can be most closely imitated by an abstraction of water. Cytoplasmic gels produced by hypertonic solutions on unfer- tilized eggs behave toward cold and ether much like the normal gel of the fertilized egg. Cytoplasmic gels produced by acid or by distilled water do not exhibit this resemblance. b. Entrance of water into the fertilized egg reverses the normal cytoplasmic gelation. LITERATURE CITED ALBRECHT, HE. 1898 Untersuchungen zur Struktur des Seeigeleies. Sitzungs- ber. d. Ges. f. Morph. u. Physiol. in Miinchen, 1898, S. 133-141. VAN BENEDEN, E. 1875 La maturation de l’oeuf, la fécondation et les prem- iéres phases du développement embryonnaire des Mammiféres, d’aprés des recherches faites chez le Lapin. Bull. de l’Acad. Roy. de Belgique, 2me ser., T. 40. CuamBerS, R. 1917 Microdissection studies. II. The cell aster; a reversible gelation phenomenon. Jour. Exp. Zodél., vol. 23, pp. 483-505. FiemmMine, W. 1882 Zellsubstanz, Kern- und Zelltheilung. Leipzig 1882 (see especially pp. 206-208). 1891 Uber Theilung und Kernformen bei Leukocyten, und iiber deren Attraktionsphiren. Arch. f. mik. Anat., Bd. 37; S. 249-298. For, H. 1879 Recherches sur la fécondation et le commencement de l’héno- génie chez divers animaux. Mem. d. 1. Soc. de Phys. et d’Hist. Nat. de Geneve, T. 26. Heitprunn L. V. 1915 Studies in artificial parthenogenesis. II. Physical changes in the egg of Arbacia. Biol. Bull., vol. 29, pp. 149-203. 1917 An experimental study of cell division. Proc. Amer. Soc. Zool., Anat. Rec., vol. 11, pp. 487-489. Hertwia,O. 1890 Experimentelle Studien am tierischen Ei, vor, wihrend, and nach der Befruchtung. Jenaische Zeitschr., Bd. 24, 8. 268-313. Levi, G. 1916 La costituzione del protoplasma studiata su cellule viventi coltivate ‘in vitro.’ Archivio di Fisiologia, T. 14, pp. 101-112. Logs, J. 1892 Investigations in physiological morphology. III. Experiments on cleavage. Jour. Morph., vol. 7, pp. 253-262. Norman, W. W. 1896 Segmentation of the nucleus without segmentation of the cytoplasm. Arch. f. Entwicklungsmechanik, Bd. 3, 8S. 106-126. Spex, J. 1918 Oberflichenspannungsdifferenzen als eine Ursache der Zellteil- ung. Arch. f. Entwicklungsmechanik, Bd. 44, S. 5-113. Witson, E. B. 1901 Experimental studies in cytology. II. Some phenomena of fertilization and cell-division in etherized eggs. III. The effect on cleavage of artificial obliteration of the first cleavage furrow. Arch. f. Entwicklungsmechanik, Bd. 138, S. 353-395. Resumen por el autor, Caswell Grave. Amaroucium pellucidum (Leidy), forma constellatum (Verrill). I. Movimientos y reacciones de la larva. El cuerpo de la larva, que como es sabido se asemeja por su forma a los renacuajos de los anfibios, presenta un movimiento de rotacién en el sentido de las agujas de un reloj, alrededor de su eje, durante la locomoci6n, a consecuencia de la forma asimétrica del cuerpo o de la torsi6én del eje de la cola. Las larvas presentan una reaccién positiva definida hacia la luz durante un intervalo muy breve que sigue inmediatamente a su liberaci6n de la colonia que las produjo, pero reaccionan negativamente hacia la luz durante la mayor parte del periodo de actividad ulterior. Las larvas permanecen invariablemente en la superficie del agua o cerca de ella durante la primera parte del periodo de actividad, pero al aproximarse el tiempo de la metamorfosis descienden a las capas inferiores y nadan en el fondo o cerca de él. La reac- cidn positiva de la gravedad se presenta con varios grados de intensidad, faltando completamente en un pequefio tanto por ciento de las larvas. Hacia el final del periodo de actividad, el contenido viscoso de los extremos glandulares de las papilas adhesivas se vierte sobre la superficie externa de la ttinica, y la fijacién inicial de la larva acaece cuando una o varias de estas gotitas viscosas se ponen en contacto con la superficie de un cuerpo extrafo. La duracién del periodo de natacién libre de la larva varia desde diez minutos hasta dos horas. Translation by José F. Nonidez Carnegie Instituion of Washington AUTHOR’S ABSTRACT OF THIS PAPER ISSUED BY THE BIBLIOGRAPHIC SERVICE, FEBRUARY 2 AMAROUCIUM PELLUCIDUM (LEIDY) FORM CONSTELLATUM (VERRILL) I. THE ACTIVITIES AND REACTIONS OF THE TADPOLE LARVA CASWELL GRAVE Washington University, St. Louis, Missouri FOUR FIGURES A study of the organization, activities, and metamorphosis of the free-swimming larva of Amaroucium pellucidum constellatum was begun during the summer of 1912 at the Marine Biological Laboratory at Woods Hole, Massachusetts, and has been con- tinued as opportunities have permitted. It has also included work on the structure and asexual reproduction of the primary ascidiozooid and the formation of colonies and the differentia- tion of germ cells by the secondary ascidiozooids. | The observations made on the activities and reactions of the tadpole during its free-swimming period are recorded in this paper. The histological structure of the organs involved in larval activities will be described and figured in a paper now being prepared for publication on the organization of the Ama- roucium tadpole. LOCOMOTION Since the discovery of the chordate affinities of ascidians by Kowalevsky, in 1866, a closer similarity in the behavior, as well as the fundamental structure, of the ascidian tadpole with chordate animals has been taken for granted than is apparently warranted. In the numerous papers which have been published on the development and metamorphosis of ascidians very few observations on the behavior of the free-swimming larva have been recorded. MacBride,! in his Text Book of Embryology, page 619, states that the tadpole larva propels itself like a fish 239 240 CASWELL GRAVE by lateral blows of its tail. In the course of my study of the Amaroucium tadpole, it was a great surprise to find that this larva does not swim in the fashion of a vertebrate, in which a constant position of the body is maintained, but that the body of the tadpole while swimming is in constant and rapid rotation on its long axis, clockwise as seen from behind. I have found this method of locomotion also in the tadpole larva of Botryllus and suspect it is characteristic of the larvae of ascidians in general. On account of the suddenness with which the tadpole begins to swim after one of its quiescent intervals, and the rapidity with which it passes through and out of the field of vision with the microscope, the rotary movement of the body during locomotion is difficult to observe; however, by the addi- tion of a narcotizing agent to the water, such as a solution of Epsom salts, the rapidity of the swimming movements of the tadpole is gradually reduced and, when moving slowly, the revolutions of the body are readily noted. The mechanism by which the body is caused to rotate is also not immediately evident. The body of the tapole is somewhat compressed and, when at rest, comes to lie on one of its flattened sides. For convenience in description, these flattened surfaces will be referred to as the right and left sides. The side of the body containing the rudiments of the oral and atrial siphons and the pigmented sense organs will be referred to as dorsal, the tail as posterior in position, and the end opposite to tail, containing the adhesive papillae, as anterior. It is generally assumed that the tail fin of the ascidian tadpole is expanded vertically, as in vertebrates, and in general this may be the case, but the tail fin of the Amaroucium tadpole is horizontal in position. Seeliger? noted that the tail of the larva of Clavelina lepadi- formis, in consequence of its forward growth along the side of the body beneath the closely fitting chorionic membrane, be- comes twisted and that the nerve tube is thus turned from the dorsal to the left side. According to his descriptions and figures of the free-swimming tadpole, however, it follows that the tail untwists when the chorion is burst, the nerve tube assuming the ACTIVITIES OF THE AMAROUCIUM TADPOLE 241 dorsal, the tail fin the vertical position. Seeliger states also that this larva swims in the manner of (amphibian) tadpoles. The slight asymmetry which exists between the right and left sides of the body may account for the rotation of the body as it is propelled through the water by the lashing of the tail. Viewed from the dorsal side, a concave depression is seen on the left located near the anterior end; moreover, the anterior tip of the body, containing the middle adhesive papilla, is found to lie slightly to the right of the median sagittal plane (fig. 1). These asymmetrical features are the result of the pressure of the tail which, during the embryonic period of development, is folded forward beneath the chorionic membrane and coiled about the anterior part of the body. The depression takes an oblique ‘course from below upward across the left side of the body and therefore gives to it the form of a screw with a single groove which tends to set up an axial rotation in the observed left to right direction when the body is propelled rapidly through the water. Considerable support for the suggestion that the screw-like form of the body plays a part at least in producing rotation during locomotion is found in the fact that when a tadpole sud- denly stops swimming, as it frequently does during its active period, the axis of the body quickly assumes a vertical position, tail upward, and begins to revolve slowly from left to right as it slowly sinks through the water. The lateral asymmetry of the body may not be the only or even the chief factor in producing rotation during locomotion. The direction and character of the strokes of the tail may be such as to cause rotation, and my observations on the character of the movements of tadpoles held captive beneath a cover-slip, and on tadpoles in which a part of the tail has been amputated, lead me to believe the strokes of the tail are not made directly to the right or left, but that the contractions of the muscle fibrillae, due to their orientation in the muscle cells, tend to slightly twist as well as to bend the axis of the tail, thus produc- ing a spiral thrust. The disposition of the contractile fibrillae in the cortical layer of the muscle cells is shown in figure 2. As 242 CASWELL GRAVE tg ines ng Norn map .-- dap L R OMa= ee -Sv ~- Lpo a s---- ie oP ie mc Fig. 1 Camera outline drawing of the fully developed tadpole larva of Ama- roucium pellucidum constellatum as seen from the dorsal surface, showing the lateral asymmetry of the body, the horizontal position of the tail fin, and the location of the sense organs in the sensory vesicle. ACTIVITIES OF THE AMAROUCIUM TADPOLE 243 - was noted and figured by Kowalevsky* and Seeliger, the muscle fibrillae take a slightly oblique course in each cell and the fibrillae of adjacent muscle cells seem to be continuous from cell to cell. When a tadpole is lightly held between a slide and cover-slip, the vibrations of the tail cause the body to oscillate, the anterior end seeming to move in a circular or oval path on a pivot located near the posterior part of the body. When a part of the tail is amputated, the vibrations of the remaining stump do not pro- —. Fig. 2 Camera outline drawing of a muscle cell, showing the oblique position of the fibrillae in the cortical layer, the continuation of the fibrillae from cell to cell, and the cross striations in one fibrilla. ABBREVIATIONS (FIG. 1) as, atrial siphon R, right side dap, dorsal adhesive papilla sc, statolith cell _ L, left side sv, sensory vesicle lpo, light-perceiving organ tf, tail fin map, middle adhesive papilla tg, groove in test resulting from pres- mc, muscle-cell sheath sure of the tail during embryonic n, notocord development os, oral siphon tv, test vesicles THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 30, NO. 2 244 CASWELL GRAVE duce locomotion, but throw the body into vibration with the appearance of oscillatory movement the same as that seen when locomotion in a normal tadpole is prevented by confining it between slide and cover. The rotary method of locomotion in the ascidian tadpole may be accidental and have no special significance, but it is of some interest to find its method of swimming is that characteristic of the early larvae of invertebrates in general and not like that of vertebrates. REACTIONS TO LIGHT When colonies of Amaroucium are placed in a rectangular glass jar lined except for one side with unglazed black paper and the jar is so placed that light enters the water through the uncovered side, the tadpole larvae, as they are liberated, swim through the water first in an undulatory course obliquely upward toward the light, then up along the side of the glass to the surface of the water where they dodge back and forth for a few seconds attempting to proceed toward the source of the light. They then leave the illuminated side and swim into the less — illuminated parts of the jar, where they remain, alternately active and quiescent, throughout their free-swimming period. Castle! noted that “Ciona and Amaroucium tadpoles avoid the light while that of Botryllus swims toward lght.” He failed to observe the Amaroucium tadpole during the brief interval when, on escaping from the parent colony, it shows a positive light reaction. Occasionally a tadpole, in its initial excursion toward the light, will turn in its course and swim away before it has reached the lighted side of the jar. That is, the period of positive reac- tion to light in some tadpoles is remarkably short, but is appar- ently never entirely absent. Tadpoles do not swim in a straight course either toward the source. of light, during the brief period when they react posi- tively to it, or away from its source during their longer period of negative response, but depart from a straight line in all direc- tions and, at times, they travel in circles and curves of con- ACTIVITIES OF THE AMAROUCIUM TADPOLE 245 siderable magnitude, as is shown by the diagram (fig. 3). When observing the movements of a single tadpole, especially during the period of its negative reaction to light, it is difficult to follow the exact path and trend of its course, but, when working in a dark-room with a large number of tadpoles in a rectangular dish with electric light bulbs at opposite ends of the dish, the nega- wo Or oO Fig. 3 Showing three characteristic paths taken by tadpoles upon emerging from the parent colony during exposure in a rectangular jar 12 cm. in width, 18 cm. in length, and 25 em. in depth, to bright directive light reflected from a south window, 12 feet distant, as seen from above. The sides of the jar, mo, op, and pn, lined on the inside with black unglazed paper. J and JJ, Amaroucium colonies on the bottom of the jar. a, b, and c, points at which tadpoles emerged. Dotted lines, paths taken by tadpoles. Arrows indicate width and direction of beam of light. tive response of the group as a whole is strikingly definite and immediate. The entire group moves slowly away from the source of light, changing its direction the instant the source of the light is reversed. The light-perceiving organ of the tadpole, composed of a series of three lenses, a cup-shaped layer of pigment granules, and a 246 CASWELL GRAVE group of retinula cells, is not located in the median sagittal plane of the body, but is displaced to the right and so oriented that those rays of light only which enter the series of lenses from the upper right side of the body will reach the pigment cup and be effective in stimulating the sensitive ends of the retinula cells (fig. 1). In its natural habitat, light rays of greatest intensity reach the tadpole from the direction of the water surface and, as a consequence of the axial rotation of the body during loco- motion, the tadpole receives its maximum stimulations from light at one instant only in each revolution when the body is so oriented that its long axis is parallel to the water surface. It is thus obvious that a series of frequent orienting responses is provided for. RESPONSE TO GRAVITY During the first part of their free-swimming period, tadpoles tend to seek and to remain at or near the surface of the water, but later on they swim away from the surface and seem to seek the deeper strata of water and, at the close of the free-swimming period, usually become attached to the less illuminated places on or near the bottom. The tendency of tadpoles to remain at the surface during the first part of their active period is subject to the interpretation that it is simply one of the results of their positive reaction to light, while their behavior a little later, when they exhibit a very definite negative response to light yet continue to remain at the surface of the water, warrants the conclusion that tad- poles are sensitive to conditions other than light, probably to their position with reference to gravity, but possibly to changes in density or water pressure or to differences in the oxygen or carbon-dioxide content of the upper and lower strata of water. I incline to the view that the response of the tadpole is to gravity, because of the presence of a statocyst-like structure within: the sensory vesicle and the similarity of this organ to sensory structures in other animals known to be end-organs for the perception of the position of the body with reference to gravity. ACTIVITIES OF THE AMAROUCIUM TADPOLE 247 In order to determine the character and definiteness of the response of the Amaroucium tadpole to gravity, a number of experiments were carried out, in each of which a large number of tadpoles, collected and used as soon as possible after their emer- gence from the parent colonies, were placed during their free- swimming and attachment periods in a glass cylinder contain- ing a column of water 5.5 cm. in diameter and 40 cm. in depth, note being made of the levels at which the tadpoles became attached. The results of twelve of these experiments are shown in table 1. The special conditions under which each experiment was made are stated in the explanation of the table. While these experiments were designed primarily for the study of the response of tadpoles to gravity, those numbered from 1 to 6 in the table show also the effect upon this response of special or unusual conditions of light, and the experiment num- bered 12 shows the effect of the absence of light in modifying the normal geotropic response. Five of the experiments, those numbered from 7 to 11 in the table, were made under conditions considered as nearly normal as it is possible to make them in the laboratory. The unjack- eted clear glass cylinder stood 2 feet directly in front of a north window on a table the top of which was 53 feet below the top of the window and 8 inches below the window sill. The rays of light entered the column of water more or less obliquely from the general direction of the window and were therefore some- what directive. Combining and summarizing the results of these experiments, it is found that 77.6 per cent of the total 496 tadpoles used in the experiments became attached either to the bottom or to the edge of the bottom on the least illuminated side of the cylinder; 9 per cent to the side below the middle; 7.6 per cent to the side above the middle; 3 per cent to the side of the cylinder at the edge of the water surface, and 2.6 per cent remained floating on the surface of the water, perhaps attached to or held by the surface film. Notes made during the course of one of the experiments will serve to show the nature of the activities observed: 248 CASWELL GRAVE TABLE 1 Showing the levels at which tadpoles became attached when subjected to various light conditions in a 1000-cc. glass cylinder filled to the 40-cm. mark with sea-water SF BB Pa") eat Pee! goat Rat Cat | Manet 1 J Site aR eh oleae Leia ee | , aie te eae eee al Bole al 2 3 2 S| | ee 35 1 | 1 ial jad Sele elie 3 ae 30 3 | os | See 1 ce EY EA) RTE fe CO Ee ee a 25 1 10)! Haale i ee | ee ee eee 20 ft) eS Sihiy 2 1 1 ee 15 2 Wiki A 8 bea a a a a a Re Ee 10 Hh 2 2 4 bale 1 Sa ee fT EE eee ee ee ee 5 1 5 72d Me VA Ya lie) Pail ng: BE 171,30). ts) -40;[. 10) ant 25 1 258 | Fo) 4) Sea ek 1 The number of tadpoles remaining at the top in experiments 3 and 4 seems abnormally large. It is possible that some of the tadpoles were not in good condition due to the fact that the parent zooids had been kept too long under laboratory conditions. ACTIVITIES OF THE AMAROUCIUM TADPOLE 249 September 12, 1917. 4.45 p.m. All tadpoles liberated by an Amaroucium colony during a fifteen-minute interval, seventy-five in number, collected in a glass cylinder and placed upright on a table before a north window. All tadpoles swimming actively at the surface of the water. A tadpole dives downward now and then to a depth of 2 to 3 cm., but returns immediately to the surface. 4.53 p.M. One tadpole dived to a depth of 6 cm. and then returned to the surface. Others do the same. None diving deeper. 4.55 p.m. One tadpole swam straight down to a depth of 32 cm., stopped swimming, turned tail upward, and slowly sank to the bottom, where it remained motionless. 4.58 p.m. One tadpole descended to a depth of 9 em., came in con- tact with the side of the cylinder and became attached, not able to free itself by repeated attempts to swim away. 5.01 p.m. One tadpole swam half way to the bottom, turned and swam to the top. 5.03 p.m. Several tadpoles dived to depths of about 15 em., all but one returning to the top, one sank motionless to the bottom. 5.08 p.m. Several tadpoles swimming about actively at and near the bottom, others at the top and in the upper strata of water, one sinking motionless with metamorphic changes taking place. 5.13 p.m. Several tadpoles swimming at various depths between the top and bottom. EXPLANATIONS, TABLE 1 SF, attached to surface film (floating); SH, attached at line in which water surface and side of cylinder meet; BE, attached along the bottom edge; B, at- tached to the bottom; figures in the columns indicate the number of tadpoles attached at various levels; figures at the bottom of the columns show the total number of tadpoles used in each experiment. Experiments 1 and 2 made simultaneously with tadpoles taken from the same colonies of ascidians. This is the case also with experiments 3 and 4 and with 5 and 6. Experiments 1 and 3 Sides and top of the cylinder covered with black paper. Light from a north window reflected directly upward through the column of water. Experiments 2 and 4 Sides and bottom of the cylinder covered with black paper. Light from a north window reflected directly downward through the column of water. Experiment 5 Same as 1 and 3, except that direct rays from the sun were reflected upward through the column of water. Experiment 6 Same as 2 and 4, except that direct rays from the sun were reflected downward through the water. Experiments 7 and 8 The cylinder without cover set before a north window. Experiments 9, 10, and 11 Same as7 and 8, except the cylinder was inclined at an angle of 45° from the perpendicular. Experiment 12 Cylinder placed in a dark room during the free-swimming and attachment periods of the tadpoles. 250 CASWELL GRAVE 5.30 p.m. a ok et PA SAT" Bins. ag) ait {yet fo) a ‘vualt uh ne 4 gE bien ty i tala. ‘Pint jane ay an geoahiaia Karpathos Olsen Abi * gai Et ar ha Su Rordeadl ees psaoatier 7D) haieetd ale’: diy apa HOE tind ues or Mite bod a meiiit rc : SS Halsid: Tusa Qutub ala eee ws aa ate, aah: ion ‘ ryt. - ‘ve A “$7 a at tng (es 12) fay apn * ¢. se Sect } wat : persire ls gn c hor Ghee fit ; P ‘ ie vie fl or R= i) aie Ms ek ee Coe a oe A Pe) Oe tag) heey ess hy a . i | Pe) g Sat al oe 2g = & ‘ irs ie ;. = 2 ~ rn rss | uo Resumen por el autor, C. H. Edmondson. Universidad de Oregon. Sobre la formacién de un nuevo estilete eristalino después de la extraccién del primitivo en Mya arenaria. La funcién del estilete cristalino de los lamelibranquios, tal cual se considera actualmente, es la de un fermento digestivo, que al ser impelido con un movimiento de rotacién por la accién de las fuertes pestanas vibratiles del saco que le contiene, pene- tra en el estémago donde es digerido por el contacto con el escudo gastrico. El estilete cristalino de Mya es muy resistente, presentando un ligero grado de disolucién después de catorce dias de dieta. El saco del estilete esta casi completamente separado del tubo digestivo, presentando a lo largo un surco manifiesto, limitado a cada lado por typhlosoles, de los cuales el derecho e st’ mas marcado que el izquierdo. Las células de los typhlosoles son mas altas, de menor didmetro, y poseen pestanas vibritiles mas cortas que las de las células que tapizan interior- mente el saco. Las células de los typhlosoles aparentemente segregan mucus. Pr6ximamente el 50 por ciento de los indi- viduos de Mya sobreviven a la extraccién del estilete cristalino. El estilete en vias de regeneracién comienza a aparecer unos seis dias después de la extraccién del contenido anteriormente en el saco. Al principio el estilete es un tracto de mucus arrollado, que contiene particulas alimenticias; después se arrolla hacia la derecha. El arrollamiento y las particulas alimenticias desa- parecen cuando el estilete aleanza la madurez. Aparentemente los animales no ingieren ni digieren alimento normalmente hasta que el estilete es de suficiente tamafo para proyectarse en el interior del estémago. Un estilete en vias de regeneracién puede aparecer en una porcidén del saco enteramente separada del resto. Bajo condiciones favorables la regeneracién completa del estilete de Mya puede verificarse en setenta y cuatro dias. Translation by José I’. Nonidez Carnegie Institution of Washington AUTHOR’S ABSTRACT OF THIS PAPER ISSUED BY THE BIBLIOGRAPHIC SERVICE, DECEMBER 15 THE REFORMATION OF THE CRYSTALLINE STYLE IN MYA ARENARIA AFTER EXTRACTION CHARLES HOWARD EDMONDSON Zoological Laboratory of the University of Oregon THIRTY FIGURES CONTENTS Ila JERAHROYSITWET ELOY Auer FROM. Seam RR a A ae iC EAR ise SA aa UR Mn 5 9) 78 2 Nicihodtor procedure Be SAY Sey SAT eod Wh Ne is SNE Oe) 3. The mature powelline. sive in nea arenaria. es . 266 4. The relation of the crystalline style to the trieerinal pact in i Mayan a arenaria, MHC teAbUreS, Of TMC Spy lerSa Cima ity sei wi tess» sakes + vcs 4bbehe eombenie mn cyeee oo iheseastrictsnield in Neyer arenaria oc) 00... Cee ae... vetoed eee lee eee ete a DeLailsior-the experimentsnss tte eee... eee wile Daal Ua EE BOTS aPbhe sourceyor theverystalline, sty less co iwsd,< 2 ceamle Surtey tocem aati boi ee Se MMM ATV LOC COME TSI OWS: h.th:.3.scis ee aisiagi > ao ORR RAO re Notes Eas EO WIAD LM 1. INTRODUCTION For more than two hundred years after Anton de Heide’s first allusion (1686) to the ‘stylus crystallinus’ in his monograph on the common mussel, the significance of this peculiar organ present in certain molluscs was a matter of mere conjecture. Nelson (’18) presented a very complete review of the many theories advanced by a host of observers in the period during which the crystalline style has been an object of attention. It is to be noted that within the past thirty-eight years, as a result of experimental work, this long list of hypotheses regarding the value of the crystalline style to those molluscs possessing it has been, by a gradual process of elimination, reduced to three, each having the support of investigators of considerable ability. Basing conclusions upon the fact of the dissolution of the crystalline style in certain lamellibranchs during periods of star- vation and hibernation, Hazay (’81), Haseloff (’88), and others have defended the theory that the organ represents reserve food 259 260 CHARLES HOWARD EDMONDSON material stored up during times of excess nutrition and assimi- lated by the animal when needed. The comprehensive investiga- tions of Barrois (89-90), Coupin (’00), Mitra (’00), and later workers have rendered the ‘reserve food’ hypothesis untenable. It has been pointed out that marine lamellibranchs are never without food except for short durations on the recession of the daily tides, moreover, it is well known that, in case of many marine bivalves, the temporary cutting off of the food supply by the recession of the tide is not followed by a dissolution of the crystalline style or by any marked change in it. The writer has kept Mya arenaria alive for fourteen days without food during which period only slight evidences of dissolution of the crystalline style were detected. Coupin (00), in a summary, says: ‘“‘The chemical investiga- tions show that the crystalline style does not contain sugars or fats and only traces of albuminoid materials.”! The same in- vestigator also shows that the crystalline style of Cardium edule when dried weighs but 0.004 gram, and remarks that this would seem an insignificant amount of food for an animal of this size. After ably disposing of the ‘reserve food’ theory championed by Haseloff (’88), Barrois (’89-’90) concludes that the function of the crystalline style is to furnish mucus for the coating of sand grains and other foreign bodies in the stomach thereby preventing injury to the epithelium of the digestive tube as they pass through. Pelseneer (’06) follows Barrois when he says, with reference to the crystalline style: ‘‘The product of its solution forms a sort of cement which encrusts any hard substance that may have been ingested and thus protects the delicate walls of the intestine from injury.” Coupin (’00), as quoted below, not only ascribes to the crystalline style a digestive function, but suggests a lubricating value when he says that the mucus of the style may agglutinate the solid particles which float in the stomach. Schultze (90) agrees with Barrois, drawing an analogy bem een the supposed function of the crystalline style and the secretion of certain glands of the larvae of batrachians which furnish a sub- 1 My translation. REFORMATION OF THE CRYSTALLINE STYLE 261 stance serving to envelop foreign materials thereby protecting the delicate mucous coat of the gills and the intestinal tract. Kellogg (’92) raises an objection to the lubricating theory, be- lieving that the crystalline style could not possibly serve as a coating substance for the large quantity of sand taken in by some species of lamellibranchs. Nelson (’18) has shown that the mu- cus which envelops foreign bodies in the stomach of the molluse, is a product of the glands of the oesophagus, which produce a thick, heavy mucus unlike that furnished by the dissolution of the crystalline style. That the crystalline style might function as an alimentary fer- ment was first suggested by de Heide in 1686. That this early conjecture represents the correct interpretation of the significance of the organ in lamellibranchs has been a growing belief among biologists during the past nineteen years. The convincing in- vestigations of Coupin (’00), Mitra (’01), List (’02), Van Rynberk (08), Gutheil (11), Nelson (18), and others have left no trace of doubt as to the true meaning of the crystalline style apparently so necessary to life and the proper functioning of lamellibranchs. Coupin (00) after testing the action of a solution of crystalline styles upon starch material says: On peut done conclure, de ces expériences, que la tige cristalline des Acéphales est un suc digestif, une sorte de comprimé de diastases, con- tenant beaucoup d’amylase et un peu de sucrose, le tout noyé dans une matiére muquese, laquelle a sans doute pour but d’empécher la trop rapide dilution de la tige dans l’eau de mer contenue dans |’estomac, et peut-étre aussi d’agglutiner les matiéres solides qui flottent dans celui-ci. The weight of authority since 1900 is essentially in accord with Coupin’s conclusion that the crystalline style functions as a digestive ferment. Mitra (’01) differed from Coupin in that he believed the crystalline style represented a solid mass of fer- ments. Van Rynberk (’08), by a series of experiments on the crystalline style of Mytilus, confirmed the results of Coupin and Mitra regarding the presence of an amylolitic ferment and con- cluded that cytases and proteases were absent. Nelson (’18) looked upon the crystalline style as a digestive ferment and has 262 CHARLES HOWARD EDMONDSON shown that the organ rotates, as was suggested by List (’02), in the style sac in a clockwise fashion, when viewed from the an- terior end, thereby serving to assist in the separation of food from foreign substances in the stomach as well as taking the place of intestinal peristalsis. A summary of the function of the crystalline style, as recog- nized at the present time, is as follows: the style is pushed for- ward into the stomach by a movement of the strong cilia of the style sac, at the same time slowly rotating in the direction of the hands of a clock, when viewed from the anterior end. The head of the style, being in contact with the gastric shield of the wall of the stomach, is gradually worn away and, during rotation, as- sisted by movements of the cilia of the stomach wall, separates the refuse matter from the food, twisting the latter up into a mass of mucus at the end of the dissolving style where it may be readily acted upon by starch converting enzymes released from the style. 2. METHOD OF PROCEDURE It is apparent that most of the work of previous investigators, with respect to the formation and function of the crystalline style, has been with those types of lamellibranchs in which there is an incomplete separation of the intestine from the caecum in which the style rests. No published reports, so far as the writer is able to determine, describe the formation of the crystalline style of molluscs in which the style sac is completely or almost com- pletely separated from the intestinal tract. With the hope of adding some small contribution toward the solution of this problem, the experiments, the results of which are set forth in the present paper, were initiated. Working upon the theory that the reformation of the crystal- line style, after extraction from the body of the mollusc, would take place in a normal manner if the animal were kept under natural conditions, it was decided to select a species, remove the crystalline styles from a large number of individuals, restore the animals to their normal surroundings, and by examination at more or less regular intervals follow the formation and develop- ment of the crystalline style in this particular species. REFORMATION OF THE CRYSTALLINE STYLE 263 After considering the list of bivalves on our west coast, Mya arenaria, the eastern long-neck clam, was selected as being the only possible species upon which such an operation as contem- plated might be successfully performed. In the first place, Mya, by reason of its anatomical features, has proved to be exceptionally well adapted for this kind of experimental work. This favorable structural feature lies in the fact that the style sac is exposed along the ventral surface of the visceral mass for a distance of from 20 to 30 mm. in specimens ranging from medium to large size (figs. 1 to 3). Furthermore, on removing Mya from the mud it will be noticed that, in a very large number of individuals, the mantle lobes are protruding some dis- tance between the gaping edges of the shell. The mantle lobes are fused along their anterior and ventral margins except for a small space through which the foot may be extended. It will also be found in most instances that the visceral mass of the body of the mollusc is pressed closely against the ventrally protruding mantle lobes. By cutting the mantle in the midline along the ventral surface for a distance of from 15 to 25 mm. posterior to the pedal opening, the exposed length of the style sac is brought into view. Using fine-pointed scissors, one may easily sever the style sac transversely or longitudinally at the distal extremity or at any level in the exposed length of it. Usually immediately upon the severance of the sac the style will be forced out at the point of division. Oc- easionally the crystalline style will be thrown entirely out of the body by a sharp muscular contraction when the sac is clipped. More often, however, the use of fine-pointed forceps is necessary to draw the style from the sac. In the majority of cases I have severed the style sac and style by a transverse cut at from 10 to 20 mm. from the distal extrem- ity (fig. 3) and removed both ends of the style. The entire op- eration requires but a few seconds of time, and in most individuals must be accomplished immediately after the clam has been taken from the mud, as the mantle lobes are soon withdrawn and the valves of the shell tightly closed, rendering the removal of the crystalline style impossible. 264 CHARLES HOWARD EDMONDSON As soon as practicable after the crystalline style has been ex- tracted, the clams should be replanted in selected areas, and each individual so marked that there may be no possibility of mis- taken identity of the specimen, even though months should elapse before it isreexamined. I have used a system of stakes, by which each individual may be identified, with entire success. The experiments on the reformation of the crystalline style of Mya arenaria were carried on in the Siuslaw River about four miles from its mouth, between the towns of Florence and Acme, where is located one of the most extensive beds of this species on the Oregon coast. Here Mya grows to a very large size, many specimens I have measured attaining a length of shell of 150 mm. The beds in the Stuslaw River are very accessible and work can be done upon them at a moderately low tide, this locality being a most convenient one in carrying on series of rapidly repeated tests or those running through longer periods of time. The physical effect of the operation, as described above, upon the molluse itself may be indicated by a summary of the result of the entire series of experiments. I have found that approxi- mately 50 per cent of the clams die as a result of the severance of the style sac and the removal of the crystalline style. In some tests extending over a period of several weeks the death rate ran as high as 75 per cent, while in others, enduring for a similar period, it was as low as 25 per cent. Unless death occurs during the second or third week following the operation, the clams usually recover and regain their normal functions, the wound in the meantime having healed and the severed ends of the style sac usually bemg completely closed. The success is dependent largely upon the carefulness with which the incision of the style sac is made—a very deep cut destroying the sur- rounding visceral tissue and permitting bacterial infection. Re- sults also show that large, mature individuals survive more readily, apparently having a greater degree of resistance than do young or half-grown ones. In general, a higher rate of mortality was reached in experi- ments carried on during the winter than the summer months. However, excessively high water and frequent flooding of the REFORMATION OF THE CRYSTALLINE STYLE 265 clam beds with fresh water from an adjacent tributary of the Siuslaw River for periods of several weeks had, I believe, its effect upon the death rate at this time. Moreover, during the fall and winter months Mya arenaria is not in the best physical condi- Fig. 1 Semidiagrammatic sketch of Mya arenaria showing the relation of the crystalline style to the digestive tract. St. stomach; Ss, style sac; Int, intestine. Det ae Fig. 2 Semidiagrammatic sketch showing the position of the crystalline style in the visceral mass. Ss, stylesac. X 1}. Fig. 3 Ventral view of the visceral mass with the exposed portion of the style sac. The usual point of severance of the style sac is indicated by the transverse constriction. Ss, style sac. xX 1}. 266 CHARLES HOWARD EDMONDSON tion. The spawning season of the species in this locality is in: the late summer, extending into September. The energy of the animal is at a low ebb after spawning, the following months being a period of physical restoration and of increasing resistant power. A relatively small number of animals survived when environ- mental modifications were brought about by restoring the mol- luses, after the operation, to the sandy soil near the river shore instead of returning them to the soft, black mud of the bed from which they were taken. Although these latter experiments were conducted during the winter months, I believe the change of soil, the longer time out of water between tides, and other accompa- nying conditions may have contributed to the high death rate in the tests carried out along the river bank. The experiments were commenced on February 23, 1918, and conducted continuously for twelve months, one series following another, each series consisting of from one to three dozen clams with crystalline styles removed, each group being reserved for examination at a definite interval after the operation, the inter- vals depending somewhat upon the condition of the tides, but ranging from six to seventy-four days. It was found that this latter time was approximately the period required for the com- plete reformation of the crystalline style in this species. 3. THE MATURE CRYSTALLINE STYLE IN MYA ARENARIA Among the common marine lamellibranchs of the northwest coast great variation in the resistant character of the crystalline style may be observed. In some forms a complete dissolution of the style readily occurs if the animal is removed from its. natural environment or subjected to an extended period of star- vation. Among the species which may be mentioned in this connection are Cardium corbis, Saxidomus giganteus, Saxido- mus nuttallii, Paphya staminea, ete. Owing to the shape of the crystalline styles of the species just. mentioned, it will be found when the animals are taken out of the water and subjected to unusual conditions that the styles soon creep into the stomachs and coil themselves up there, resulting in the style sacs being empty long before the crystalline styles are actually dissolved. REFORMATION OF THE CRYSTALLINE STYLE 267 The complete dissolution of the styles of these species occurs, however, within a few days at most. I have found the style of Paphya staminea to have entirely disappeared after a starvation period of forty-six hours, and the complete dissolution of the crystalline style of Saxidomus giganteus usually takes place within two or three days. Nelson (18) has shown that in the eastern oyster, Ostrea virginica, the crystalline style disappears within an hour after the animal is uncovered by a recession of the tide and may be reformed in fifteen minutes after active feeding has commenced again. On the other hand, crystalline styles of certain species are found to be very resistant, not readily dissolved, and persisting throughout the life of the animals even when the latter are re- moved from their normal surroundings or when death by starva- tion occurs. This latter group of bivalves includes Siliqua patula, Schizothaerus nuttallii, Macoma nasuta, Mya arenaria, and others. Among these Mya arenaria possesses a crystalline style of a very high degree of resistance. It is certainly never dissolved dur- ing the intervals of the recession of the daily tides and presents a remarkable degree of persistence when the clam is placed under unnatural conditions. After eight days of starvation the crys- talline style of Mya is still very firm except at the extremities where it has begun to soften. The end projecting into the stom- ach has been broken down into a jelly-like mass and the oppo- site extremity shows slight indications of dissolution. Mya may be kept alive, out of water and undergoing starvation, for a duration of fourteen days. At the end of this period the crystal- line style shows but a slightly imcreased degree of dissolution over that presented after eight days of starvation. In Siliqua patula and Schizothaerus nuttallii the crystalline style is found to persist throughout the life of the individual, while subjected to starvation, without much noticeable alteration in its sub- stance. The tenacity of life of these species, however, is not so great as that of Mya arenaria. In the forms just mentioned, showing crystalline styles with great power of resistance, the style sacs are completely or, as in case 268 CHARLES HOWARD EDMONDSON of Mya arenaria, almost completely separated from the intestinal tract. I have not observed spirochaetes in the styles of any forms in which the style sacs are separated from the intestinal tracts, al- though the parasites are at times abundant in the stomachs and intestines of some of these clams, especially Schizothaerus nut- tallii and Mya arenaria. I have commonly found spirochaetes in the crystalline styles of Saxidomus giganteus and Paphya staminea, in both of which species the style sac is connected with the intestinal tract throughout the length of the former. The presence or absence of these free-moving parasites in the crys- talline style also assists, I believe, in establishing the relative consistency of this organ in different species of lamellibranchs. The mature crystalline style of Mya arenaria (fig. 4) is eylindri- cal, becoming slightly thicker toward the extremity which pro- jects into the stomach, which I shall designate as the proximal end, tapering gradually toward the rounded distal extremity which rests against the base of the foot of the clam. In its course from the base of the foot of the clam to the stomach the style approximates, in an imperfect manner, the arc of a circle (figs. 1 and 2). On the ventral surface of the visceral mass, for a dis- tance of from 20 to 30 mm. in a medium-sized individual, the dis- tal portion of the style sac is exposed to view in the midline. From this exposed region, as the crystalline style disappears into the visceral mass, it inclines to the left of the midline in its course toward the stomach entering that organ through its posterior ventral wall. The size of the crystalline style of Mya is dependent, in gen- eral, upon the size of the animal possessing it. During digestion the proximal or stomach end of the style is worn away more or Fig. 4 Mature crystalline style of Mya arenaria. X 2%. Fig. 5 Transverse section of the crystalline style showing the concentric lay- ers and spindle-shaped masses of mucus. Mc, mucus masses. X 22. Fig. 6 Longitudinal section of a portion of the crystalline style, greatly en- larged. Mc, spindle-shaped masses of mucus. Fig. 7 Horizontal section through the floor of the stomach showing the begin- ning of the style sac and the intestine. St, stomach cavity; Ss, style sac; Gs, gastric shield; Int, intestine. X 25. 269 REFORMATION OF THE CRYSTALLINE STYLE Fes es ATED OUP ES Yi®. Ce 5A ZN Be SV uit, == A ZS Wee AN ESS WWE ZA = ISS ZA AN re, ANSF See eS CECI ees - Se J Ee emit AW Us A = OMA LA | OA Spee XN NAAN 4053 ea =) FAY SNS) Zea) Hs I NYY NY ENR ES iN) ES A = eA BN uN ZEN ai a es AN ZA ‘ a Ai = ZZ SS: ZH Zz ‘ EG Zi to ety RS Gi, Ve: ZN i 'SSS 2 SS Sera! sf Sex == ae == yA SK —— A. —= as (= “2 = = | a= a ‘a ed \ n Hh t ath = Co : mammal Za Ha e SILL 270 CHARLES HOWARD EDMONDSON less rapidly. Nelson (18) found that the rotary movement of the crystalline style in Modiolus was not continuous during diges- tion, but was inhibited at intervals, periods of rest alternating with periods of activity. Repeated attempts to verify the rotary movements of the crystalline style in Mya have so far been without success on my part, but from observations on the de- velopment of the organ, as presented later in this paper, one may safely conclude that rotation occurs. By comparing the lengths of the crystalline styles in a large number of individuals with the shells, one finds that the style of Mya arenaria averages approximately 72 per cent of the length . of the shell. In a specimen with a shell length of 120 mm. the crystalline style usually measures from 85 to 90 mm. In diame- ter the style of a medium-sized specimen of Mya is approximately 3 mm. at the larger extremity. The style is firm, quite solid, and translucent in a fresh specimen, but lacks the luster so characteristic of the same organ in Schizothaerus nuttallii and Macoma nasuta. Cross-sections of the crystalline style of Mya arenaria reveal the concentric layers in the substance of the organ (fig. 5). I have been able to count from eighty to one hundred of such layers, which vary considerably in thickness. At irregular intervals between the concentric layers of the erys- talline style of Mya arenaria I have been able to make out minute, spindle-shaped areas which have the morphological ap- pearance and the staining reaction of mucus. These are repre- sented in a portion of a longitudinal section under high magnifica- tion (fig. 6). I have at no time observed a central core of food matter in a mature crystalline style of Mya arenaria. 4, THE RELATION OF THE CRYSTALLINE STYLE TO THE INTESTINAL TRACT IN MYA ARENARIA, AND FEATURES OF THE STYLE SAC In Mya arenaria, for a distance of from 10 to 15 mm. from the floor of the stomach, the style sac and the proximal extremity of the intestine are incompletely separated (figs. 1 and 17). Cross- | sections through the style sac and intestine just below the floor of the stomach show the union between the two tubes, in all es- REFORMATION OF THE CRYSTALLINE STYLE 200 sential features, to be similar to the union of the intestine and the style sac existing throughout the entire length of the latter in Cardium corbis, Saxidomus giganteus, Paphya staminea, and others. In these cross-sections prominent typhlosoles are ob- served to be developed, one on either side, separating the intesti- nal tube from the style sac (figs. 7 and 8). Here the intestine is on the anterior surface of the style sac, and I have designated one typhlosole as the left and the opposite one the right. After being in union for a short distance below the floor of stomach, the intestine breaks away from the anterior border of the style sac, making a sharp turn toward the stomach wall, bending ventrally again and beginning the characteristic series of loops in the visceral mass (figs. 1 and 17). The intestine crosses to the right of the crystalline style as it ascends toward the dorsal border to traverse the pericardial cavity. At the point of separation of the intestine from the style sac, and as a result of this separation, an evagination of the style sac occurs (fig. 9). The groove thus formed continues on the antero- lateral. border of the tube, becoming less prominent toward the distal end, and fades away as a distinct groove among the numerous folds of the inner wall at this extremity. Transverse sections at the point of separation of the two tubes clearly indicate that the groove is a remnant of the intestinal tract formed by the latter as it is drawn away from the style sac. In the formation of this groove in Mya arenaria the evidence points toward this species being a transitional form between lamellibranchs in which the style sac is united throughout its length with the intestinal tract, and those species in which the intestine and the style sac are completely separated. ‘The cells lining the groove have the same general characteristics as the cells of the epithelium of the intestinal tube. In the bottom of the groove thus formed the cells are short, becoming longer on the sides where they merge into the epithelium of the style sac proper. The position and arrangement of the nuclei of cells of the groove vary somewhat, due to modification of the cells in the different regions of it, but the nuclear arrangement char- acteristic of the epithelium of the digestive tube is generally 262 CHARLES HOWARD EDMONDSON maintained. Cilia similar to those of the intestine cover the free ends of the cells of the groove. The typhlosoles separating the style sac from the intestinal tract near the stomach wall now become the typhlosoles separat- ing the groove from the style sac proper and maintain their relative position and importance, the more prominent one being on the right side. Histological preparations of the style sac, by cross-sections through the region of its union with the intestine, give a view of the cells covering the typhlosoles (fig. 8). On the right typhlo- sole, extending nearly one-third the circumference of the tube, the cells are very long, narrow, so closely crowded together that their nuclei have become compressed and are situated at differ- ent levels. The free ends of the cells are provided with a dense layer of cilia somewhat shorter than those covering the cells of the general surface of the style sac. The opposite typhlosole may lack entirely or carry but a small group of these long, narrow cells. The epithelium of the general lining of the style sac con- sists of cells characterized by their uniformity of size, being ap- proximately 0.13.mm. in length and 0.016 mm. in diameter in a medium-sized individual (fig. 14). In these cells the nuclei are uniformly oval, each with a distinct nucleolus, and are located near the middle region of the cells all approximately at the same level. A basement membrane resting upon a thin layer of loose connective tissue supports the cells. Attached to the free ends of these cells is a very dense layer of strong cilia slightly longer than those carried by the cells of the typhlosoles. In a medium- sized individual the cilia of this dense layer measure about 0.04 mm. in length or one-third the length of the cells which bear them. The cytoplasm of the cells just described is granular and often their free extremities are crowded with minute granules of brown pigment. Sabatier (77) concludes that this brown, granular pigment, characteristic of the cells of the general surface of the style sac, is derived from the substance of diatoms, protozoa, etc., collected between the crystalline style and the epithelium, of the lining of the tube and squeezed out by the action of the strong cilia. This involves the function of the crystalline style as pro- REFORMATION OF THE CRYSTALLINE STYLE DHS, Me, Zz aN Ps Sst: = LQSS SS SARTO NS ra >= as ESAS = SSS ZS BS SSS: fe WAS 5 SS ULM TTT Coo aut an > Ci 2 on os SPIES. BSS Ws ? Se GAS 7 r toa sd ig Rp aoe Zs WN -" SATIN HAIN ZX op L ER A BS 1 Ue ZX. Tp aoa! =e ‘3 a b hi Ups. 5 24) Uf Ga I) a My cS ey AY iN at \ ae: 2; a \\ L) MS H Te \ Nt Ne? aN ia . Ss ——— ah ee rah a) {i Wis Gre \ tw pore. u GS ins Nail, “a, i i 14. 276 CHARLES HOWARD EDMONDSON The circular folds of the style sac do not continue to the mar- gin of the groove, but abruptly terminate at the lateral border of the long, narrow cells on either side of the groove. Therefore the typhlosoles are not marked by undulating epithelium as is the rest of the tube, but are smooth on their inner surfaces. This latter feature is made plain by longitudinal sections, close to the margin of the groove, in which the epithelium is seen to rest upon a basement membrane disposed in an even line without folds or undulations. 5. THE GASTRIC SHIELD IN MYA ARENARIA Spreading over the left wall, the roof, and the floor of the pos- terior half of the stomach of Mya arenaria is found the structure which Nelson (’18) called the gastric shield. In this species the shield does not cover the right wall of the stomach (fig. 18). The gastric shield consists of a cartilage-like layer applied to the epithelium of the stomach wall growing thicker from the posterior region forward where it develops into an irregular struc- ture with three curved processes which clasp prominent folds of the dorsolateral wall of the stomach (figs. 15, 16, and 19). Poli (1791) first described this organ under the name of ‘fleche tricuspide,’ and suggested that it might control the flow of bile into the stomach by the extension of its processes into the bile crypts. Histological sections through the stomach wall and gastric shield of Mya arenaria show the latter to be closely applied to the columnar epithelium. Beneath the shield the epithelial cells are not provided with cilia. It will also be observed that the gastric shield consists of numerous strata comparable to the con- centric layers of the crystalline style in cross-section. Fig. 17 Semidiagrammatic section showing the relation of the crystalline style to the gastric shield in Mya arenaria and also tothe proximal extremity of the intestine. St, stomach; Gs, gastric shield; Cs, crystalline style; nt, intestine. x 2. Fig. 18 A transverse section of the stomach of Mya arenaria showing the position of the gastric shield on the roof and the left wall of the stomach. Sf, stomach; Gs, gastric shield; Ss, style sac. X 3}. 207 STYLE REFORMATION OF THE CRYSTALLINE a c) 3 * e i] @, n , ooo 18 278 CHARLES HOWARD EDMONDSON According to Gutheil (11), the gastric shield is built up by a hardening of the material secreted by the epithelium on which it rests. Nelson (18) states that it is probably in the nature of chondrin. In Mya arenaria the structure is very resistant, slight dissolution of its thickest region occurring during a period of about thirty days after the removal of the crystalline style. This peculiar structure, without doubt, serves as a protection to the epithelium of the wall of the stomach which it covers, and at the same time it assists in the wearing away of the end of the crystalline style which, in its rotation, presses against the concave surface of the thickened portion of the shield (fig. 17). 6. DETAILS OF THE EXPERIMENTS In presenting the details of the various experiments carried on for the determination of the reformation of the crystalline style, following its extraction, in Mya arenaria, it is my purpose to con- sider the series of experiments in the order of the progressive de- velopment of the crystalline style rather than the chronological order in which the experiments were conducted. Experiment 1. September 19, 1918 Six medium-sized clams, with crystalline styles removed by trans- verse sectioning of the style sacs about 10 mm. from the distal extremi- ties, were replanted under normal conditions. The experiment was one of short duration, continuing for six days, with the object in view of determining the exact character and position of the crystalline style in its initial stage. Result, September 25, 1918. After six days all of the clams operated upon were found to be alive and in good condition. The severed ends of the style sac were healed in each individual, and in one or two speci- Fig. 19 A section of the left wall of the stomach of Mya arenaria with the gastric shield closely applied to the inner surface of the epithelium. Bm, base- ment membrane; Gs, gastric shield. X 84. Fig. 20 The regenerating crystalline style four days after extraction, the clams being out of water and without food but subjected to low temperature. X 23. Fig. 21 The crystalline style after a growth of six days under normal condi- tions. xX 2%. Fig. 22 The same as the preceding figure. X 2. . Fig. 23 A regenerating crystalline style thirteen days after extraction. X 2. REFORMATION OF THE CRYSTALLINE STYLE peters SEs & 2! 280 CHARLES HOWARD EDMONDSON mens were closed. In each of the animals the short, distal portion of the style sac, separated from the remainder by the operation, was filled with mud and other foreign material. In each of five of the specimens, on opening the long, proximal section of the style sac by a dorsal, longitudinal incision, the beginning of a crystalline style was evident. At this early stage the crystalline style consisted of a delicate mucilaginous sheath enclosing an axillary core of food and foreign material which included diatoms, sand _ particles, spicules, etc. The delicate style at this period occupied the entire length of the sectioned style sac and was greatly convoluted and begin- ning to coil (figs. 21 and 22). In three specimens the style was closely applied to the right typhlosole;in one animal it occupied the left typhlo- sole, and in one it alternately lay on the right typhlosole and in the groove throughout its length. The diameter of the crystalline style at this stage of its development -averaged about 0.3 mm., but varied at different levels due to the food and foreign material being massed in the axis in greater amounts at some points than at others. Apparently the digestive processes were inhibited in the clam during the interval of six days, the stomach being empty and the intestinal tract entirely free of food material. No change was observed in the gastric shield of the wall of the stomach at the close of this period. Experiment 2. November 16, 1918 Eight clams, each with the crystalline style removed by a transverse severance of the style sac near the distal extremity, were replanted under normal conditions. The experiment extended over a period of thirteen days. Result, November 29, 1918. Five of the eight clams were found to ‘be alive and in good condition. The severed ends of the style sac were healed in each surviving specimen, and in one or two animals were ‘entirely closed. In each individual the short, blind end of the style sac was filled with mud, while the proximal division was occupied by a crystalline style in the course of development. At this period the crys- talline style presented a considerable degree of advancement over one of six days’ growth. It was tightly coiled throughout its length, being twisted to the right when viewed from the proximal end. A food core occupied the central axis of the style (fig. 23). Fig. 24 A sketch of the style sac opened along the posterior border showing the crystalline style of thirteen days growth lying on the right typhlosole. Sz, -stomach; Ss, style sac; Cs, crystalline style. Natural size. Fig. 25 The crystalline style thirty-four days after extraction. X 2. Fig. 26 A reformed crystalline style fifty-two days after extraction. X 2. Fig. 27. A crystalline style sixty-nine days after extraction. X 13. Figs. 28-30 Crystalline styles approaching maturity, seventy-four days after extraction. > 134. Some crystalline styles were completely reformed after this period of regeneration. = | LTRS Sea ea aaah tes ah 281 282 CHARLES HOWARD EDMONDSON In each individual the organ lay on the right typhlosole, occupying nearly the entire length of the tube, but did not protrude into the stomach (fig. 24). In diameter the style measured 0.5 mm. at the dis- tal extremity, which at this time was somewhat thicker than the proxi- mal end. That no food had been taken by the animal during the in- terval of thirteen days seemed apparent by the empty and blanched condition of the intestines. The stomach was without food in each surviving individual. ‘ The material forming the food core of the crystalline style at this early period of its development is probably furnished from that which was inthe stomach at the timeof the extraction of the style. Although the style sac is an open tube leading from the stomach after the re- moval of the crystalline style, in no specimen examined have I been able to detect a particle of food or foreign material being carried down this tube from the stomach except that enclosed by the style itself. Experiment 8. February 23, 1918 Twelve clams, ranging in size from 40 to 125 mm., with styles re- moved, as in the preceding experiment, were planted under normal conditions. This test extended for a period of sixteen days. Result, March 9, 1918. On this date nine of the twelve clams were surviving and apparently in good condition. The cut surfaces were well healed, and in a number of specimens the style sacs were also closed at the severed ends. On opening the proximal portion of the — style sac of each of these clams no appearance of a regenerating crystal- line style was evident. The tube was free of foreign material, but ap- parently the reformation of the crystalline style had not yet made its beginning. An examination of the stomach revealed the absence of food, and the intestinal tract was also empty. It was quite evident that the clams had not been feeding since the removal of the crystalline styles. This experiment resulted negatively in so far as the reforma- tion of the crystalline style is concerned, but, I believe, is evidence of the slow development of this organ during the winter months. Experiment 4. November 30, 1918 Twelve clams with crystalline styles removed were taken from their native bed and planted in a sandy locality near the bank of the river where they would be exposed for a much longer period than usual by on recession of the tides. The experiment continued for twenty-six ays. Result, December 26, 1918. Of the twelve clams planted in the sandy soil near the shore line four were alive after a period of twenty-six days. Of these survivors two specimens showed no evidences of the: reformation of crystalline styles. The other two possessed styles very rudimentary in development, correspondng closely to those of six days’ growth as recorded in experiment 1. REFORMATION OF THE CRYSTALLINE STYLE 283 The stomachs and intestines of these animals were empty, the latter exhibiting a blanched appearance indicative of the lack of activity for a considerable length of time. It was evident that the ingestion of food had been completely inhibited during the entire twenty-six days. All surviving animals were in a very weakened condition. Experiment 5. December 27, 1918 Thirty-seven clams with styles removed as in previous experiments were returned to the natural bed. This test extended over a period of thirty-four days. Result, January 80, 1919. Five clams survived the operation for thirty-four days. On examining these animals it wasfound that each had reformed a very rudimentary crystalline style. In each case the organ rested on the right typhlosole, occupying nearly the entire length of the proximal division of the style sac, but not projecting into the stomach. It was convoluted and coiled throughout its course (fig. 25). frequencies —> rf actorial class units Fig. 2 The class distributions according to factorial units affecting facet numbers in populations of ultra-bar females (U), bar females of the second low selected white-eyed generation from which ultra-bar was derived (B), and full- eye (Ff). The class frequencies are in per cents of the whole population and are therefore directly comparable. The zero of the factorial scale is the mean value of the unselected white bar. The same material is represented in tables 1 and 3. U —— > frequencies el Beth Pat WODAMKAPANUVPOOPNUARUARIDO ° s eeee#eeee e O} 105 10.0 O01: 6) (6. @ DOCOCODODIDD ODDO SCO WwWNNHNWNWWO AAAAATAAMAAiIAnInngnaanig»ngnan ry p ° ct ° Le ial class units Fig. 3 The class distributions for males. The same material is represented in tables 2 and 4. 300 CHARLES ZELENY Tice further located the gene for bar eye in the sex chromosome, and on the basis of percentage of crossing over with other sex- linked factors it was put at 57 cross-over units from yellow. She considered bar as a dominant to full eye, and it is so described in various papers by Morgan and others. Strictly speaking, the heterozygous condition is intermediate between bar and full. According to the method used here, the heterozygotes between low bar (F 24) and full are +23.89 factorial units from bar and only —7.23 such units from full (table 7). If 100 per cent repre- sents a condition of complete dominance and 0 per cent of com- plete recessiveness, bar has a dominance coefficient of 23.2 per cent and full of 76.8 per cent. If either is to be considered as a dominant, the term should be applied to full eye rather than to bar eye. Bar eye has a high degree of stability, though there is an occa- sional mutation such as the reverse mutation to full. Such re- verse mutations are apparently actual returns of the gene to the original condition. The writer has noticed several of them in his stocks and several other germinal changes, among which is the one described in the present paper. THE ORIGIN OF ULTRA-BAR Ultra-bar also appeared in a single male. This male with but 19 facets appeared in the second low selected generation of the white-bar line on October 20, 1917. A cross with a 44-+facet sister gave a pure stock in F;.. The ultra-bar thus established has remained stable in character except for the mutations de- scribed on page 304 ff. The average facet number is 21.96 in the females and 23.04 in the males as opposed to 61.8 and 75.6 in the second low gener- ation of white bar from which it was derived and 810.6 and 849.8 in the full eye. In factorial units the second low generation of the white-bar line is —25.93 and —24.59 units from full. Ultra- bar is —10.53 and —11.52 units from the second low generation of white bar and —36.46 and —36.11 units from full eye. Typ- ical representatives of full eye, bar, and ultra-bar are shown in figure 1. Table 1 gives the values of the means, the ranges, and CHANGE IN THE BAR GENE OF DROSOPHILA 301 the standard deviations. In the case of factorial units the stan- dard deviation may be used directly as the coefficient of varia- tion. The values for full eye are based upon such a small num- ber of cases wholly because of the tedious character of the count. In the bar stocks counts can be made by placing the flies on their sides on a small block of wood with a surface painted black and so inclined as to bring the eye facets in a horizontal plane under the microscope. The rows of facets can then be followed by the observer and counts can be made with a fair degree of accuracy. In practice there is, however, always a small per cent of error which increases with the facet number. In full eye the large number of facets combined with the curvature of the eye makes this procedure impossible. It is necessary to resort to the mount- ing of the eye on a slide under a cover-glass. The non-chitinous parts are removed with caustic potash and the faceted part mounted in Canada balsam and flattened out under a cover- glass. With a high objective, a mechanical stage, and cross hairs in the ocular, it is possible to count the facets with a fair degree of accuracy, but it is not profitable to count a large num- ber of individual eyes. Tables 3 and 4 and figures 2 and 3 give comparisons of the range and variability of the three stocks in graphic form. There can be no question that the stocks are absolutely distinct. Bar and ultra-bar barely overlap in the females and not at all in the males. As shown in the tables, the variability of ultra-bar, here expressed directly by the standard deviation, is significantly less than that of the bar stock from which it was derived, and this difference between bar and ultra-bar remains throughout the selection generations. Figures 2 and 3 bring out the same point. The standard deviation value for full eye of course has no special significance because of the small number of cases. It may per- haps indicate that full eye is less variable than bar. TABLE 3 Variability of full, bar, and ultra-bar. Females BAR SECOND LOW me ractontaL OuASS8 IN HSReN EL ee SELECTED GENERATION LER SE Number Per cent Number Per cent Number Per cent —15.93 11 ne 0.2 —14.93 12 3 0.2 —13.93 13++ a 1.4 —12.93 15++ 493 3.1 —11.93 16-17 155 9.7 —10.93 18-19 317 19.9 — 9.93 20-21 387 24.3 — 8.93 22-23 294 18.4 — 7.938 24-26 282 Lea — 6.93 27-29 57 3.9 — 5.93 30-32 il 0.6 il7/ Iva il — 4.98 33-385 1 0.6 4 0.3 — 3.93 36-39 3 iL, — 2.93 40-43 4 22 — 1.93 44-48 22 188} — 0.93 49-53 23 12.8 + 0.07 54-59 34 19.0 + 1.07 60-65 30 16.8 + 2.07 66-72 26. 5 14.5 + 3.07 73-80 11 Gn + 4.07 81-88 16 9.0 + 5.07 89-97 7 3.9 + 6.07 98-107 + 7.07 108-118 + 8.07 119-131 1 0.6 aioe Od 132-145 +10.07 146-160 Sell Oye 161-177 +12.07 178-196 +13 .07 197-217 +14.07 218-240 +15.07 241-265 +16 .07 266-293 = alles Uh 294-324 +18.07 325-358 +19.07 359-396 +20.07 397-438 +2107 439-484 +22.07 485-535 +23 .07 536-591 +24.07 592-653 1 10 +25 .07 654-722 +26 .07 723-798 3 +27 .07 799-882 4 40 +28 .07 &83-975 2 MPotals i. ect are 10 100 179 100 1590 100 302 TABLE 4 Variability of full, bar, and ultra-bar, Males. BAR SECOND LOW =f ULTRA-BAR CLASSES Aenea ken eh 20a SELECT&D GENERATION IN FACTORIAL FACETS UNITS Number Per cent Number Per cent Number Per cent Oo 12 1 eal —20.05 13+ z 0.2 —19.05 15+ 123 0.8 —18.05 16-17 68 4.3 —17.05 18-19 i 0.6 148 9.3 —16.05 20-21 | | 294 18.4 —15.05 22-23 395 24.8 —14.05 24-26 429 26.9 —13.05 27-29 185 11.6 —12.05 30-32 43 2.7 i 05) 33-35 11 0.7 —10.05 36-39 4 0.3 0 40-43 2 1.3 1 0.1 (SU) 44-48 a 4.4 — 7.05 49-53 it 4.4 — 6.05 54-59" 19 12.0 — 5.05 60-65 21 13°3 — 4.05 66-72 18 11.4 — 3.05 73-80 23 14.6 pe aeUo 81-88 26 16.5 aH) 89-97 15 a5 — 0.05 98-107 8 5.1 + 0.95 108-118 6 3.8 ae IL86) 119-131 3 19 2200 132-145 2 1.3 se) 146-160 + 4.95 161-177 + 5.95 178-196 + 6.95 197-217 see (iS) 218-240 ato /a) 241-265 Siu OD 266-293 +10.95 294-324 +11.95 325-358 +12.95 309-396 +13.95 397-438 +14.95 439-484 +15.95 485-535 +16.95 536-591 +17.95 592-653 +18.95 654-722 1 10 eral OEOe 723-798 1 10 +20.95 799-882 ft 40 +21.95 883-975 OMe 30 +22.95 | 976-1078 1 10 fro Fre FT ne oa 10 100 158 100 1594 100 1 The mutant from which the ultra-bar stock arose. 303 304 CHARLES ZELENY CONSTANCY OF ULTRA-BAR The single mutant male appeared on October 20, 1917, and the new stock was isolated in pure condition on December 27th. Numerous counts at 27° show no essential change in character except for the mutants mentioned in a later paragraph. The mean, range, and standard deviation remain essentially un- changed. The data are given in tables 5 and 6. During the period of establishment of the stock it was subjected to selection, but without any noticeable effect. There is no indication that this selection had any part in the production of the constancy. The selection data are treated in a separate paper. Later counts made during July, 1919, and not given in the tables show that the stock had remained without essential change for twenty months. The differences in the various cultures as shown in tables 5 and 6 are in some cases well within the probable error of random sampling, but in several cases the departures can not be explained in this way. For instance, take such a difference as that between the means of cultures no. 178 and no. 184 which is equal to 1.86 units. Its probable error, E, = + VE; + E; = + 0.17, is only one-eleventh of the difference. Correspond- ingly, the difference in the,males of the two cultures is 1.71 units and the probable error is 0.16. These differences are therefore in all probability not due to random sampling. Their lack of permanence indicates that they are environmental and not germinal. On four occasions individuals have appeared in the ultra-bar stock which differ markedly from the ordinary ones. Three of these may be considered as probably reverse mutations to full and the fourth as a new mutation upward in the direction of bar. Mutant A. On July 9, 1918, there appeared in bottle no. 496.4 of the ultra-bar stock at 27°C. a female with 46 facets in the left eye and 50 in the right eye. The upper limit of ultra- bar stock females is 35 facets (table 5). It appeared that this female might be a heterozygote with one ultra-bar and one full- eye factor as the facet range of 54 such heterozygotes in test no. 307, table 7, is 28 to 49. Accordingly, she was mated with four CHANGE IN THE BAR GENE OF DROSOPHILA 305 | of her brothers and gave. according to expectation both full-eye and ultra-bar sons and ultra-bar and heterozygous daughters. Other crosses followed expectation in a similar manner. That she is not the result of contamination is indicated, first, by the fact that she is white-eyed, and that hypothesis therefore necessitates contamination by a white full-eyed fly. Trap tests showed that there were no such flies free in the laboratory, though other kinds were caught. In the second place, only a single such individual appeared. Thirdly, the fact that she is a heterozygote makes it improbable that she is a stranger in the bottle. Fourthly, the general laboratory procedure is the same as that pursued when the reverse mutations of bar appeared, and the arguments given for those cases by May (1917) hold in this case also. Deficiency tests were made. If a piece of the chromosome carrying the ultra-bar factor has dropped out, either the forked gene which is —0.5 units from ultra-bar or the fused gene which is +2.5 units away or both would in all probability be carried with it. When full males derived from a cross of the mutant female with ultra-bar males are mated with heterozygous fused females, they should give fused sons if there is deficiency on the plus side of ultra-bar. Such crosses gave no fused or forked sons and deficiency cannot be considered as the explanation of the mutant. Mutant B. In a cross between wild red-eyed females and a 22-facet white male from bottle no. 150 sp., 59 of the females were heterozygotes according to expectation and there was one full-eyed female. This exceptional female was crossed with wild red males and gave 23 full-eyed females and 25 full-eyed males. Both chromosomes therefore have the full-eyed factor, and the case is not due to a failure of the ultra-bar factor to dominate. Dr. A. H. Sturtevant has suggested that this may be a case of non-disjunction. Unfortunately the eye color was not recorded at the crucial point. Mutant C. A single full-eyed white male appeared in the white ultra-bar stock no. 158.1 on December 26, 1917. It was shown by test no. 291.1 to be like ordinary full-eye, but no further tests were made. THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 30, NO. 3 CHARLES ZELENY 306 FOG ‘SOU Bopeywo) [NJ YIM IVq-v1Z[N SNOSAZOLOJoyY 9G 0} So} AQ UMOYS SBM cs II |90°0*96'0Z|/09 F— \€6'°SI—-| ZOO CT €0'0F 61°6— *(GOg pus YOIyM ‘fa[BULoy YOORJ-9p OUO Burpnypoxny y OOT &@ ILA 8161-41 ILX LI6I |P—® 008 6¢ ILA-@ ITA 8161 T ITA-8T IA 8I6T ¥V IA-IG A 8161 6¢ A-IT A 8I6I VI A-92 AI 8I6T 92 AI-92 III 8I6T G@ III-22 II 8161 IT III-91 IL st6r LIT Oates tor FL II-91 I 8I6r é II-ST I 8161 &@ I-¥ I 8161 G I-71 IX LI6IL 66F-96P TLé 8&6 GEG 966 916 961 061 v8I IE il 16°06 GG II FS 0G 96 oT 0& 0G SG él co 61 62 ST 6916 0€ ial GG IG 8% &1 86°06 62 real OL 0G && II LL 61 1g roe 89° 6 6 v1 L806 o& at FO 16 Cee aie ch 98°16 4soysry | ySoMo'T uBoyy SLOOVaA ysoyspyT | JSaMOT | UOTeIAep prepuryg uvoyl SLINO TVIHOLOV4 GOFO' 46 “Saqypuaf wog-pajyjn fo hounjsuog $ HTAVL Salva SugdaWon DOTVLVO 307 CHANGE IN THE BAR GENE OF DROSOPHILA ‘a9A9 [[NJ ouo Surpnjpoxg ; *qUBYNUL [VUISIIO OY], 1 2OP GL 90: 0+F70 S2icr 6— |SO'1IZ—| 20 0F I¢°T £0'OF 96° FI— | FEST | &2 IIA 8IGI-FI ILX LIGT |P—8 008 GE at G6 Eo =|GL OL—|S0 IZ-| 80° OF OST FO OF O06 FI— | 009 &@ IIA-Z IIA S8I6T |66%92F 62 9T LE&S |6L ¢1—|0€ 8I— SOO St T GliOte 922i Bieor, T ITA-8T IA 8I6T TZé 8G OT cf ce §=6S0 €I— 08 SI—-| CI OF 6E'T AN Ogee an Si ae FI TA-Té A 8I6T 86 Ig 91 COMGe 0 Gl 08 Sl" —2010==" Goo. OM0t Lr Sl 1592 66 A-II A 8I6T G&G O& ill TiVGe se Gh ise cl) “S0s04= Ze 1 GLOF 69 7I— | 19 VI A-96 AI 8I6T 966 && ST COrcGw 185 LlS5.8l =} eeZ0004= 2p ar OleOAS SOS S6 96 AI—-9¢ ITI SI6T 916 96 9T L3.6G™= Gy Ol—|0E Sl! SOK07E SET 60°0F 6° FI— | SOL SG III-2z II 8I6r 96T 66 v1 [STG |c2 GI |SS 61>) SOOO 08 1 60°0F 9F ST— | 8&6 IL TIE-OF If St6r O6T 6& Lal IGG |8E TL 184 Gi >| 600-5 291 Sl Ose Gale amc T III-9 II 816. PST OF 61 cvSG |cb6— |08°9I—| 20°0F 8F'T OL 05 SO nh eaieco VI II-9I I SI6r SLT OS SI O8'GZ = 8E°SI—|88°8I—| 90° 0F 02'T 60'0F 66°FI— | 8 G II-ST I 816. OLT 6€ CT DeeGG 2 |LONG— see 8h oa) eZ 0"04tes 162) Tl 60: 0F 60°ST— | 89I &c-F I S161 POL 1g ZT ZCG \80 cl—08 21=| 8600+ 29° Gln ele Gh— ee? & I-VI ITX ZLI61 8ST 00°6T (Oe |) base ar 0G X LIGT | ISP JOYS] | JSomorT uve, JSeqsty | JoMorT UOlVBIAVP plepuRyg uvoyy Bh -IGNI sSaLva Ho tinars JO uaa SLaOVa SLINA TVINOLVA -WON GOFO £6 “SaypU svg-nuzyjn fo hounjsuog 9 HIAVL 308 CHARLES ZELENY Mutant D. In bottle no. 196.2 of the ultra-bar stock there appeared on March 20, 1918, both males and females with a facet number coming within the range of bar of the low selection line. When this particular mass culture was started on February 2, 1918, all the individuals were ultra-bar. The number of indi- viduals with the new character makes it probable that the muta- tion occurred at least two generations before the first observa- tion was made. A 62-facet and two 99-facet males were mated separately with full-eyed wild females and gave heterozygotes which differed from those of ultra-bar x full and also of bar x full. The mean value in about five hundred flies as determined by an estimate without counts is very distinctly between the other two types of heterozygotes, and the range is extended so as to overlap their ranges. This extension of the range indicates the probable presence of accessory factors, but its limits make it seem that the new form is a mutation of ultra-bar in the di- rection of bar which is not exactly like bar. Other cases that may be considered as mutations are given in the section on the determination of the locus of ultra-bar (p. Silay) The frequency of occurrence of mutations in ultra-bar is about | the same as that in bar, and the stability of the two stocks may be considered to be of about the same order. THE CHANGE IN DOMINANCE Even more striking than the difference in facet number be- tween bar and ultra-bar is the difference in dominance. A _ hetero- zygote between ultra-bar and full comes very close to ultra-bar, while the heterozygote between bar and full is closer to full than to bar (fig. 1 and table 7). For purposes of comparison, the fol- lowing method of determining the coefficient of dominance has been devised. Let A and A’ be the two members of an allelo- morphic pair of factors, then the coefficient of dominance / co. = eau < 100 in which AA and AVA’ are the mean values, respectively, of the homozygous stocks in factorial units and AA’ is the mean value of the heterozygous individuals. CHANGE IN THE BAR GENE OF DROSOPHILA 309 As discussed in a separate paper, the value of a factor or com- bination of factors affecting facet number is put on the basis of 10 per cent units. An arbitrary point, the mean facet value in the unselected stock, is taken as the point of reference, and dis- tribution classes are so arranged that the facet range of each class is 10 per cent of the mean facet value of that class. In this scheme any facet value may be represented as being a de- parture plus or minus a certain number of 10 per cent units from the point of reference. The method is based on the view that TABLE 7 Dominance values in ultra-bar, bar, and full eye FACTORIAL NUMBERS MEANS ; DOMI- FEMALES CN ag ero ola ee | UALS UNITS ZYGOTES CENTS Ultra-bar stock...........| 158-499) 1590 21.96 | —9.79 | —5.53 | 84.8 MOMS OCK:. fetches .c ce tess) OAD 10 | 810.6 |+26.67 |+380.93 | 15.2 Heterozygotes............| 357 54 36.54 | —4.26 Low selected bar (Fas). .... 391.2) 129 35.1 | —4.45 |+23.89 | 23.2 BNUSCOCK a ius. dso odes 2, s(t ny O20 10 | 810.6 |+26.67 | —7.23 | 76.8 Heterozygotes. oo... 55---) 109 19 | 399.9 |+19.44 Ultra-bar stock...........]| 158-499) 1590 21.96 | —9.79 | —1.84]| 82.6 Low selected bar (F2)..... 144-145) 179 61.8 | +0.74 | +8.69 | 17.4 Heterozygotes.....<......| 742 87 200 \ =F oOo change in facet number is not a matter of accretion, but that the whole facet-producing mass is involved. According to this view, the factorial value of the difference between a 50-facet and a 55-facet individual is the same as that of the difference between a 500- and a 550-facet individual. In other words, an environ- mental or germinal change in what would otherwise have been a 500-facet stock and which produces instead 550 facets would, if acting upon a 50-facet stock, produce 55 facets and not 100 facets. The ordinary tabulations of variation in which classes have equal character values over the whole range of variation do not give equal factorial values. 310 CHARLES ZELENY Using factorial values and reducing facet numbers to departures plus or minus from the mean value of the unselected white-bar stock, the following percentage coefficients of dominance were determined: a. Low selected bar (F.4) over full eye _ full — heterozygote 1.26.67 1924 full — low selected bar (Fx) 26.67 — (—4.45) = 23.2 per cent. b. Full eye over low selected bar (F.:) __ low selected bar (F.s) — heterozygote < 100 ~ low selected bar — full —4.45 — 19.44 NES ee eee 100 = 76.8 if SAE ae Be tien full — heterozygote . Wltra; b ver full eecet WE wiltare set ees Tid EG year ban Tesi S850 4-26.67 — (—4.26) eee COI SEE t. +26.67 — (—9.79) poeta dias d. Full eye over ultra-bar = ultra-bar — heterozygote ultra-bar — full —9.79 — (—4.26) ie 9.79 — 96.67 factorial class units Fig.9 Ultra-bar and bar males from heterozygous females by ultra-bar males. ‘The same material is given in table 9. are put down as high heterozygotes, and without further evi- dence might have been supposed to have a bar instead of an ultra-bar factor. TABLE 10 F2 from ultra-bar X full and full X ultra-bar TOO HIGH LOW HIGH ULTRA-BAR | ULTRA-BAR+ FOR HETEROZY- HETEROZY- ULTRA-BAR GOTES GOTES IMB GSks es, cucccscntae cen coe 1235 3 3 320 CHARLES ZELENY Among the F, males of both the reciprocal crosses there are 1235 males within the range of ultra-bar, three of which come just above this range having respectively 42, 42, and 44 facets, and three which are considerably higher having 51, 60, and 88 facets. Of the first three, one of the 42-facet males was tested by crossing to full and shown by its dominance to be an ultra- bar. It is probable from this and other evidence that this slight extension of the range of ultra-bar is due-to accessory factors introduced by the full stock. The other three males are of a different character. The 60-facet male died without a test. The 51- and 88-facet males when mated to full gave heterozygotes which were higher than those of ultra-bar x full, but on the other hand lower than those of bar x full. These heterozygotes agree in character with those of the mutant in the pure ultra- bar stock which was described above on p. 308. The high ex- ceptional males cannot therefore be considered as bar males produced by the separation of the U factor from B’ by crossing over. Among the F, females of the cross between full females and ultra-bar males there are 553 with the characteristics of typical low heterozygotes of the F; generation and four with higher facet numbers. One of the latter has 154 facets and is considerably below the range of heterozygotes between full and bar as given in table 9. The other three females come within this range. The one which was tested by crossing to full, however, gave only ultra- bar and full males instead of bar and full as its facet number seemed to indicate that it should give. It is therefore an instance of failure of the ultra-bar factor to dominate full in the ordinary manner. As in the case of the males, crossing over cannot be given as an explanation of these exceptional females, though it is unfortunate that some of them died before a test could be made. However, even if all the exceptional untested males and females were classed as bars produced by crossing over of an assumed accessory ultra-bar factor, the locus of this factor would have to be placed very close to bar. Absolute proof is impossible because it can always be held that the new gene is so close to the old one that crossing over cannot certainly be expected in any CHANGE IN THE BAR GENE OF DROSOPHILA ik practicable number of individuals. Besides there may be a ‘no- crossover’ accessory factor. The reasonable conclusion to be drawn from the evidence is that ultra-bar has the same locus as bar or a locus so close to it that the two are a single unit in all demonstrable cases. It follows that ultra-bar has been produced by a change in the bar gene, and not by a change at some other locus. CONCLUSIONS EKye-facet number in Drosophila furnishes an excellent material for the quantitative study of both germinal and environmental factors. The demonstration of the striking and regular temper- ature effect makes possible the recognition and analysis of ger- minal factors. As outlined in preliminary reports (see bibliog- raphy) selection for facet number in bar-eye has shown that the effect of selection here is the result, first, of a sorting out of ger- minal diversities in the original stock and, second, of the isolation of new diversities as they arise. Besides occasional reverse mu- tations to full, the original diversities as well as the new changes have been due to accessory factors. The mutation described in the present paper, however, resembles the reverse mutations in that it involves change in the bar gene. The effect here, how- ever, is essentially an intensification of that produced by bar. There is further decrease in facet number and a great increase in dominance. It is interesting to note also that the change occurred in a line in which low selection was being carried on. In the light of all the evidence concerning the direction of mutations, no special significance can be attached to this fact at present. For instance, reverse mutations to full occur in the low selection lines of bar and in ultra-bar as well as in the high selection lines. While, therefore, a connection between the direction of selection and the direction of mutations in general seems improbable, the matter is of sufficient importance to warrant careful record of all instances. THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 30, No. 3 a22 CHARLES ZELENY The constancy of the new gene seems to be of about the same order as that of bar. Reverse mutations to full occur in both. In ultra-bar there is also a reverse mutation in the direction of bar, but apparently not to the same point because, while the mutant comes within the facet range of bar, it does not have the same dominance over full. | The locus of the new gene is the same as that of bar or is so close to it that our methods of analysis are not capable of sepa- rating the two. The crossing-over test is the only one that can be applied, and it has the obvious difficulties due to the improba- bility of any crossing over between two genes which are very close together, and it is also subject to the criticism that the crossing over may itself be inhibited by an accessory factor. However, it is our only test and its findings must be taken at their face value. In the particular test applied, the presence or absence of crossing over in F, of crosses between ultra-bar and full, a few individuals appeared which came within the facet range of bar. The untested ones might of course have been bar. It is un- fortunate that absolute certainty was not obtained because of the interest in progressive changes of the genes. Most of the progressive changes in characters observed in Drosophila are due to accessory factors and not to changes in a single gene. The presence of three genes at the same locus made possible an attempt to see if ultra-bar bears a direct quantitative relation to bar. If this is true, it might be considered as merely two or more bar units closely held together. The values for full, full x bar, bar, ultra-bar x full, ultra-bar x bar, and ultra-bar fe- males could in that case be represented as resulting from differ- ent amounts of a common inhibitor of facet number and there should be a consistent relation in the facet values. This, how- ever, is not true, and it seems probable therefore that the change ‘from bar to ultra-bar is specific and not merely a quantitative intensification of the bar factor. CHANGE IN THE BAR GENE OF DROSOPHILA 323 SUMMARY 1. The recognition of a pronounced and regular effect of tem- perature upon facet number in Drosophila has made possible a more accurate analysis of the germinal changes taking place in the bar-eyed races. 2. A single male with but 19 facets appeared in the second generation of downward selection in white bar, the normal range of that generation being 41 to 134 facets. 3. From this single male there was derived by crossing with his sisters a new race which has been called ultra-bar. The symbol for its gene is B’U. 4, The mean facet value of the males of this race is 23 as com- pared with 75.6 in the parental stock. 5. The ranges of ultra-bar and bar do not overlap in the males and barely overlap in the females. 6. The new stock has remained unchanged for over twenty months except for the appearance of a few individuals with marked departures from the normal range. 7. Among these mutations there have appeared reversals to full-eye and returns to a condition resembling bar in facet num- ber, but differing from it in dominance. 8. Ultra-bar differs strikingly from bar in having a much greater dominance over full, 84.8 per cent as compared with 23 per cent. 9. Correspondingly ultra-bar has a dominance of 82.6 per cent over bar. 10. The ultra-bar gene is located in the sex chromosome. 11. Crossing-over tests show that the new gene has the same locus as bar or is so close to it as to be identical for all practical purposes. 12. The change involved is to be considered as a change in the bar gene itself which is in the direction of the original change from full to bar. 13. The dominance relations, however, make it improbable that ultra-bar can be considered merely as a quantitative in- crease in the bar reaction. 324 CHARLES ZELENY 14. The fact that ultra-bar occurred in the course of a down- ward selection of bar is interesting, though probably not signif- icant, in view of other mutations affecting facet number which have not always been in the direction of selection. Acknowledgments In common with other students of Drosophila, I have had the generous help with materials and suggestions of Prof. T. H. Morgan and Drs. A. H. Sturtevant and C. B. Bridges. It isa pleasure also to express my indebtedness to Dr. Joseph Krafka, Jr., to whose untiring care of the stocks and tests the success of these experiments is largely due. I am also under obligation to Prof. A. R. Crathorne, who was kind enough to go over the method of factorial units which I have used in tabulating the data. BIBLIOGRAPHY Krarka, JosepH, Jr. 1920 The effect of temperature upon facet number in the bareyed mutant of Drosophila melanogaster. J. Gen. Physiol., v. 2. May, H.G. 1917 Selection for higher and lower facet numbers in the bar-eyed race of Drosophila and the appearance of reverse mutations. Biol. Bull., vol. 33, pp. 361-395. Sryster, E. W. 1919 Eye-facet number as influenced by temperature in the bar-eyed mutant of Drosophila melanogaster (ampelophila). Biol. Bull., vol. 37, pp. 168-182. Ticz, S. A. 1914 A new sex-linked character in Drosophila. Biol. Bull., vol. 26, pp. 221-230. ZELENY, C. 1917 Full-eye and emarginate eye from bar-eye in Drosophila without change in the bar gene. Abstracts of papers read before the fifteenth annual meeting of the American Soc. of Zoologists at Minne- apolis, p. 7; Anat. Rec., vol. 14, p. 89. 1917 Selection for high-facet and low-facet number in the bar-eyed race of Drosophila. Abstracts of papers read before the fifteenth an- nual meeting of the American Society of Zoologists, Minneapolis, p. 9, Anat. Rec., vol. 14, p. 91. 1918 Germinal changes in the bar-eyed race of Drosophila during the course of selection for facet number. Proc. Indiana Acad. Sc., 1917, pp. 73-77. 1919 A change in the bar gene of Drosophila involving further decrease in facet number and increase in dominance. Journ. Gen. Physiol., vol. 2, pp. 69-71. ZELENY, C., AND Marroon, E. W. 1915 The effect of selection upon the ‘bar eye’ mutant of Drosophila. J. Exp. Zool., vol. 19, pp. 515-529. re) fA ws B's ai Wine itis) guhe a) oe ae ns RESATRS 24 ee ar Al La / fe Resumen por el autor Harold Cummins. Universidad Tulane. El papel de la voz y la coloracién en la emigracién primaveral de la rana y en el reconocimiento sexual. Mediante una trampa que rodeaba casi por completo a una chareca en un bosque, el autor capturé ranas pertenecientes a cuatro especies, cuando intentaban entrar en la charca para criar. La emigracién tiene lugar en oleadas sucesivas, durante periodos de humedad relativa elevada coincidentes con una tem- peratura comprendida entre unos 41° y 52°F. El periodo de emigracién se prolongé hasta cuarenta y tres dias en el caso de la rana leopardo. La intensa emigracién siguié a periodos du- rante los cuales no se oy6 el canto de las ranas en la charea o en sus alrededores. Por otra parte un aumento en la emigracién no acompafé o siguié a un periodo de actividad vocal considerable. De donde el autor deduce que el impulso migratorio esta regido por factores diferentes de la voz y que esta ultima no es un factor incitante o directivo esencial. La vista no es necesaria para la copula y no parece jugar un papel importante en el reconoci- miento sexual. El autor pudo observar algunos machos intentando aparearse con otros machos tal como si fueran hembras. Los machos normales objeto de tales intentos de cépula luchan, inflan los sacos vocales, y cantan, obteniendo de este modo la libertad. Las hembras por el contrario, ofrecen generalmente poca o débil resistencia y en la mayor parte de los casos se dejan retener por los machos. El reconocimiento de los sexos, tal como se mani- fiesta en el apareamiento normal, se atribuye al comportamiento diferencial de los dos sexos cuando son retenidos por un macho. Es pues un reconocimiento ulterior que depende de la reaccién del macho que intenta la cépula hacia este comportamiento diferencial. ; Translation by José F. Nonidez Carnegie Institution of Washington AUTHOR’S ABSTRACT OF THIS PAPER ISSUED BY THE BIBLIOGRAPHIC SERVICE, FEBRUARY 2 THE ROLE OF VOICE AND COLORATION IN SPRING MIGRATION AND SEX RECOGNITION IN FROGS! HAROLD CUMMINS Tulane University Medical School While the habits of frogs have been described by numerous writers, but little emphasis has been accorded the important phases of migration into the ponds for breeding and the method of sex recognition. Following the suggestions of Prof. Jacob Reighard, the writer carried out in the spring of 1914 a series of observations devoted to these aspects of the breeding behavior. Work was conducted in White’s Wood, near Ann Arbor, Michigan, a hardwood tract containing five small ponds, one of which was selected for intensive study. This pond is bounded on one side and one end by wooded land and elsewhere by a cultivated field which adjoins the wood. During the high water of spring the pond is about 320 feet long, with an average width of 35 feet and a maximum depth of about 3 feet. Apparatus and camp supplies were installed on March 23rd; the writer remained at the pond until May 10th. Sex-recognition observations were made upon frogs both in the pond and in terraria. Data on migration, except for a few instances mentioned later, were obtained from the catch of a frog trap. The trap consisted of a 14-inch white-cloth fence supported by wooden stakes which made an angle of 135° with the earth on the inner or pondward side. The fence was placed about two feet from the edge of the water and extended more than two-thirds around the pond, on the side bordered by the wood and on the end and side next to the field. At intervals leaders of similar construction, but vertically supported, were 1 Contribution from the Zodlogical Laboratory of the University of Michigan. 325 326 HAROLD CUMMINS connected at right angles to the main fence. The lower edge of the cloth, both of fence and leaders, was imbedded in the soil and fastened with wire pins. Underneath the junction of the main fence with each leader a 2-gallon pail was sunk to the ground level and kept half filled with water. The apparatus proved to be very effective. Frogs apparently did not perceive its pres- ence until actually in contact with the cloth. When a frog approached the vicinity of a leader but a short time elapsed before he dropped into the pail, with no chance of escape. If the frog came in contact with the main fence at some distance from a leader, he usually moved along the fence and was even- tually trapped in a pail. But a less active individual might remain nestling in the angle between the fence and the ground. A careful examination of the fences and pails was made at least twice each day, morning and evening, the catch was recorded, and the frogs were transferred to terraria for observation. The fence was originally planned to completely enclose a smaller pond. Because the ice first disappeared from the larger pond, more exposed to the sun, the plan was altered, allowing only a partial enclosure of the larger pond. Future work of this sort should be done with the observation pond completely enclosed, thus ensuring the trapping of all migrating frogs. The frog fauna of White’s Wood includes the leopard frog (Rana pipiens Shreber), the wood frog (Rana cantabrigensis Baird), the pickerel frog (Rana palustris Le Conte), the green frog (Rana clamitans Latreille), the swamp tree frog (Choro- philus nigritus Le Conte), the spring peeper (Hyla pickeringii Holbrook), and the common tree frog (Hyla versicolor Le Conte). Migration data for the leopard frog, wood frog, swamp tree frog, and spring peeper were obtained, but for sex recognition work only the first two species were used. MIGRATION With the approach of spring, frogs desert their hibernation quarters for breeding places. Doubtless many of them hibernate in the mud at the bottoms of the same ponds where they breed, MIGRATION AND SEX RECOGNITION IN FROGS PAT but some winter elsewhere, perhaps in near-by bodies of water or possibly in such favorable locations on land as in masses of dead vegetation. In other locations than White’s Wood the writer has found leopard frogs hibernating under conditions fulfilling all three of the above possibilities: in the mud of bodies of water where later the species was commonly found breeding, in mud and plant material in the bottoms of streams which were swift flow- ing and unfavorable for breeding, and, lastly, in masses of swamp grass in lowland. In White’s Wood frogs might have wintered in the pond, in adjoining ponds, in springs or a creek near the observation pond, or in the adjacent wood and field. Wherever they may have been for the winter, many frogs resorted to the observation pond to breed. The questions of what incited a spring migration of those frogs hibernating away from the pond and of the conditions under which it occurred are of interest. But little benefit can be expected from direct observations, because of the relatively small number of frogs that can be noted in their migration. Banta (14) mentions his observation of two female wood frogs making their way to the pond where he studied the mating behavior of this species. While at White’s Wood no effort was made to collect data by direct observation, the following records were made. On March 27th, at 10:30 a.m., a female leopard frog was found in the field adjoining the pond, headed toward the water. In several instances ovaries and ovi- ducts, presumably of the leopard frog, were found 90 to 100 feet distant from the pond. Possibly these remains represented migrating frogs captured on their way by crows. Between 9 and 10 p.m. of March 31st, during a warm rain, dozens of spring peepers were captured while they were rapidly hopping toward the pond from its wooded side. Extended data on migration were obtained from the trap catches, presented below in table 1. Following the table are extracts from field notes which describe weather conditions for three days preceding the beginning of trap catches and during the period of maximum migration, in- cluding also observations of frogs appearing within the pond. TABLE 1 Showing separately for each sex of the four species the frogs trapped between March 26th and May 7th, inclusive, also temperature and humidity records for the period: (footnote, p. 329) [Sales ae o LEOPARD FROG | WOOD FROG He pees Hl SPRING PEEPER e : 5 3 B 3 DATE < AN & z 2 - BRN oe Male | Female) Male | Female} Male | Female} Male | Female 5 5 iz Bla 2 3 < = AERA z = aE 4 2. 3 4 1 March 26 | 41.7 94 2 1 4 Pay 51.9 100: 28 37.9 91 1 2 29 | 49.0 87 a 21 1 il 3 3 3 30 | 44.6 | 100 1 1 1 31 | 41.0 87 3 2 1 April 1 | 45.0] 100 1 1 2 42.1 92 3. 73720) | 296 AD N3Obe 48 5 | 27.9 92 6 | 30.0 78 We \a2ed 90 SESCZSaL 86 9 | 23.6 98 10 | 30.4 74 A oj: 78 12) new, 62 13 | 40.7 78 14 | 37.3 91 1 3 1 1 1 1 15) | O0ES 91 1 1 2 1 2 16 | 44.6] 100 2 i 1 1 1 lr 50.4 91 2 18 58.5 96 19 | 63.1 93 20 | 46.0] 100 1 1 21 | 33.6 94 22 | 48.0 85 1 23 «| 45.3 90 24 | 46.5 84 25 49.9 100 1 1 1 26 | 62.0 94 2 rT 56.1 Ot 2 1 28 | 61.5] 100 1 1 1 29 67.5 93 1 30 | 48.5 90 May 1 | 42.9 87 2 | 4329) |= 82 Sie anleyy 88 4) 57.7 98 Hy 65.4. 96 6 | 56.9 88 2, 7 55cll 75 23 37 9 9 ial 12 6 o MIGRATION AND SEX RECOGNITION IN FROGS 329 Extracts from field notes, March 23rd to April 3rd, inclusive March 23rd. ‘The southern two-thirds of observation pond was free from ice, while other ponds in wood were entirely frozen over. March 24th. A film of ice had frozen over the observation pond during the night. By 3 o’clock in the afternoon this film was melted with the exception of a thin strip along the west edge of the pond. In the afternoon three specimens of the swamp tree frog were observed. One of them was first noticed in the grass at edge of water; when startled it jumped into the pond, swam for a few inches and then bur- rowed into the mud. The other two were first observed swimming; they were caught and proved to be male and female. March 25th. Two leopard frogs, sexes undetermined, were first noted in the grass on west side of pond. When startled they jumped into the water. Jour swamp tree frogs were seen. All six frogs were observed in the afternoon as on the preceding day, the morning having been too cool for them to be active. March 26th. In the early morning there was a heavy mist, followed in the late morning by a hard rain which lasted until about 1 p.m. At 11:15 a.m., during the rain, a single swamp tree frog croaked inter- mittently for about an hour. At 1 p.m. the voice of one swamp tree frog was noted; it continued for 45 minutes, when two other frogs took up the chorus. Their voices continued throughout the after- noon. All afternoon, too, the voice of the leopard frog was evident. Apparently a number of individuals were croaking. At 11 a.m. three leopard frogs were seen in the water at west edge of pond, at 1:45 four of this species, and in the late afternoon about ten individuals. None were collected. March 27th. There was a rain throughout the preceding night. The air was cool and no frogs were seen or heard. March 28th. There was no croaking during the morning, but for a half-hour in the afternoon a few isolated calls of the swamp tree frog were heard. The following leopard frogs were caught in the pond: at 11:30 a.m., three males and five females, which included one clasping pair; at 1:45 p.m., one clasping pair, and in the late afternoon one male and four females. March 29th. There was no croaking during the morning until about 11 a.m., when the occasional calls of the leopard frog were noted. Occa- sionally through the afternoon a call of this species, of the wood frog, and of the spring peeper was heard, but there were no calls in the 1 Temperature and humidity records were obtained through the courtesy of Professor Hussey, Director of the University of Michigan Observatory. These records were made by instruments situated at a distance of over two miles from the breeding pond and at a higher elevation. The temperature records in general correspond with readings made at the pond; no instrument for humidity records was available for use at the pond. 330 HAROLD CUMMINS pond after 8 p.m. At 12:30 p.m., two clasping pairs; at 1:30 P.M., three females; at 3 p.M., one clasping pair, and at 4 p.m. one male, all leopard frogs, were captured in the pond. At 3 P.M., one male wood frog was caught. March 30th. No croaking occurred during the preceding night and none was heard until 9 p.m. of March 30th, when the calls of the leopard frog began. In the morning a number of leopard frogs were seen in the water, none of them except one clasping pair being taken. One female wood frog was seen. March 31st. Great numbers of unpaired wood frogs and clasping leopard frogs were noted in the pond in the afternoon. The voices of R. pipiens were first heard-at 9:30 a.m.; from that time and through the night their voices were heard. Beginning at 10 a.m., the swamp tree frogs began to croak in numbers, and in the evening the voices of the spring peeper and the wood frog were added to the chorus. At 9 p.m. all four species were found at the edge of the water. The light of the lantern dazed them, for, while continuing their calls, they did not jump readily when disturbed. April 1st. The chorus of the preceding night continued all morn- ing. Several clasping pairs of the leopard frog were noted in the water. April 2nd. A few swamp tree frogs croaked all day. April 3rd. No croaking was heard until evening, when the swamp tree frog and the spring peeper began. The numerous frogs which appeared in the pond either had hibernated therein or had evaded the trap. Both sources might have furnished individuals, the latter being possible for two reasons. All species could have entered where the fence was not continuous, and the spring peeper, also perhaps the swamp tree frog, could have adhered to the fence and climbed over. It is unlikely that the leopard frog and wood frog could have done this, for experiments demonstrated the effectiveness of the fence as a barrier for these species. With the exception of three examples, all trapped frogs were adult and sexually mature. These instances were excluded from the records of the trap, to be recorded separately. On March 29th an immature leopard frog was trapped, on April 2nd an immature wood frog, and on April 16th an immature wood frog. Unfortunately, no examinations of the reproductive organs were made for the determination of sex, but the three were so small that they could not have been sexually mature. MIGRATION AND SEX RECOGNITION IN FROGS aol, Migration occurred both during day and night. For the period including March 26th through April 2nd, the time of concen- trated migration, table 2 shows an equal distribution of day and night migrations, the totals including both sexes of the four species. The numbers of individuals of any one species or sex are too small to admit of further conclusions, except that night time seems to be more favorable for the migration of the wood frog and spring peeper. It appears, then, that the inciting TABLE 2 Showing, separately for each sex of the four species, the day and night distribution of migrations between March 26th and April 2nd, inclusive mm DAY NIGHT BA o5 in Leopard| Wood | Swamp| Spring |Leopard} Wood | Swamp} Spring ae or frog frog |tree frog] peeper | frog frog |tree frog| peeper z : Be & £ = 2) os | Oe S Oales A March 26 4| 2 34 | @ 14 27 2 4 7 28 29 1 2 3 30 a) & 3/3 12a 3 | 39 31 1 il il 3 April 1 3 2/1 6 2 1 1 2 MOGalS: seit Wty ayy al Gy al 1 | 16,4)3)4)4)2)3 74 Day total, 37. Night total, 37. stimulus for migration and the factors controlling the migration behavior operate independent of light and darkness. Temperature was an important factor in the control of mi- gration. No catches were made in the period between April 3rd and April 14th, inclusive, a time when the thermometer was low. The lowest temperature accompanied by migration was 33.6° F., and only two frogs were trapped on that date. The largest catches were made between 41° and 52° F., this range indicating an optimum temperature. No temperatures were high enough to determine a maximum. 332 HAROLD CUMMINS Humidity as well played an important réle in migration. The lowest relative humidity accompanied by migratory activity was 75 (table 1, May 7). The optimum humidity, however, was much higher, ranging between 90 and 100. It is not sur- prising to find a high humidity requirement in this group. From the data at hand it is difficult to ascertain the relative importance of temperature and humidity. Since in the cool period of April the humidity was at times as high as that favoring migration in warmer weather, and since migration temperatures were asso- ciated with high humidities, the inference is that migration necessitated a coincidence of favorable temperature and favorable humidity. The duration of the migration period was much longer than would be expected. The periods in 1914 for four species were as follows: leopard frog, 43 days; wood frog, 33 days; swamp tree frog, 35 days; spring peeper, 27 days. In other years these periods might be shortened by more favorable weather condi- tions or lengthened by adverse circumstances. If the cool weather of April had not intervened, the periods in 1914 might have been shortened considerably. Inasmuch as the frogs were hibernating under variable conditions, it is conceivable that the duration of warm weather necessary to arouse some of them would be longer than the requirement of other frogs hibernating in readily warmed localities. Further, in localities permitting frogs to hibernate under similar conditions, perhaps the wave of migration would have an early climax and be very short. In ‘support of the supposition that some frogs were aroused late, only one type of concrete evidence can be offered here. As late as April 27th there was taken in the trap an occasional leopard frog with the flabby edoematous appearance of individuals fresh from hibernation. We know nothing of the wanderings of frogs en route to breeding ponds. Inasmuch as the references to this point lead one to infer that the route is always a direct one and probably continuous, the results of an experiment leading to partially contradictory evidence is here recorded. On March 31st fourteen leopard frogs captured in the pond, two males and twelve females, were each marked by a white string tied care- MIGRATION AND SEX RECOGNITION IN FROGS 333 fully about the joint between femur and crus, then liberated in the wood about 1000 feet away from the pond. Only four indi- viduals were recovered, the remainder probably having migrated to a pond on the opposite side of the wood. These catches were not included in the trap record. The dates of capture follow: a male, April 1st; a male, April 12th; a female, April 16th, and a female, April 18th. Reference to table 1 shows that a cool spell intervened between the trapping of the first individual and of the last three. That this fact alone was not responsible for the extended time requirement of the last three individuals is evidenced first by the return of one frog as early as the next day after liberation, and second by the migrations of other frogs in the interval. To voice has been generally attributed the function of at- tracting frogs to their breeding ponds. In the species under con- sideration the males are provided with huge resonating sacs opening into the mouth cavities, whose presence results in a marked increase in the volume of sound. A general account of voice is extracted from Holmes (’12): ‘‘The voice of the male is louder and deeper than that of the female and is more often heard. In large frogs the notes are deeper than in small ones. The notes of frogs are more often heard in the breeding season, when they are supposed to serve the purpose of a sex call. In the summer, however, it is not unusual to hear the croaking of frogs, especially in the evening. A damp atmosphere is con- ducive to their song, and for this reason the voices of these ~ animals are often heard upon the approach of a shower.’”’ The writer did not hear female leopard frogs croaking during the breeding season of 1914 except under circumstances which would permit of explanation by reflex croaking in response to tactual stimulation. Males, on the other hand, not only exhibited the croaking reflex, but also croaked of their own accord. The chorus of male toads has been studied by Courtis (’07) and Miller (09). Their results indicate that voice in this form does serve as a sex call, but the writer’s study of frogs does not sup- port this view, at least it indicates that the chorus is not neces- sary to incite or direct migration. 334 HAROLD CUMMINS Trap records for certain days were made after and during periods when no calls were heard in the pond. The writer is able to discriminate the voices of the White’s Wood species and was at all times on the alert to hear them. The particular days cited below were selected because the night observations as well as those during the day were continuous. ‘Thirty-nine frogs were trapped on March 30th, more individuals than were caught in any other day; in fact, this number is about one-third of the number obtained during the entire period of trapping. Table 3 presents the catch of March 30th, the twenty-four hours be- tween 8 p.m. of March 29th and 8 p.m. of March 30th. On TABLE 3 Showing the number of frogs of each sex of the four species migrating between 8 p.m. of March 29th and 8 p.m. of March 30th, with the times of collections from the trap on March 30th SWAMP LEOPARD FROG| WOOD FROG SPRING PEEPER TIME OF COLLECTION TREE FROG FROM TRAP TOTALS- Male | Female} Male | Female} Male | Female} Male | Female 8.00 a.m. 1 1 1 3 ale 1.45 P.M. 4 6 10 5.00 P.M. 3 3 3 3 12 Motels. ave eee 7 21 1 1 3 3 3 39 March 29th there was no croaking in the pond until about 11 A.M., when an occasional croak of the leopard frog was heard. In the afternoon, and continuing until 8 p.m., the voices of the leopard frog, wood frog, and spring peeper were noted, but the chorus did not continue thereafter. All frogs were silent until about 9 p.m. of March 30th. Conversely, the catch of April 1st, while not as small as that of some other days, is interesting because of the relatively small total trapped when climatic conditions were favorable and much croaking occurred. The average temperature and hu- midity for the twenty-four hours ending at 7 a.m. April Ist (table 1) were favorable to migration. The total catch, three female leopard frogs, two female swamp tree frogs, and one MIGRATION AND SEX RECOGNITION IN FROGS 335 male spring peeper, represent frogs migrating between 9 p.m. of March 31st and 8 a.m. of April Ist. Throughout March 3lst, beginning at 9:30 a.m. the leopard frogs were croaking, the swamp tree frogs began at 10 a.m., and in the evening the other two species started their chorus. All four species called during the night. Comparison may be made with the catch of March 27th, a day characterized likewise by a small catch, favorable weather conditions, and croaking, but differing in that it pre- ceded rather than followed the day of largest total catch, March 30th. As on April Ist, the temperature lay within the optimum range and the air was saturated to 100. The catch included two male wood frogs, one female wood frog, and four male swamp tree frogs, captured in the period between the evening of March 26th and 8 a.m. March 27th. All afternoon and into the night of March 26th the swamp tree frog and leopard frog chor- used, and at 11:30 p.m. the spring peeper began to call. One of the direct observations can be applied in this connection. The single female leopard frog which was picked up in the field on the morning of March 27th was making her way toward the pond despite the fact that there was no croaking. To ascribe a directive function to voice seems to be unwar- ranted in the light of migrations which occur without the pres- ence of this suggested factor. In order to perform a directive function, the chorus must be accompanied by migration, and we do not always find the two coincident. On March 30th large numbers of frogs migrated because of favorable climatic con- ditions; warm weather had been prolonged enough to arouse them and the weather conditions were such as to allow an overland trip. Few individuals migrated on April 1st; the small number admits of explanation on the ground that the migratory climax had passed, that a longer duration of warm weather would be necessary to arouse the frogs remaining in less readily warmed locations. The small catch of March 27th does not indicate necessarily that voice was not effective, rather that up to this date only those frogs in the most easily warmed situations had been aroused. The immediate inception of the migratory im- pulse must be intrinsic; physiological processes with which this THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 30, No. 3 336 HAROLD CUMMINS problem is not concerned (probably associated with the state of the reproductive organs, since with the exception of three imma- ture frogs only sexually mature individuals migrated) govern the impulse, and are operative when external circumstances become favorable. Because migration is successfully accomplished without the directive influence of voice, we must look elsewhere for factors controlling the direction of migration. In such a closely related group as the salamanders there are some forms which lead a terrestrial existence at all times except during the breeding season, when they resors to water. (In the trap a num- ber of examples of Ambystoma punctatum and several A. tigri- num were obtained.) Here as well as in other groups which are voiceless the factor of a vocal attraction is unquestionably eliminated. Neither can we explain by a vocal factor the exodus of frogs from their breeding places after the termination of the breeding season, or the emigration of newly transformed frogs. To substantiate the idea that voice is not effective, there are the results of Yerkes (’05) on auditory responses in frogs. Yerkes finds that while frogs possess a fairly well-developed sense of hearing, its function seems to be “a warning sense which modi- fies reactions to other simultaneous or succeeding stimuli.” He does not find evidence that it serves as an independent control of motor reactions. SEX RECOGNITION The results of Holmes on Amphipods (’03), of Pearse on cray- fish (09), and of Reighard on the brook lamprey (’03) show that in these animals sex recognition is established through the reac- tions of those individuals with which the breeding act is attempted by males. That is, there is really no precopulatory recognition, rather a male may attempt union with any individual, and the reactions of that individual determine whether or not the union shall continue. Holt’s observations on the dragonet (’98) and Miss Reeves’ study of Etheostoma (’07) show that behavior is the criterion of recognition in these fishes. However, in these instances there is a visual recognition before the onset of the breeding act. Banta (’14), working on the wood frog, concludes MIGRATION AND SEX RECOGNITION IN FROGS 337 that “the color of the female may possibly be a factor and that the behavior of the female is probably a factor in sex recognition.” While Banta notes that the usual procedure is the indiscriminate testing of many individuals regardless of their sex, he neverthe- ess postulates a precopulatory recognition established through sensory channels. This writer does not attribute any signifi- cance to the postcopulatory reactions of clasped frogs; in fact, he finds no consistent difference in the resistance of the two sexes when clasped. Miller (’09), on the other hand, finds in toads that ‘‘males cannot distinguish at sight males from females. For this reason they are continually clasping one another. They have a call of three or four notes which they utter in rapid suc- cession when taken up between the finger and thumb, or clasped by another male. This seems to be a warning signal, for a male will release another as soon as he chirps.”’ In an effort to obtain evidence on sex recognition, the writer observed mating reactions in the wood frog and leopard frog, both under natural condi- tions in the pond and under experimental conditions in terraria In both species, pursuit under natural conditions did not differ whether the frog pursued was male or female. In this respect the observations are contrary to those of Banta, who states that ‘‘The beginning of the attempt of a male upon a female is of course not in any way different from his approach toward another male, but when he actually touches or often only nears the female his actions are usually very different, for instead of the vigor and aggressiveness of the assailant rapidly falling off, as in the case of one male approaching another, the aggressiveness is tremendously increased.’”’ In practically every instance none but moving frogs were pursued, although at a distance of a foot or less a quiet frog might be attacked, and a quiet frog if touched by a male usually was attacked, in both cases regardless of sex. A male upon which clasping was attempted sometimes only croaked when touched; if the attempt to clasp him were con- tinued, violent struggles, croaking, and inflation of the vocal sacs were followed by dislodgment of the clasping male. Fe- males when clasped were occasionally passive, sometimes strug- gled very vigorously, and sometimes struggled only a little. 338 HAROLD CUMMINS Usually a male continued his efforts to clasp a female despite her struggles, but frequently he desisted, even in a few instances when her opposition was but weak. More instructive data were obtained from observations in terraria. Some of them are presented below; with the exception of nos. 5 and 10, each observation or experiment was repeated, with results no different from the cited cases. Observations in terraria 1. Wood frog, April 1st. The frogs of a pair were separated and placed in a large glass container with 14 inches of water. The male swam first in one direction and then in another, with no relation to the position of the female and without touching her. After a short time he stopped about 2 inches behind and to the right of the female. Then with a sudden jump he mounted, encountering no resistance. After being again separated, the male swam about as before. The female alternately crouched at the bottom of the dish and swam erratically. Activities of this nature followed for twenty minutes, when the frogs encountered each other snout to snout. Suddenly the male clasped the female about her head and then gradually worked backward until he secured a hold in front of her fore legs. After two minutes a sudden maneuvre placed the male in the correct position for a pectoral clasp. During this procedure neither frog croaked and the female did not resist. 2. Wood frog, April 1st. A male and a female which had not been clasping were placed in a container with leaves but no water at the bottom. With his fore legs touching her, the male crouched at the side of the female. Suddenly he jumped upon her and attempted to clasp, at the same time squawking loudly. But the female resisted, jumping up on her hind legs and turning the ventral surface upward. She succeeded in dislodging the male and he did not again try to clasp her. This clasping. and dislodgment occupied not more than a half minute. 3. Wood frog, April 1st. A number of individuals of both sexes were confined in a terrarium. A male secured a clasp upon another male. The clasped male resisted, croaking and turning on his back. After a few seconds the clasping male loosed his hold. 4. Wood frog, April 2nd. Three males which had been clasping were placed in a terrarium with a male so disabled that he could not use the hind legs. After a few minutes the disabled male was clasped, and was held so for several hours. 5. Wood frog and leopard frog, April 1st. Two male wood frogs were confined in a terrarium with about twenty leopard frogs of both sexes. On the morning of April Ist, after two days in the terrarium, © one of the wood frogs clasped a female leopard frog. He experienced MIGRATION AND SEX RECOGNITION IN FROGS 339 a little difficulty in gaining a hold upon so large a mate, but as she did not oppose his efforts, sueceeded. The clasp continued for eight days. 6. Wood frog, April 1st. At 4 p.m. on this date five males and three females were placed together in a light-proof container. At 8 a.m. of the next day the container was opened, and all three females were found clasped by males. In the interval, especially at first, croaking was heard within the container. 7. Wood frog, April 1st. In a terrarium containing a number of this species one male clasped another. The clasped male resisted, turning on his back and croaking loudly. After a half minute the clasping male loosed his hold. 8. Leopard frog, March 30th. Two males which had not been clasp- ing and one male from a pair were placed in a terrarium with a female from a pair. One of the males obtained a clasp on the female, but she struggled so violently as to succeed in dislodging the male. 9. Wood frog, March 31st. A female was introduced into a terrarium containing two clasping pairs and three single males. Within ten minutes one of the males tried to mount her; he approached from her left side, but she pushed him back with her fore leg. He persisted in attempting to mount, and after three minutes her resistance became more marked—jumping about and turning the ventral surface upward. In spite of this opposition, the male grasped her; within seven minutes he secured the usual clasp and remained. 10. Wood frog and Ambystoma, April 1st. Three single males and two females which had not been pairing and a clasping pair were placed in a terrarium with a female Ambystoma punctatum. The terrarium contained no water, only damp leaves and grass at the bottom. In a few minutes two males clasped the salamander, both with their heads directed toward her posterior end, one of them clasp- ing her head and the other her body a little behind the forelegs. In nine minutes the third male clasped the salamander midway between the two pairs of legs. All three held tightly. When she tried to dis- lodge them their clasp tightened and their hind legs were braced against the body of the salamander. In forty-five minutes the third male dropped off, but in twelve minutes returned and secured a clasp in approximately the same position as before; he kept his vocal sacs inflated and croaked repeatedly. After two or three minutes he let go, the salamander writhing and struggling. At 1:05 p.m. the male which had clasped her head let go, and at 9 p.m. the one clasping back of her fore legs did likewise. During the last hours of clasping the salamander lay quietly. 11. Leopard frog, March 30th. On the evening of March 30th seven- teen females and twelve males were placed in a terrarium, rather closely crowded. After an elapse of twelve hours there was only one clasping pair. 12. Leopard frog, March 31st. Two single males, a single female, _ and the male and female separated from their clasp were placed in a terrarium at 9:15 a.m. At 10:45 both females were clasped; at this 340 HAROLD CUMMINS time the pairs were separated and the frogs returned to the terrarium. At 1:45 p.m. only one of the females was clasped. 13. Leopard frog, March 30th. Several individuals of both sexes were placed in a terrarium. When touched by a female one of the males croaked forcibly. The terrarium was crowded and the female was climbing over him. 14. Leopard frog, March 30th. Whenrubbed or touched, even lightly, by other males (in the same terrarium as no. 13) the response was the same as when a female touched a male. 15. Wood frog, April 1st. A clasping pair of wood frogs, three males, and two females were placed in a terrarium. A single male approached the pair, and when he touched them the clasping male croaked and warded off the intruder with his hind legs. The approaching male made no effort to clasp. From the foregoing observations, it is evident that males attempt to clasp individuals of both sexes under experimental con- ditions as well as in the pond. In no. 1 the female did not resist and the male retained his clasp. On the other hand, in no. 9 she resisted strenuously, but the male succeeded in clasp- ing and remained in spite of the resistance. In no. 2 and no. 8 the females dislodged the males. Thus there seemed to be no consistent reaction of clasped females. When males were touched by other frogs (nos. 13, 14 and 15) they croaked, and this alone seemed to be at times sufficient to frustrate further attempts of the approaching males. But sometimes, as in no. 3 both resistance and vocal remonstrance followed an actual attempt to clasp. In no. 4 the clasped male was unable to offer resistance, and the clasp in this case was retained for several hours. In agreement with Banta, there was no consistent difference in the resistance offered by the two sexes, except that males, unless experimentally disabled, always croaked or resisted, or did both, while females sometimes resisted and sometimes failed to do so. Neither was there a consistent ardor of the at- tacking males; sometimes, in spite of resistance offered by a fe- male, he struggled until a hold was established; at other times the males were not persistent in their attempts upon females. But attempts on males were not continued. Pairing did not occur in all cases when frogs were placed in terraria, asin no. 11. It is significant that on the two days preceding this experiment - MIGRATION AND SEX RECOGNITION IN FROGS 341 and on the day of the experiment only few pairs were seen in the pond, while numerous single individuals were noted. The necessity of a visual factor was eliminated by no. 6, where correct coupling occurred in the dark. The efforts of clasping frogs were not confined to their own species, or even to frogs, as no. 5 and no. 10 indicate. Holmes (’12) shows that these extraspe- cific claspings are even more extensive. The variation in resistance offered by females may be tenta- tively explained by a gradual development of the physiological state favorable for clasping. Passive acceptance of the clasping male may be associated with the optimum development of this state, while resistance may indicate that the female is not yet ready or perhaps has already undergone the climax. In the same manner, there may be a gradual development of the clasp- ing impulse in the male, for under natural conditions it is lost after the termination of the breeding season. ‘The less persistent efforts to maintain a hold may be associated with a small degree of development of the impulse. And the ardor which results in the clasping and retention of the clasp on females, disabled males, salamanders, frogs of other species, etc., may be associ- ated with the maximum development of the impulse. Now, if males attempt clasping with both sexes of their own species, with other animals and objects, it seems that sight plays no part except to inform the male that there is something to be clasped. In the light of extraspecific pairing (no. 5 and no. 10), it seems absurd to attribute any role ot sight in sex recog- nition, either on a basis of color or behavior. That sight is not even essential, that other stimuli are responsible for correct coupling, was shown by no. 6. It has been pointed out that females will sometimes resist when clasped by males, that normal males always resist. ‘Too, it was shown that following the resistance of the female the male might or might not continue his efforts, while in the case of males he did not persist. The resistance of the males consisted not only in struggling, but also in the inflation of the vocal sacs and croaking; in fact, attacked males did at times cause the desistance of the attacking males by only croaking when they 342 HAROLD CUMMINS were touched. Apparently, then, the recognition is based upon the reaction of males. The reaction which establishes recogni- tion is a combination of vocal remonstrance and struggling which stimulates the clasping male through tactual (and kines- thetic ?) sensations. Again, according to Yerkes (’05), sounds reinforce tactual stimuli and result in motor responses, and the sense of hearing serves as a warning sense to modify reactions to other stimuli. - SUMMARY 1. By means of a trap nearly enclosing a pond, frogs of four species (Rana pipiens, R. cantabrigensis, Chorophilus nigritus, Hyla pickeringii) were caught as they attempted to enter the pond for breeding. 2. It was found that migration occurred in waves, during periods of high relative humidity coincident with temperature ranging between about 41° and 52° F. 3. By continuous short-period day and night records of the croaking of frogs in the pond it was found that intense migration followed periods during which there was no croaking in the pond or about it and that great vocal activity was not followed or accompanied by increased migration. It is concluded that voice does not direct the movement of the frogs into the pond. 4, Observations in the open and numerous experiments on frogs in terraria lead to the conclusion that sight plays no role in the attempt of the male to clasp the female except to inform him that there is something to be clasped. Sight was not found to be essential for correct coupling and is believed to play no réle in sex recognition. 5. Males were found to clasp other males as well as females. Clasped normal males struggle, inflate the vocal sacs and croak, and are always released. Clasped females show usually brief and weak resistance and the clasp is nearly always retained. Sex ‘recognition’ as manifested in normal pairing thus results from the differential behavior of the two sexes when clasped, and depends on the reaction of the clasping male toward this differential behavior. MIGRATION AND SEX RECOGNITION IN FROGS 343 LITERATURE CITED Banta, A.M. 1914 Sex recognition and the mating behavior of the wood frog, Rana sylvatica. Biol. Bull., vol. 26, pp. 171-183. Courtis, 8. A. 1907 Response of toads to sound stimuli. Amer. Nat., vol. 41, pp. 677-682. Homes, 8. J. 1903 Sex recognition among amphipods. Biol. Bull., vol. 5, pp. 288-292. 1912 Biology of the frog. Macmillan Company. Hott, E.W.L. 1898 On the breeding of the dragonet. Proc. Zool. Soc. Lond., pp. 281-315. Miturr, N. 1909 The American toad. Amer. Nat., vol. 48, pp. 641-688 and 730-745. Pearse, A. 8. 1910 Observations on copulation among crawfishes with special reference to sex recognition. Amer. Nat., vol. 43, pp. 746-753. ReEEvES, Cora D. 1907 The breeding habits of the rainbow darter (Etheostoma coeruleum Storer), a study in sexual selection. Biol. Bull., vol. 14, pp. 35-59. REIGHARD, JacoB 1903 An experimental study of the spawning behavior of Lampetra wilderi. Science, N.S., vol. 17, p. 529. Yerkes, R. M. 1905 The sense of hearing in frogs. Jour. Comp. Neur. and Psych., vol. 15, pp. 279-304. Resumen por los autores, Henry Laurens y Henry Daggett Hooker. Laboratorios Zool6gico y Botdnico Osborn, Yale University. Estudios sobre el valor fisiol6gico relativo de las luces espectrales. II. La sensibilidad del Volvox a longitudes de onda de igual contenido de energia. La sensibilidad del Volvox a la radiacién de diferentes longi- tudes de onda e igual energia (sensibilidad a la radiacién de la misma energia) ha sido investigada por los autores siguiendo dos métodos: a) la duracién relativa del tiempo de presentacién, y b) la velocidad relativa de la locomoci6n (y precisién de la ori- entacién). Una longitud de onda de \ 494u tiene el mayor valor estimulante, como demuestran ambos métodos. La eficacia de las otras longitudes de onda presenta una disminucién gradual cuando se emplean longitudes de onda mas o menos largas. Los autores insisten sobre la necesidad de usar un espectro de igual energia en tales investigaciones. También llaman la atencién acerca de la presente incapacidad para hacer comparaciones con la vision humana (normal o dalténica), es decir, con las curvas de luminosidad fotépica y scotépica (o acromatica) por las sigui- entes razones: a) a causa de nuestra ignorancia sobre las reac- ciones fotoquimicas y la naturaleza de las substancias fotosensi- tivas, y b) porque las investigaciones efectuadas recientemente sobre la visibilidad de la radiaci6n por el ojo normal y dalténico demuestran que el efecto maximo no depende de las longitudes de onda mencionadas en investigaciones mas antiguas. Translation by José F. Nonidez Carnegie Institution of Washington AUTHOR’S ABSTRACY OF THIS PAPER ISSUED BY THE BIBLIOGRAPHIC SERVICE, FEBRUARY 23 STUDIES ON THE RELATIVE PHYSIOLOGICAL VALUE OF SPECTRAL LIGHTS II. THE SENSIBILITY OF VOLVOX TO WAVE-LENGTHS OF EQUAL j ENERGY CONTENT HENRY LAURENS AND HENRY D. HOOKER, JR. Osborn Zoological and Botanical Laboratories, Yale University TWO FIGURES INTRODUCTION The present paper is one of a series dealing with the deter- mination of the relative stimulating effect of radiation in dif- ferent parts of the spectrum. The experiments to be described were made during the summer of 1917, but due to the pressure of other duties occasioned by the war it has been impossible to prepare the paper for publication until the present time. The work as it is now presented is not complete, delay in the receipt of apparatus has prevented us from continuing the work, and a number of questions are reserved for future investigation. In- complete as the present work is, however, it is considered inad- visable to delay publication longer. A preliminary account of the results has appeared (Laurens and Hooker, 718). The determination of the relative visibility of radiation, or the visibility of light of different wave-lengths, consists in find- ing the relation between luminous sensation—‘light’—and radiant energy. The sensibility of the human eye, or the visi- bility of the radiation, at any wave-length is the ratio of the luminous intensity, measured in light units, to the intensity of the light measured in energy units. The curve obtained is the visibility for an equal-energy spectrum. The same general procedure can, and should, be applied to the determination of the relative physiological value of light of 345 346 HENRY LAURENS AND HENRY D. HOOKER, JR. different wave-lengths for other organisms. In our work the energy units are equated as described in the first paper of the series (Laurens and Hooker, 717). The ‘light units’ are the figures calculated from rate of movement, or time taken to tra- verse a certain distance, or the presentation or action time, ete. Under these conditions, the sensibility of the organisms to the various lights of different wave-lengths is sundae to the reciprocal of the ‘light units.’ In our first paper (Laurens and Hooker, *17) are set forth additional reasons why it is essential that work of this sort be done with an ‘equal-energy spectrum’ (also Sheppard, pp 102-3). It will not be out of place to mention in passing the work which has been done more or less recently, principally by phy- sicists, or from the physicists’ point of view, on the visibility of radiation, or luminosity at constant spectral energy. In this work the distribution of the energy in the spectrum is always taken into consideration. We wish to call attention to the con- tributions to this subject by Nutting (’08), Thiirmel (710), Ives (12), Bender (’14), Nutting (15), Coblentz and Emerson (17), Luckiesh (’17), Reeves (18), and Hyde, Forsythe and Cady (18). It is with these and similar results concerning the relative stimulating value of lights of different wave-lengths for the human eye, in which the luminosity process is considered independently of the color process, that comparisons of the relative stimulating value of light of different wave-lengths for photosensitive organisms, or for photosensitive protoplasm in general, must be made. It seems best to avoid the use of the word ‘color’ in the pres- ent connection, since we are dealing primarily with the objective properties of radiation. Color is a psychological fact, depen- dent upon the integration of physical and physiological bases. In our work on a variety of organisms we are using a physical basis to gain information concerning the physiological basis, and from a strictly objective point of view, for comparison let us say, with the human eye. To say that the physiological effect of radiation consisting of a limited number of wave- SENSIBILITY OF VOLVOX TO SPECTRAL LIGHTS 347 lengths represents ‘color’ to a non-differentiated bit of proto- plasm, is to confuse physical, physiological, and psychological qualities of radiation. The plan of work outlined includes the carrying out of experi- ments with lights such as we have described, to determine the visibility of light in the different parts of the spectrum, as well as the relative effects on the size of the pupil. Furthermore, the study of the action currents of the eye of various animals when stimulated by the various wave-lengths will furnish data and information regarding the relations between the photochemical and photo-electric effects. Determinations of the relative effectiveness of the same or similar lights for different organisms will furnish a fundamental basis for comparison with the relative stimulating value, as determined by luminosity curves, for the human eye. APPARATUS The apparatus used to obtain the lights of different wave- lengths but of equal radiant power is described in the first paper of the series (Laurens and Hooker, 717). In addition to the twenty-three lights there listed, each 30 my wide, and extending from d 420 mu to » 670 mu, a white light was made equal in ra- diant energy content to the various spectral lights. This was used in a balanced relation to the various spectral lights, as will be described below. A small glass aquarium (26 mm. x 26 mm. and 10 mm. deep) was made to hold the organisms under observation. For initial orientation in the beam of light, the organisms were placed in a trough made by placing two strips of celluloid parallel to the direction of the beam of light. The aquarium was placed on the stage of an ordinary dissecting microscope, with a glass plate in the stage aperture. Twelve centimeters below the level of the stage there was a ruby glow-lamp on a weak Columbia dry cell, giving a very faint illumination. This lamp was needed only when the shorter or longer wave-lengihs, or the white light were being used, the organisms being clearly visible in the majority of the spectral lights. It had no demonstrable 348 HENRY LAURENS AND HENRY D. HOOKER, JR. effect, the results obtained from the lights in which the organ- isms were visible being the same when it was lighted as when it was not. a MATERIAL Volvox globator was found in the spring and summer and early fall of 1917 in Mill River near New Haven. Material was collected two to three times a week, and kept fresh and cool by placing the glass vessel containing the colonies in running water in a battery jar. The colonies were dark-adapted for at least an hour before exposure to the light, the stimulating effect of which was to be tested. Only photopositive colonies were used. EXPERIMENTAL The reactions, orientation, etc., of Volvox to white light have been carefully observed and analyzed (Holmes, ’03; Mast, ’07, 11). We are not primarily interested at the present time in the problem of orientation, that is, as to whether Volvox orients directly or indirectly; nor in the question as to whether the response to light is occasioned by a change in light intensity (time rate of change) or by the continuous action of light, al- though our results have a bearing on these questions, particu- larly the latter. The experiments may be divided into two main parts: a) those dealing with the determination of the presentation or action time, and, 6) those dealing with the determination of the relative rate of locomotion. a. Determination of the presentation or action time By the presentation time is meant the minimum time for which the organism must be stimulated or acted upon, by a stimulus of constant strength in order that a motor reaction be elicited. For the human eye, the action time, the time required for a stimu- lus to produce a sensation of maximum luminosity, is a function of intensity, not of color. SENSIBILITY OF VOLVOX TO SPECTRAL LIGHTS 349 Between the ocular end of the telescope of the spectrometer and the small glass aquarium there was placed a shutter with stops for 1, 0.5, 0.2, and 0.04 seconds. For exposures of longer duration than one second, the ‘bulb’ was used and timed by a stop-watch. The aquarium was put in place on the microscope stage, so that the band of spectral light impinged on one wall, and the white light on the opposite wall, the shutter was then closed and a Volvox colony placed in the aquarium. Being stimulated by the white light it moved toward its source, that is, into the portion of the trough of the aquarium farther away from the source of the spectral light. The white light was then screened and the colony observed for a moment or two from above and from the side to make certain that there was no hori- zontal movement. It was easily seen that some of the colonies, after forward motion was thus stopped, swam slowly upward, while others hung apparently motionless. The colony was then exposed to the spectral light by opening the shutter set for the shortest exposure, or for one which it was reasonably certain would have no effect. If no reaction fol- lowed, the organism was given the same exposure again, and usually a third time, allowing adequate intervals of time and taking precautions that no fortuitous horizontal movement either in the direction of or away from the source of light was taking place at the moment of exposure. The duration of the exposure was then increased until a reaction—movement toward the source of light—was obtained. In the more effective lights the colonies would often continue to the end of the trough, although, as described, they had been exposed to the light for only the short presentation time, which was succeeded by the reaction time. Ten colonies were used for each light and each colony exposed several times and the results averaged. They are shown in table 1, column 3, the figures representing the average minimal duration of exposure below which no effect is produced. It is now well established that in order for light to produce a definite degree of effect the time required is inversely propor- tional to the intensity of the light The primary photochemical 300 HENRY LAURENS AND HENRY D. HOOKER, JR. action of the light will not be effective until it has reached a certain minimum. To the production of this minimal photo- chemical effect a certain fixed amount of energy is necessary. The minimal duration of exposure below which no effect is pro- duced is an indication of the physiological intensity, or stimu- TABLE 1 The relative stimulating value of spectral lights of equal energy content as ascer- tained from the determination of the minimal duration of exposure necessary to produce a reaction RELATIVE STIMULATING VALUE PRESENTATION TIME NUMBER WAVE-LENGTH IN MUL IN SECONDS Bees any sec A Max. = 100 1 434.0 4.6 2.17 2.21 2 444.0 3.2 3.12 3.18 3 454.0 1.8 5.90 5.66 + 464.0 1.4 7.14 7.29 5 474.0 0.53 18.8 19.2 6 484.0 0.182 55.5 56.6 7 494.0 0.102 98.0 100.0 8 504.0 0.116 83.4 85.1 9 514.0 0.23 43.0 43.9 10 524.0 0.77 13.0 13.3 ih 534.0 1.8 5.59 5.66 12 544.0 2.0 5.0 5.10 13 554.0 3.6 Pott 2.83 14 564.0 4.1 2.44 2.49 15 574.5 4.6 Ze 17/ 2.21 16 584.5 5.9 1.69 1.72 17 594.5 6.1 1.64 1.67 18 605.0 1.2 1.39 1.42 19 615.0 7.5 1.33 1.36 20 625.0 8.8 1.14 1.16 21 635.0 10.3 0.98 1.00 22 645.0 ib 7 0.85 0.87 23 655.0 12.4 0.81 0.83 lating value, of the light. The relative stimulating value may be ascertained by taking the reciprocal of the presentation time, since this represents the time necessary to produce a constant quotient of change and since all the lights are of equal physical intensity. These values are shown in column 4 of table 1. For plotting, the relative stimulating efficiencies are expressed as SENSIBILITY OF VOLVOX 10 SPECTRAL LIGHTS aor percentages of the wave-length of maximum efficiency. The stimulating values thus obtained are to be found in column 5, the curve in figure 1. The band of spectral light with its center at \ 494 mu has the greatest stimulating value. The stimulating efficiency of the wave-lengths decrease rapidly and fairly sym- metrically on either side of this wave-length, being the same for \ 454 my and 534 mu. Ea Zes ce Ea ere MES BAe pale sar La FR ed RELANVE STIMULATING VALUE 420 440 A6O 480 500 S20 S40 560 580 GOO 620 G40 cso MU Fig. 1 The relative stimulating value of spectral lights of equal energy con- tent as ascertained from the determination of the minimal duration of exposure necessary to produce a reaction. Between the application of the stimulus and the initiation of the motor reaction, there is an interval, the reaction time. There is also an interval between the cessation of the external action of the stimulus (i.e., the exposure to the light) and the first evidence of the motor reaction, the latent period, during which the organism need not be illuminated. The reaction time therefore may be divided into two parts: the presentation time, which is the minimal duration of exposure below which no effect THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 30, NO. 3 352 HENRY LAURENS AND HENRY D. HOOKER, JR. is produced, and during which the primary photochemical effect proceeds to a certain minimal amount, e.g., by transformation of a substance already existing into another, or ‘inner stimulus,’ which gives rise to the impulse; and the latent period, during which a chemical process, either a continuation of the one started during the presentation time, or others, are taking place. These latter’processes are not completed until at least a threshold amount of substance has been formed, or transformed, and the excitation, thus set up, conducted to the locomotor organs, or cilia, in Volvox. Unfortunately, complete records were not made of the duration of either the reaction time or of the latent period. This is one of the points which, as mentioned above, we desire to clear up later. In our early work on the determination of the relative velocity of locomotion when the colonies were exposed to the various spectral lights, we recorded in a number of cases the time from the beginning of the exposure to the light up to the first evidence of locomotor response, that is the reaction time. We will refer to this again in the portion dealing with the rate of movement. Hecht (’18) has obtained some very striking results in connec- tion with the latent period portion of the reaction time of Ciona to light. He demonstrated that the latent period was a constant quantity under considerable variations in intensity, while the sensitization period (the presentation time) varied with the in- tensity of the light. The sensitization period, which is the reac- tion time minus the constant latent period, is the important factor of the reaction in connection with the intensity, since it indicates the quantity of stimulus received by the organism. The product of the sensitization period and the intensity proved to be a constant, and thus the applicability of the reciprocity law of Bunsen and Roscoe to the stimulative action of light has again been confirmed. Mention may be made here of certain other points in Hecht’s (19 a, b, ec) contributions to the analysis of photochemical re- actions. In the first place, he shows good reason for believing that the photochemical process involved in the sensitization SENSIBILITY OF VOLVOX TO SPECTRAL LIGHTS 353 period is of the nature of a simple reversible chemical reaction with increase of energy, taking place in a homogeneous medium. The latent period, however, involves a catalytic reaction in which the catalyst remains after the action of light, and is therefore a reaction with loss of energy. There are two points in Hecht’s work which we have not been able to make clear to ourselves, the first of these concerns the latent, the second the sensitization period, into which two parts Hecht (18, p. 152) divides the reaction time. Although he repeatedly states (’18, pp. 162, 165; 719 a, pp. 547, 548) that under given conditions of temperature and intensity the latent period is constant, he makes the following contradictory remarks (Hecht, 18, p. 153): “It will be seen that, within limits, the shorter the exposure time (sensitization period) the longer the reaction time and consequently the latent period;” (’19b, p. 661): ‘There- fore the latent period also varies inversely with the duration of the exposure.” In the above quotations exposure time and sensitization period appear to be synonymous, but in the following passage (Hecht, 19 b, p. 659) a distinction is made: All the experiments agree in showing that for a given intensity the reaction time varies inversely with the exposure for exposure periods shorter than the sensitization period. Exposures for intervals greater than the sensitization period make no change in the duration of the reaction time. The sensitization period may thus be defined as the minimum exposure necessary to produce the minimum reaction time. This distinction leads to confusion. Moreover, it is contra- dictory to the Bunsen-Roscoe law, which Hecht (718, pp. 155, 165, and fig. 1, p. 156) found applied to Ciona and later (719 a, p. 548) stated ‘‘has been shown to apply to such sensitization processes.” For a given intensity there can be but one presentation time. This term is used instead of sensitization period to avoid con- fusion. An exposure shorter than the presentation time is sub- liminal. The striking similarity between the reaction-exposure curve (Hecht, ’19b, p. 659, fig. 1) and the reaction-temperature curve (Hecht, ’19 ¢, p. 671, fig. 1) suggests a possible explanation 304 HENRY LAURENS AND HENRY D. HOOKER, JR. of the effects of exposures longer than the presentation time, which Hecht describes. In any case, the date given in experi- ment 9.4 (Hecht, ’18, p. 153) and figures 1 and 2 (Hecht, ’19b, - pp. 659, 660) are insufficient, and the experimental error is too great to justify the setting ‘aside of the Bunsen-Roscoe law. b. Determination of the relative rate of locomotion and precision of orventation The same general procedure was followed in the determination of the relative rate of locomotion as was described for the deter- mination of the presentation time. A colony of Volvox was placed in the trough of the small glass aquarium. ‘The spec- trometer was set to deliver wave-lengths of a certain distribution, but the aquarium screened so that the organism was not exposed to their action. By means of the white light of equal radiant power the colony was stimulated to swim to the end of the trough away from the source of the spectral light the stimulating value of which was to be tested. After, or just before, the colony reached the ends of the trough, the white light was screened and the organism, as it stopped forward motion and hung motion- less or moved slowly upward, was observed by the light from the ruby glow-lamp. The colony was then exposed to illumination by the spectral light. At the first indication of movement toward the source of light a stop-watch was started, and when the colony reached the other end of the trough it was stopped. The time it took the colony to swim the distance of the trough was thus obtained. As the colony reached the end of the trough, the light - being tested was turned off, and the colony observed for a few seconds as it again hung motionless or moved slowly upward. The white light was then turned on and the time it took the col- ony to reach the other end (the end nearer the source of white light) was taken. The reciprocal of the ratio of these two ‘times’ was taken as the index of the relative stimulating value of the wave-lengths in question. By taking in this way alternately the times required to swim toward the source of the white light and toward the source of the spectral light, the stimulating value of SENSIBILITY OF VOLVOX TO SPECTRAL LIGHTS 355 which was being tested, the influence of change in ‘physiological condition,’ or of an increase or decrease in speed of locomotion and of precision in orientation, in successive tests, was eliminated. A colony was usually given five trials, in each direction and the respective times taken to swim toward the spectral light averaged, as were those taken to swim toward the white, and the ratio between the two rates taken. From four to sixteen colonies were thus given trials in each light. In the results obtained from the determination of the presen- tation time the reciprocal of this value was taken as the stimu- lating efficiency of the light. In comparing the relative rate of locomotion, we likewise take the reciprocals of the time re- quired to travel a certain distance. In table 2 the ratios of the rate of locomotion toward the source of the spectral lights to the rate of movement toward the white light are listed in column 3. The relative stimulating values of the various spectral lights calculated by taking the reciprocals of these ratios are listed in column 4, and as percent- ages of \ maximum in column 5. The curve is shown in figure 2. An extensive series of tests was also made of the stimulating value of the various wave-lengths while the colonies were ex- posed at the same time to the white light and to the spectral light, the particular light whose stimulating value was being tested impinging at right angles on one wall of the aquarium, the white light on the opposite wall. The speed of locomotion to the first fifteen of the spectral lights while the white light was also burning was then ascertained, and the speed of locomotion to - the white light alone, as above, and the ratio again taken. These values are shown in column 6, table 2; the reciprocal of them, or the relative stimulating value, in column 7, and the values computed by regarding } maximum as 100 in column 8, which values have also been plotted (fig. 2). In the region of the spec- trum of highest stimulating effects the values are not very differ- ent from those obtained for the spectral light alone (columns 5 and 8). The reason for this is probably that the stimulating effect of the wave-lengths of greatest efficiency, and of those in the immediate neighborhood in the spectrum, is relatively so great 356 HENRY LAURENS AND HENRY D. HOOKER, JR. that the stimulating influence of the white light, though of equal radiant energy power, does not, or cannot, influence the organ- isms. But in the regions of the spectrum of wave-lengths of less stimulating effect, the influence of the white light reveals itself, TABLE 2 The relative stimulating value of spectral lights of equal energy content as determined from the relative rate of locomotion SPECTRAL LIGHT ALONE SPECTRAL LIGHT VS. WHITE LIGHT Relative stimulating TT SN Sao aati WAVE- value value NUMBER| LENGTH | Rate of locomo- |——-—————————|_ Rate of locomo- et ate cantign Reciprocal eas eee Reciprocal Sates (|e | as) | eens (eae tion tion 1 434.0 0.86 0.116 76 1.02 0.098 64 2 444.0 0.82 0.122 80 0.92 0.109 72 3 454.0 0.78 0.128 84 0.79 0.127 84 4 464.0 0.74 0.135 89 0.77 0.130 86 5 474.0 0.72 0.1389 91 0.73 0.137 90 6 484.0 0.69 0.145 95 0.69 0.145 95 7 494.0 0.66 0.152 100 0.66 0.152 | 100 8 504.0 0.71 0.141 93 0.69 0.145 95 9 514.0 0.75 0.133 88 0.75 0.133 88 10 524.0 0.79 0.127 84 0.80 0.125 82 il 534.0 0.85 0.118 78 0.89 0.112 74 12 544.0 1.02 0.098 65 Og 0.093 61 13 554.0 ileealat 0.090 59 1.22 0.081 53 14 564.0 1.17 0.085 56 1.41 0.071 47 15 574.5 1.28 0.078 51 1.70 0.059 39 16 584.5 1.36 0.074 49 17 594.5 ill 0.066 43 18 605.0 1.54 0.065 42 19 615.0 1.60 0.063 41 20 625.0 1.63 0.061 40 21 635.0 1.64 0.061 40 22 645.0 1.79 0.056 37 23 655.0 ESTAS) 0.056 37 becoming relatively stronger and stronger, so that orientation and locomotion toward the source of the spectral lights are pro- gressively made less precise and retarded, with the result that the ratio of the rate of locomotion to spectral light as compared with the rate to white light becomes greater and greater. The SENSIBILITY OF VOLVOX TO SPECTRAL LIGHTS 357 influence of the white light thus relatively increased until its effect was practically equal to that of \574muy (light no. 15), the colonies showing an almost equal tendency to proceed to one or the other source. When exposed to the influence of the white light and to that of wave-lengths greater than \574muy, the white light had an increasingly greater stimulating effect as the wave- lengths became longer so that the organisms moved toward the So es ae: mT 2 a “HA we W = so J c ? 70 v L fF co 5 5 50 c ® 40 U z F 30 ry a 20 ite) 420 440 460 480 500 520 540 560 580 GOO G20 G40 S6Gc Mp Fig. 2 The relative stimulating value of spectral lights of equal energy con- tent as determined from the relative rate of locomotion. x——x——x, spectral light alone; e - - -*---e, spectral light vs. white light. white light. No further record was taken of the relative rate of locomotion. The curves show that \494my is the wave-length of maximum stimulating value, as was found by the method of determining the presentation time, and that the relative stimulating effect decreases by degrees toward both ends of the spectrum. The intensity of the spectrum used, however, was so great that all of 398 HENRY LAURENS AND HENRY D. HOOKER, JR. the wave-lengths tested have a stimulating effect. With this high-intensity spectrum it will be interesting to continue the experiments into the longer and shorter waves. This it is planned to do as soon as practicable. As mentioned earlier, the reaction time was also taken in certain cases. That is the time that elapsed between the mo- ment of exposure of the organism to the light, and the first indi- cation of a locomotor response, or a swimming movement, to- ward the source of the light. The results of computing the ratio of the reaction time in the spectral light to the reaction time in the white light show that the most effective wave-lengths (\494my) are the same as ascertained from the determination of the pre- sentation time and of the rate of movement. The relative ef- fectiveness of the various lights, due to the incompleteness of the records, is, however, not clearly shown. We have used the time taken to traverse a certain fixed dis- tance as an index of the relative stimulating efficiency of lights of different wave-length. Although Volvox orients fairly precisely, the colonies deflect somewhat, traveling in a wave-like course, up and down or from side to side. This deflection is less when the colonies are swimming toward the source of the lights of greater stimulating effect. In other words, the degree of de- flection varies with the stimulating efficiency of the light. We did not note any consistent change in the precision of orientation or of the rate of movement in the five tests given each colony. The relative time taken to swim a certain distance was not, however, due merely to the amount of deflection or precision of orientation, but also to the actual rate (strength) of swimming. It was easy to see in the lights of low efficiency that the colonies loafed along, now and then speeding up a little to slow down- again, and so on. These variations in the energy output, or the speed of swimming movements, being irrespective of the distance which they had traversed or as to how near they were to their goal. The conclusion seems justified, therefore, that the rate of locomotion, as well as the precision of orientation, is dependent upon the relative stimulating efficiency (or physio- logical intensity) of the lights, those of greater efficiency causing SENSIBILITY OF VOLVOX TO SPECTRAL LIGHTS 359 -a greater chemical change or transformation resulting in a greater expenditure of mechanical energy. Holmes (’03, p. 323) described the swimming of Volvox toward a source of light as follows: It was found that, as the Volvox traveled toward the light, their movement was at first slow, their orientation not precise, and their course crooked. Gradually their path became straighter, the orienta- tion to the light rays more exact and their speed more rapid. After traveling over a few spaces, however, their speed became remarkably uniform until the end of the trough was reached. If the light is so ‘intense that one end of the trough is above the optimum intensity of illumination, the speed of the Volvox is decreased as it approaches this optimum where it finally stops. In our experiments the distance to be traversed was only 26 mm. We observed that the rate of locomotion became uniform, as Holmes describes, but the organisms always went the whole distance, so that the effect of intensity on the rate of movement which he described is not in evidence. We have, however, often observed that the rate of movement can be from the first rel- atively very rapid, one might say, explosive, as if the colony had been catapulted or had bounded away in the direction of the source of lights. The impression was gained that in many cases the velocity of movement slackened a little after the first out- burst of speed, and that the colony settled down to a uniform rate which was held until the end of the trough was reached. This type of movement was only to be observed in the reactions to the lights of greatest stimulating efficiency. Its rate was not measured, and we are not sure that it can be measured, although it is possible that by dividing the distance by transverse lines it may be done. It was also observed with great clearness in the experiments on the determination of the presentation time. In these experiments, particularly in the responses to the most ef- ficient lights, the Volvox would often continue to the end of the trough, although, as above mentioned, the exposure lasted for a very short time and was followed by the latent period. In many cases, however, the colony would only proceed a short distance. This sometimes happened even in those cases where the first movement was of the explosive nature described, the colony stop- 360 HENRY LAURENS AND HENRY D. HOOKER, JR. ping its forward movement and hanging apparently motionless or swimming slowly upward. The investigation of the question as to whether the rate of loco- motion, particularly in the lights of high stimulating effect, is uniform throughout the course or whether there are variations in the rate of locomotion which can be correlated with the dis- tance already gone, the length of exposure, the relative proximity to the source (intensity), etc., is one which is reserved for future investigation. Some interesting data may thus be obtained, e.g., from the comparison between the average rate for the whole’ distance, or period, with the rates at different parts. In other words, the determination as to whether the rate is constant for the whole period or variable, and thus whether the curve ex- pressing the average rate is a straight line or a curve. DISCUSSION There is no doubt that light stimulates nerve-endings through a photochemical reaction, the stimulation being mediated by photoreceptors, not necessarily structurally defined, but sensi- tive to photochemical change in the substance with which they are in contact. The effect of light is therefore due not to light directly, but to the chemical changes which it causes, these changes involving the formation of a substance or of substances which, according to mass action and reaction velocity, act as ‘mner stimuli.’ Although the effect of light is due to the mass action of the chemical compounds which it produces (the photochemical re- action product), it is not assumed that the rates of the photochem- ical reactions themselves follow the simple law of mass action, since the rate is controlled by the amount of the light energy absorbed per unit time, and not by the actual number of molecules present (Sheppard, pp. 211, 217 ff). The photochemical process is fundamentally an electric one, in that there is a raising of potential, due to ionization, as the immediate effect of light (electrolytic dissociation is governed by the law of mass action). An increase in permeability, resulting SENSIBILITY OF VOLVOX TO SPECTRAL LIGHTS 361 from stimulation, does away with the polarized condition, by allowing the two layers of ions to mix freely, and the condition of excitation spreads or is propagated. We prefer to leave to later development and consideration the exact relations between ionization and permeability (‘cause and effect’), and at this time merely call attention to a few mat- ters of rather general nature and import, as they bear upon our subsequent work. The study of the relative stimulating effects of lights of dif- ferent wave-lengths on organisms will undoubtedly give informa- tion concerning the nature of the photochemical changes in- volved in light reactions. Any particular chemical reaction is » produced by a certain group of wave-lengths only, so that the possibility is presented of distinguishing between light of differ- ent wave-lengths. A study of the stimulating effect of various wave-lengths on organisms, combined with radiometric and spectrophotometric examination, showing the degree of ‘energy’ as well as ‘light’ absorption of solutions or extracts of the organ- — isms, and of the ‘visual purple’ of some of the invertebrates will - be of interest and value. Although there is a wave-length which is most effective in action on a given photochemical substance, the same wave- length may be maximally effective for more than one substance. Therefore, the fact that the same wave-lengths may be found equally effective for two species of organisms does not per se signify that the photochemical substances or reactions are for that reason the same. This invalidates the value of comparisons which are made, on this basis, between the photochemical sub- stance in a certain organism and the visual purple in the amphib- ian, mammalian, or human eye. The fact that a certain organ- ism shows a maximum sensibility to certain wave-lengths which happen to be the same as those which cause most rapid bleaching of visual purple, and which are maximally absorbed by it may be merely fortuitous, and have no fundamental significance. How otherwise can it be explained that in certain organisms which are nearly related the wave-lengths of light of maximal stimulating efficiency are quite different, while in many that are 362 HENRY LAURENS AND HENRY D. HOOKER, JR. only distantly, if at all, related, the wave-lengths of maximal stimulating efficiency are practically the same? This does not mean that we cannot compare the effects of lights of different wave-lengths on photosensitive protoplasm with the effects on the photosensitive substance in the eye of man, as indicated by their respective relative stimulating values. In this connection attention is called to the objections stated by Loeb and Wasteneys (’16, p. 224; Loeb, 718, p. 102) against the drawing of conclusions from comparisons made between the sensations of brightness in color-blind human beings and the wave-lengths of maximal stimulating efficiency for lower organ- isms to the effect that these lower organisms are therefore also ‘color-blind.’ The two photosensitive substances, visual purple and the photosensitive substance in the organism in question, are merely affected in a similar way by the same wave-lengths. We do not wish to seem hypercritical, but feel that attention should also be called to the probability that the determinations of the wave-lengths of maximal efficiency in bleaching visual purple, and of those maximally absorbed by it, will be found to — be erroneous. The work of Trendelenburg (’04) is far from per- fect, particularly in the application of the energy corrections and in the determination of the region of maximum absorption (Garten, ’06, ’08), as indicated by the shifting of the relative absorption values. Furthermore, Henri (711) has ascertained the threshold energy, the bleaching effect on visual purple, and the amount of light absorbed by it at various wave-lengths with due consideration to the distribution of energy in the spectrum. These three factors all follow the same course with maxima be- tween \520mu and \}500mu. Also Bender (14, ’16) has recently found that the luminosity curve for the totally color-blind eye and for the peripheral field of the normal, dark-adapted eye. has a maximum between \515muyu and 520mu. Perhaps importance may be attached to the fact that the chemical processes in the retina are assumed to be pseudorevers- ible reactions, while there is as yet no positive evidence that any other reactions involving photosensitive protoplasm are of this type. Furthermore, visual purple is an optical sensitizer, al- SENSIBILITY OF VOLVOX TO SPECTRAL LIGHTS 363 though we do not know whether the products of its change them- selves stimulate the receptors or whether they act catalytically. It seems to us justifiable to conclude that the influence of dif- ferent wave-lengths, as far as the chemical reactions which are associated with the action of light are concerned, are funda- mentally the same for animals and plants. ‘This is all that we understand that Loeb claims when he speaks of the ‘‘identity of heliotropism in plants and animals,” viz., that the reactions of plants and animals are both due to the action of light on the photosenstitive substances, resulting in transformation. For dif- ferent organisms (plants and animals) there are different wave- lengths which have a maximum of stimulating effect. This sig- nifies that there is a difference in the photosensitive substances, photochemistry telling us that the most efficient wave-length varies with the nature of the photochemical substance. We find it difficult to understand just what Mast (717, p. 522) means when he says, ‘‘the reactions are not wholly dependent upon wave-lengths, for while there is clearly a region of maximum stimulating efficiency in the spectrum, stimulation is not confined to this region and the stimulating effect of the wave-lengths on either side of it can be made greater by simply increasing their intensity.” It is found in plants and animals that certain wave-lengths have a maximal stimulating effect as compared with the effect of other wave-lengths. It stands to reason, of course, that this effect is due to the energy of radiation (i.e., light and heat effects, which together make up the stimulating value of the wave- lengths in question). For different wave-lengths to produce the same effect would require that the amount of energy absorbed by the photochemical substance would be the same in all cases (that is, to have the same relative penetrating power). There- fore it follows that by increasing the intensity of any wave- length we would increase its stimulating effect. It is thus pos- sible that if the entire series of wave-lengths be made increasingly intense or greater in radiant power beyond a certain limit, a considerable portion of the spectrum in the neighborhood of the wave-lengths of maximum stimulating efficiency at a lower in- 364 HENRY LAURENS AND HENRY D. HOOKER, JR. tensity would be found to be equally efficient as stimuli, because the absolute maximum effect of the wave-lengths of greatest stimulating effect has been reached at a lower intensity of the spectrum, while that of the other wave-lengths was below their possible maximal effect, or because of the influence of some lim- iting factor. It is probably also true that if the absolute intensity of an equal-energy spectrum be decreased, certain wave-lengths, which at a greater intensity have a relatively weak stimulating efficiency, will, as the intensity is decreased, have finally none at all, and this will spread and thus involve more and more of the spectrum. But the wave-lengths of maximum effect will remain the same for all intensities. This is of course pure assumption. We know of no work on the influence of the relative intensity of the spectrum on the ~ location of the wave-lengths of maximum stimulating efficiency for other photosensitive protoplasm than the human eye. But a low-intensity spectrum has for the human eye a region of max- imum stimulating efficiency (achromatic scotopic luminosity curve) nearer the blue end of the spectrum than a high-intensity spectrum (photopic luminosity curve), and its luminosity curve is similar to that of the totally color-blind under all intensities (Parsons, pp. 189, 209), as well as to that of rod vision (peripheral vision) under high illumination (according to Bender, 714, figs. 2 and 4), the curve for the peripheral retina coinciding with that of the foveal visibility curve of totally color-blind persons, with maxima at \515myu. (The question of the quantitative and quali- tative differences between peripheral vision and scotopia on the one hand and central vision on the other, either need revision or reinvestigation. Parsons, pp. 71-72.) Now it is a question, as above indicated, whether anything comparable can be found in the sensibility to different wave- lengths in lower organisms. Can we, by varying the absolute intensity, keeping the spectrum equal in energy throughout, shift the region of maximum stimulating efficiency? Probably not. But it is a question well worth considering and investigating. If the luminosity curve for the peripheral retina (normal rod SENSIBILITY OF VOLVOX TO SPECTRAL LIGHTS 365 vision) coincides with the curve of visibility for totally color- blind persons (isolated rod vision), this latter being the same as the achromatic scotopic curve, we conclude that the rods are always stimulated in the same relative degree. High and low intensities produce the same relative effects as to wave-lengths. Applied to lower organisms, we conclude that there is nothing in the sensibility to light of different wave-lengths which would lead us to assume anything comparable to the photosensitive substance in the cones, and that no matter with what intensity of the spectrum we work, we will always obtain the same rela- tive sensibility wave-length curve. If a substance is sensitive to light of a particular wave-length, it must absorb this light and show an absorption band in the region in question, the absorption spectrum of a chemical system - being intimately connected with its photochemical behavior. Light waves are absorbed in ponderable media by particles capable of a free period of vibration. Vibrations not synchro- nized to these produce only forced vibrations of the particles and ~would hence be only slightly absorbed. In a heterogeneous mixture, if the substances do not interact to form a new combination, the light absorption of a mixed solu- tion will be equal to the sums of the absorptions which the com- ponents would exert separately. That is, when chemical inter- actions are excluded, the behavior of summed absorptions are purely additive. The photochemical substances in organisms are probably het- erogeneous, comprising a number of different substances, so that a number of different groups of wave-lengths are absorbed, pro- ducing, by resonant vibration of different rates, chemical re- actions, resulting in photochemical products. That is, in the formation of particular photochemical products, vibrations (mo- lecular, atomic, or electronic) of a certain rate are excited by resonance, with excitation as the result of the energy set free. For recent applications of this chemical and physical conception involving absorption, vibration and resonance in theories of color vision, see the papers by Houstoun, ’16; Guild, ’18, and Troland, *167and 717: 366 HENRY LAURENS AND HENRY D. HOOKER, JR. It does not seem possible at the present time to assign a satis-- factory explanation to the maximal stimulating effect of light of a particular wave-length. The following considerations are, however, of interest. The effect of the light is a result of its absorption by the photosensitive substances. Only the light which is absorbed is chemically active, though all of the rays absorbed are not necessarily active in producing chemical change. When light is absorbed, the amplitude of the vibrations is a maximum when the free periods of the vibrating particles co- incide with the period of the incident light. The wave-lengths of maximum stimulating efficiency owe their action to maximum absorption and to the hypersensitivity of the photosensitive substance to the influence of these particular rays. But absorption alone cannot be used as a measure of physio- logical action, because it depends upon the kind of processes initiated by the transformation of the absorbed energy (Bovie, 18, p. 253). The kind of processes initiated are dependent upon the nature of the original photochemical substance. SUMMARY 1. The sensibility of Volvox to radiation of different wave- lengths but of equal energy (sensibility to radiation at equal energy) has been investigated by two methods: (a) the relative duration of the presentation time, and (b) the relative rate of locomotion (and precision of orientation). 2. Wave-length \494mu is found by both methods to have the highest stimulating value. The efficiency of the other wave- lengths show a gradual decrease as shorter and longer wave- lengths are tested. 3. The necessity of using an equal-energy spectrum for such work is emphasized. 4. Attention is directed to the present inability of making comparisons with human vision (either normal or color-blind), that is, with photopic and scotopic (or achromatic) luminosity curves: (a) because of the lack of sufficiently exact work on lower organisms; (6) because of our lack of knowledge regarding photo- SENSIBILITY OF VOLVOX TO SPECTRAL LIGHTS 367 chemical reactions and the nature of the photosensitive sub- stances, and, (c) because recent work on the visibility of radiation for the normal and color-blind eye shows that the maximum effect is not at the wave-lengths indicated by earlier work. BIBLIOGRAPHY Bayutiss, W. M. 1918 Principles of general physiology, 2nd edition. Long- mans, Green & Co. Benper, H. 1914 Untersuchungen am Lummer-Pringsheimschen Spectral- flickerphotometer. Ann. d. Physik, Bd. 45, S. 105-132 (see also 1916. Zeitschr. f. Psychol. u. Physiol. d. Sinnesorgane, II. Abt., Bd. 50, S. 1-41). Bovis, W. T. 1918a An approximation of the value of the absorption index of fluorite rays in protoplasm. Jrl. Med. Research, vol. 39, pp. 239-249. 1918 b The localization of the physiological effects of radiation within the cell. Jrl. Med. Research, vol. 39, pp. 251-265. 1918 c The physiological action of radiation. Jrl. Med. Research, vol. 39, 271-277. CoBLentTz, W. W., AND Emerson, W. B. 1917 The relative sensibility of the average eye to different colors and some practical applications to radia- tion problems. Bull. Bur. Standards, Scientific paper no. 303, pp. 167-236. Fréscuet, P. 1910 Uber allgemeine, im Tier- und Pflanzenreich geltende Gesetze der Reizphysiologie. Zeitschr. f. allg. Physiol., Bd. 11, S. 43-65. Garren, 8. 1906 Uber die Verinderungen des Sehpurpurs durch Licht. Arch. f. Ophthal., Bd. 63, S. 112-187. 1908 Verinderungen vorgebildeter Farbstoffe durch Licht, insbeson- derer Bleichung des Sehpurpurs. Graefe-Saemisch Handb. d. Ges. Augenheilk., 2te Aufl., S. 130-212. Guitp, J. 1917* The mechanism of colour vision. Proceed. of the Physical Soc. of London, vol. 29. pp. 354-361. Hecut,S. 1918 The photic sensitivity of Ciona intestinalis. Jrl. gen. Physiol., vol. 1, pp. 147-166. 1919 a Sensory equilibrium and dark adaptation in Mya arenaria. Jrl. gen. Physiol., vol. 1, pp. 545-558. 1919 b The nature of the latent period in the photic response of Mya arenaria. Jrl. gen. Physiol., vol. 1, pp. 657-666. 1919 c¢ The effect of temperature on the latent period in the photic response of Mya arenaria. Jrl. gen. Physiol., vol. 1, pp. 667-685. Hennri,‘V., er Des Bancets, J. L. 1911 Photochimie de la rétine. Jrl. Physiol. et de Pathol. gén., T. 13, p. 841-856. Houmgs, 8. J. 1903 Phototaxis in Volvox. Biol. Bull., vol. 4, pp. 319-326. Hovstoun, R. A. 1916 A theory of colour-vision. Proc. Roy. Soc., vol. 92 A, pp. 424482. THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 30, NO. 3 368 HENRY LAURENS AND HENRY D. HOOKER, JR. Hype, E. P., Forsyrue, W. E., anp Capy, F. E. 1918 The visibility of radia- tion. Astrophysical Jrl., vol. 48, pp. 65-88. Ives, H. E. 1912 The spectral luminosity curve of the average eye. Phil. Mag., 8S. 6, vol. 24, pp. 853-863. Laurens, H., anp Hooker, H. D., Jr. 1917 Studies on the relative physio- logical value of spectral lights. I. Apparatus. Amer. Jr]. Physiol., vol. 44, pp. 504-516. 1918 The relative sensitivity of Volvox to spectral lights of equal radiant energy content. Anat. Rec., vol. 14, pp. 97-98. Logs, J.. AND WasTENEYsS, H. 1916 The relative efficiency of various parts of the spectrum for the heliotropic reactions of animals and plants. Jour. Exp. Zool., vol. 20, pp. 217-236. Lors, J. 1918 Forced movements, tropisms and animal conduct. Lippincott. LucxigesH, N. 1917 The physical basis of color-technology. Jrl. Franklin Institute, vol. 184, pp. 73-94, 227-250. Mast, 8. O. 1907 Light reactions in lower organisms. II. Volvox globator. Jrl. Comp. Neur. and Psychol., vol. 17, pp. 99-180. 1911 Light and the behavior of organisms. John Wiley & Sons. 1917 ~The relation between spectral color and stimulation in the lower organisms. Jour. Exp. Zool., vol. 22, pp. 471-528. Nourtina, P.G. 1908 The luminous equivalent of radiation. Bull. Bur. Stand- ards, vol. 5, pp. 261-308. 1915 The visibility of radiation. Phil. Mag., 8.6, vol. 29, pp. 301-309. Parsons, J. H. 1915 An introduction to the study of colour vision. Putnam’s Sons. Reeves, “P. 1918 The visibility of radiation. Trans. Illum. Engineer Soc., vol. 13, pp. 101-109. SuHepparp, 8. E. 1914 Photo-chemistry. Longmans. TuHiRMEL, E. 1910 Das Lummer-Pringsheimschen Spectralflickerphotometer als optisches Pyrometer. Ann. d. Physik., Bd. 33, S. 1139-1160. TRENDELENBURG, W. 1904 Quantitative Untersuchungen iiber die Bleichung des Sehpurpurs in monochromatischem Licht. Zeitschr. f. Physiol. d. Sinnesorg., Bd. 37, S. 1-55. Trotanp, L. T. 1916 Apparent brightness: its conditions and properties. Trans. Illum. Eng. Soc., vol. 11, pp. 947-966. 1917 The nature of the visual receptor process. Jrl. Optical Soc. of America, vol. 1, pp. 3-15. wo ih a7 st) i “ital Lat ies hed ae A Magy pt aL ’ + A MT ‘adh, Tt at ah ia” | i Resumen por el autor, F. B. Sumner. Institucién Scripps. La variacién geografica y la herencia mendeliana. En ocho lotes de individuos del rat6én Peromyscus maniculatus, procedentes de diversas partes de California, he encontrado el autor diferencias locales significativas en lo referente a la longi- tud media de la cola, pié, oreja, pelvis, fémur y crdneo, anchura de la raya caudal dorsal, color del pelaje, pigmentacién del pié y numero de vértebras caudales. Animales pertenecientes a lo que ordinariamente se considera como la misma subespecie, pue- den diferir considerablemente en diferentes localidades. Para un cardcter determinado el ntiimero de variaciones no coincide con el de lotes. En determinados caracteres y en cierto grado, puede observarse una variacién gradual geografica y climatolog- ica en estas variedades locales. El orden de gradacién coincide bastante bien con ciertos caracteres, aun cuando esto no puede aplicarse a todos ellos. Los coeficientes de correlacién indican que los caracteres que varian juntos, en sucesién geogrdafica, ‘pueden variar o no variar juntos en un lote local y vice-versa. De este modo, circunstancias especiales que actuian de un modo local deben causar la modificacién simultdanea de partes que ordinariamente no varian juntas. Las diferencias entre estas razas locales no se comportan en los cruzamientos como simples factores mendelianos, aunque la teoria de los “‘factores miulti- ples”’ se invocaria indudablemente en este caso por muchos gen- eticistas. El autor ha comparado las desviaciones tipos de va- rios caracteres en las generaciones F; y F;. En la mayor parte de los casos el ligero aumento en la variabilidad se presenta en la generacién F;, aunque en una considerable minoria los num- eros obtenidos son préximamente los mismos en las do: gener- aciones, 0 las condiciones pueden aun invertirse. Los experi- mentos mencionados arrojan incidentalmente luz sobre algunas diferencias sexuales secundarias inesperadas. Translation by José F. Nonidez Carnegie Institution of Washington AUTHOR’S ABSTRACT OF THIS PAPER ISSUED BY THE BIBLIOGRAPHIC SERVICE, MARCH 8 GEOGRAPHIC VARIATION AND MENDELIAN INHERITANCE FRANCIS B. SUMNER SEVEN FIGURES INTRODUCTION AND SUMMARY OF RESULTS In several previous papers! I have reported upon the results of a biometric and genetic study of the geographic races of a single species of deer-mouse (Peromyscus maniculatus) found within the State of California. The present paper, like the former ones, is merely a report of progress. I shall here discuss the more important additions to my previous findings, based upon the work of the last two years. I shall, however, make only incidental reference to certain prominent phases of this work, such as the study of the various color mutations which have appeared and the field studies of environmental conditions. These will be embodied in separate papers, to be published before long by Mr. H. H. Collins and myself. Thus far, collections sufficient for statistical purposes have been made in eight different localities and minor collections at five or six others. Over three thousand animals, wild and cage- born, have been subjected to measurement. Only wild speci- mens, unmodified by captivity, have been dealt with in com- puting the differences between the various races here discussed. The characters chosen for study have been, as far as possible, ones which are capable of quantitative expression. As empha- sized in earlier papers, the methods employed have been stand- ardized, so as largely to eliminate differences due to irrelevant circumstances and to ‘personal equation.’ What these charac- ters are will appear in the ensuing pages. Most of them have 1 American Naturalist, November, 1915; 7d., March, 1917; Genetics, May, 1917; Bulletin of the Scripps Institution, No. 3, October 19, 1917; American Naturalist, April-May, June-July, August-September, 1918. 369 370 FRANCIS B. SUMNER been dealt with in previous reports. A further character of great importance, which I plan to study quantitatively, is the color of the pelage, but no data from this source will be considered in the present paper.? I shall not regard it as necessary here to discuss either the technique followed in making the measurements or the methods employed in the subsequent statistical treatment. Some of my former papers have covered this ground sufficiently well for present purposes. To reverse the conventional procedure, I propose here to sum- marize and discuss my results in advance of the more detailed account. This will serve to make clear at. the outset the purpose of the paper, and will perhaps render the detailed statements more intelligible to such readers as may stray beyond this introduction. To proceed with this general discussion, significant racial dif- ferences have been found in respect to the mean length of the tail, foot, ear, pelvis, femur, and skull, the width of the dorsal tail stripe, the color of the pelage, depth of pigmentation of the foot, and number of vertebrae in the tail. There are not, how- ever, for any one character, as many grades to be distinguished as there are localities from which collections have been made. For example, I have thus far found only six distinguishable grades in respect to tail length and only four or five in respect to foot length. Local collections from different parts of the ranges of the same ‘subspecies,’ as thus far recognized by systematists, have been found to differ considerably. Thus, the two extremes, in re- spect to ear length, are presented by the Berkeley and the La Jolla collections, both of which are generally assigned* to the subspecies ‘gambeli,’ while the subspecies ‘rubidus,’ as regards several characters, presents at least three well-marked grada- 2 Mr. H. H. Collins is preparing a paper which includes color analyses made in accordance with a method which we have developed jointly. 3 Stephens (California Mammals, San Diego, 1906); Osgood (Revision of Mice of the American genus Peromyscus, Bureau of Biological Survey, Washington, 1909); Grinnell (Distributional List of the Mammals of California, Proceedings California Academy of Sciences, August 28, 1913). VARIATION AND MENDELIAN INHERITANCE 3¢1 tions, as we pass southward from Humboldt to Sonoma County. In fact, the collection from Duncan Mills, near the mouth of the Russian River, is about intermediate between ‘gambeli’ and the more extreme ‘rubidus’ type of Humboldt County. Neverthe- less, it is a significant and striking fact that the ‘gambeli’ mice from the vicinity of Calistoga, in Napa County, much more closely resemble those from La Jolla, 500 miles distant, than they do the ‘rubidus’ from Duncan Mills, only twenty-seven miles away. To a certain extent, and for certain characters, the gradations here considered follow a geographic—and also a climatic—se- quence. ‘The degrees of difference in the characters are, however, by no means proportional to the geographic intervals between the ‘races,’ and there are other incongruities which greatly com- plicate the situation. Thus, while there is strong evidence for some sort of causal relation between these local differences in the mice and some element or elements in the physical environment, it is not possible to state with confidence at present just what these elements are nor how they operate to modify the organism. The order of gradation for certain of the characters agrees with that for certain others. Thus the order for tail length, foot length, and tail stripe is similar, though not identical, while the order for general color and that for skull length each presents some significant points of resemblance to that for the other characters just named. , Certain characters, on the other hand, present a totally dif- ferent arrangement. Thus ear length and the length of the pelvis (innominate bone) offer gradations which bear no relation to one another or to those of the other characters. Again, it must be emphasized that there are well-marked differences even among those characters which approach one another most nearly in their order of arrangement. In respect to tail length, for example, the Berkeley and Victorville races do not present sig- nificant differences, while in regard to the width of the tail stripe they stand far removed. Coefficients have been computed, showing the extent of the correlation among certain of these characters within each single race. This, needless to say, is another way of ascertaining Saf FRANCIS B. SUMNER whether or not the respective characters tend to vary together. Of course, the size of all the parts of the body is correlated with the general size of the animal. Thus, larger animals have longer tails, feet, etc. Also, in a series of mixed size, any parts which are correlated with body size are necessarily correlated with one another. But this is not the sort of thing which concerns us here. Of chief interest are the correlations which are found to exist between two characters, when the element of body size has been eliminated. For this purpose, we may either parcel out our animals into groups having approximately the same body length, and determine our coefficients within each of these, or we may employ material of varying size and resort to the method of ‘multiple correlation.’ For certain reasons the former method has been employed exclusively, at least for such characters as vary with the size of the body.‘ Thus proceeding, I find that there is a small, though probably significant, positive correlation between the length of the tail and that of the foot and skull, and a less certain one between tail and ear, while the length of the foot likewise appears to be positively correlated with that of the pelvis. The tail-to-body ratio is negatively correlated with body length; i.e., longer mice have relatively slighter shorter tails. On the other hand, there is probably no correlation between body length and the relative width of the dorsal tail stripe (ratio of are to entire circumference of tail). The number of tail vertebrae is positively correlated with the relative, but not with the absolute, length of the tail. In other words, animals, large or small, which have relatively long tails possess, on the average, a slightly greater number of vertebrae. But larger animals have no more vertebrae than smaller ones, despite the greater absolute length of the tail. In any case, the length of the tail is chiefly determined by the size of the indi- vidual vertebrae, rather than by their number. 4 Certain other characters, such as the relative length of the tail (regarding it as a percentage of body length), the relative width of the tail stripe (expressed as a percentage of the circumference of the tail), and the number of vertebrae are so slightly correlated with the general size of the body that correlations have been computed directly in populations of mixed size. VARIATION AND MENDELIAN INHERITANCE 373 The width of the tail stripe (relative, as defined above) appears to be not at all correlated with relative tail length. The mean coefficient actually obtained, which is probably not significant, is slightly negative. This is surprising when we consider that the two characters in question follow much the same order in their degree of manifestation among the various geographic races, and that when these local collections are thrown together and treated as a single population, a fairly high positive corre- lation is found to obtain.® Thus, characters which vary together, when geographic se- quence is considered, may or may not vary together within any single local collection, while, conversely, characters which are correlated within these various local populations may or may not be found to have undergone concomitant modification, when we pass from one locality to another. These relations raise the question whether the interracial dif- ferences in the mean values of various characters belong to the same type as the intraracial or individual differences. By be- longing to the same type I mean having the same sensible prop- erties, and the same mode of hereditary transmission. I think that the ensuing discussion will make clear that the sensible prop- erties are the same, so far as inspection reveals, and that the behavior in heredity is probably likewise the same in the two cases. As to their respective causes, on the other hand, we know too little at present regarding the causes of variation in general to draw any very useful distinctions upon that basis. It would seem obvious, however, that special factors, operating locally, must be responsible for the simultaneous modification of parts 5 A rather obvious explanation of this apparent contradiction suggests itself here, which, however, I am certain is not the correct one. It might be supposed that an actual positive correlation between these two characters exists within each of the local populations, but that this is too feeble to be appreciable, owing to the limited variability of these populations considered separately. It need only be pointed out that the standard deviations of the local collections, taken singly, are more than half (55 to 80 per cent) as great as those of the mixed assemblage which results when the data are combined. (This has been done for four races only.) Significant positive coefficients of correlation might, therefore, reason- ably be expected within each local race, so far as the extent of variability is con- cerned. 374 FRANCIS B. SUMNER which do not ordinarily vary together. I shall return to this point presently. It must be borne in mind that all of the differences which I have dealt with between the local races relate to the average condition, and do not hold constantly for every individual of the groups under comparison. In fact, a large proportion of the individuals belonging to two adjacent groups might be placed indifferently in either, without the transfer being detectable by any test known to me. It is only the most widely separated collections, e.g., those from Humboldt Bay and the Mohave Desert, which differ so much in respect to certain characters that the frequency poly- gons for these last do not overlap more or less broadly. Between such a condition of distinctness as this and one of practically complete identity we have many gradations. In some cases it is only by the comparison of probable errors that we are enabled to say whether or not two collections differ significantly in respect to a given character. It must be remembered, too, that the local ‘races’ with which I am dealing are highly artificial groups. My collections are simply samples, taken at various arbitrarily chosen points from a perfectly continuous population. Whether or not these local differences in the average or modal condition would be completely bridged by collections taken at stations sufficiently close to one another remains problematic. It would seem almost inevitable, however, that interbreeding would lead to such a complete con- tinuity, at least in the absence of some sort of geographic bar- riers. For mice of even the most widely separated of these races appear to be fully fertile inter se. It appeared early in the course of these studies that the various racial differences were hereditary. The races ‘bred true,’ so far as could be detected by the methods employed and allowing for certain abnormalities of form to which all of the races were sub- ject when reared in captivity. I have also shown that the variations within each local race— or some, at least, among them—are rather strongly hereditary. As stated in an earlier paper, the parental-filial correlation in _ respect to tail stripe and to relative tail length averages, in each ~ VARIATION AND MENDELIAN INHERITANCE BD case, about 0.3. This means, as ordinarily interpreted, that in each local race part of the variability is hereditary and part non- hereditary, this last component being regarded as ‘somatic’ in origin. One essential feature of these geographic races remains, how- ever, to be examined somewhat further. It has been shown that the shifting of mode by which one ‘race’ arose historically from another must have involved the simultaneous shifting of a con- siderable number of different modes. And this occurred even among characters which, in their every-day inheritance, do not seem to be linked together to any appreciable extent. More- over, characters (e.g., foot and pelvis), which appear to be cor- related positively in the individual, appear in some cases to have been modified in opposite directions in the course of phylogeny. I have pointed out in earlier papers that in respect to both coat color and the width of the dorsal tail stripe a general cli- matic sequence is discernible among these races, and this con- clusion appears to be borne out, on the whole, by the additional data presented below. I have also called attention to the agree- ment between my own findings in this regard and those of vari- ous mammalogists and ornithologists, who have recognized the existence of an increase in pigmentation pari passu with an in- crease in the atmospheric humidity of their habitat. If we con- sider only the coastal stations from San Francisco Bay north- ward (Berkeley, Duncan Mills, Fort Bragg, Eureka), which probably present a graded series in respect to both temperature and atmospheric humidity, we find likewise a similar gradation in respect to the mean width of the tail stripe and the mean length of the tail, foot, and ear. The suggestion lies close at hand that we have to do with some more or less direct influence of environ- ment, which, in the course of time, has modified the hereditary characters of the animals dwelling at these various points.’ Perhaps the four characters just named have undergone simul- taneous modification by some single external agency, and this 6 American Naturalist, June-July, 1918, p. 294. 7 This supposition might, of course, be expressed in such terms as would ex- clude the .‘inheritance of acquired characters.’ 376 FRANCIS B. SUMNER might be held to account for the parallel modification of parts which ordinarily do not vary together. Now, I regard it as highly probable that such racial differ- ences as relate to pigmentation have been produced in some way by environmental agencies. And our problem would doubtless be simplified if we could regard the other differences named as hay- ing arisen simultaneously through this same set of environmental factors. Against this supposition, however, is to be set the fact that these various characters are not always modified in the same direction throughout the entire range of the species. For example, the desert mice have narrower tail stripes and less pig- mentation generally than those of Berkeley, whereas the mean tail length is almost identical in the two races. Of course, it is possible to rejoin that the effects of environmental influences in any given case are probably very complex, and that while one set of conditions might call forth parallel modifications in two differ- ent characters, another set of conditions might call forth diver- gent ones. In making such assumptions, we should, of course, be venturing upon very uncertain ground, but I cannot conceive of an explanation of the curious relations here considered, with- out some sort of appeal to local (i. e. environmental) factors. I have already stated that when we throw together certain of the local collections and treat them as a single population, a decided positive correlation appears between two characters (tail length and tail stripe)® which were not correlated within the local collections taken separately. This, indeed, is a mathematical necessity when two groups of individuals, differing in the mean values of two characters, are mixed together. The characters in question are inevitably found to be correlated in the mixed popu- lation. It might seem, at first, to be equally inevitable that the gradual migration and dispersal of these mice would bring about a similar intermixture, resulting in a measurable correlation be- tween two such characters. That this has not actually resulted is doubtless owing to the slowness of the process of dispersal. If these animals were continually traveling great distances and ? This would doubtless be true of certain other pairs of characters, for which my data are not yet complete. ; VARIATION AND MENDELIAN INHERITANCE 377 in large numbers, there would, without question, occur such a mingling of types as would suffice to bring about this result. The case would be indistinguishable from that of our combining in the same table the measurements of individuals from different local collections. But it does not seem at all probable that these animals indulge in such extensive migrations within their com- paratively brief lives. It would probably take many generations of mice before the descendants of any particular local strain would reach a point only twenty miles distant, and this would be doubly true if any form of geographic or ecologic barrier intervened. Now it is important to point out that such a gradual inter- mixture of two local races, even if complete, would not result in bringing about the correlation of two characters which did not previously tend to vary together.® For, as will be shown in detail later, there seems to be no tendency in hybridization for these various racial characters to be transmitted together. The ‘rubidus’ mice from Humboldt County have both a consider- ably longer tail and considerably wider tail stripe than the ‘gambeli’ mice from Calistoga. But neither in the F, nor the F, generation of hybrids do we find any more evidence of cor- relation between these characters than we find in the pure races, considered separately. The same independence in trans- mission is probably true in respect to tail stripe and foot pigmentation. Thus it would seem that the racial complex of characters is permanent only so long as mice of the same ‘race’ breed toge ‘her, as happens of necessity in nature. There is no linkage among these characters in heredity, or at least this is true of some of the most distinctive ones. Such a condition of independent transmission is, of course, the familiar one in Mendelian inheritance, where the various unit factors segregate, for the most part, without relation to one another. But here the resemblance ends. As will shortly be ® It would, of course, eliminate all local differences of type, unless the modify- ing agencies continued to operate at a rate sufficient to outweigh the process of diffusion, an assumption which we must make wherever local differences are encountered. 378 FRANCIS B. SUMNER shown, the racial differences under consideration do not depend upon single Mendelian factors. As to those cases in which two characters are positively cor- related in the individual (e.g., tail and foot), this may be due to some sort of “‘linkage”’ in inheritance, but may likewise be due to parallel modification by environmental agencies. For the white mouse I demonstrated many years ago that the length of both tail and foot was to a considerable extent dependent upon the temperature of the atmosphere in which the animals were reared, and in the case of Peromyscus I have more than once pointed out that cage-born animals tend to have both of these parts shortened. Such facts as these point to the possi- bility that the correlation of these parts in nature may be due to the varying incidence of external modifying agencies of some sort. I shall now pass to the second phase of these studies to be reported upon in the present paper—that, namely, which con- cerns itself with the crossing of different geographic races. It is significant that the word ‘genetics’ has, to an increasing extent, come to mean the experimental study of Mendelian unit fac- tors—real or alleged—as revealed by hybridization..°. And indeed to genetics, in this unwarrantably restricted sense of the word, we are ourselves giving considerable attention in our work with Peromyscus. We have followed the inheritance of several different color mutations, and obtained fairly typical mono- hybrid and dihybrid ratios, along with some cases which still puzzle us. Some of these results I have already published in preliminary form, and a further report will probably be made during the next few months by Mr. Collins and myself. My only quarrel is with the contention that all inheritance is Mendelian, whether it seems so or not, and with the endless creation of hypothetical ‘unit factors,’ to explain every de- parture from the expected manner of transmission. In this latter class I include the so-called ‘multiple factor’ explanation of blended inheritance and of the modification of characters through selection. 10 T recall finding this meaning explicitly given to the word in a recent work. VARIATION AND MENDELIAN INHERITANCE 379 As is well known, the exponents of this latter hypothesis lay stress upon those undoubted cases in which the second hybrid generation, while showing an intermediate condition, like the first, nevertheless displays a higher variability than the latter. The necessity for such an increase in variability, as a result of segregation, is obvious where a single pair of unit factors is con- cerned, as in ordinary Mendelian inheritance. That it would be equally necessary if a given character difference were con- ditioned by two or more pairs of independently segregating factors may readily be proved." Such a general increase in variability, in later hybrid genera- tions, was, it is worth while noting, well recognized by the early hybridists before the work of Mendel became known. ‘To one who is not committed to the doctrine of the immutability of the ‘gene,’ such an increased variability is intelligible upon the assumption of a tendency for the parental contributions to seg- regate from one another during the formation of the germ cells. This tendency may be completely realized, as in the case of typi- cal Mendelian inheritance. It may be overcome, wholly or partially, by a tendency toward fusion, in those numerous cases in which we have a permanent blending of characters, whether or not an increased variability is shown in later hybrid gen- erations. As between these two theories, the case is by no means as definitely closed as the confident assertions of various recent Mendelian writers would lead one to suppose. Both are still legitimate scientific hypotheses.” I .personally lean toward the view which seems to me to involve the fewest unproved assump- tions—the view, namely, that characters, genetic as well as somatic, may and do actually blend with one another per- manently. I have thus far reared F, and F. generations from three dif- ferent crosses among my various races of mice. I am able to 11 This point has been treated satisfactorily in various recent works (e.g., Babcock and Clausen’s ‘Genetics,’ pages 183-186) and need not receive further consideration here. 2 This I feel warranted in asserting, despite Castle’s recent defection from the ranks of those who uphold the view-point here advocated. 380 FRANCIS B. SUMNER report in the present paper upon unpublished data derived from two of these crosses. The series included here comprise, in one case 97 F; and 87 F. animals, in the other case, 154 F, and 84 F, animals. Two further series representing the widest of these crosses still remain to be studied, but the F, generations are not yet old enough to kill and measure. The mice here considered are much freer from abnormalities due to captivity than were the hybrids upon which I have re- ported in previous papers. I feel, moreover, that I am more nearly in a position to make proper allowance for such abnor- malities, and to know when they do and when they do not affect the validity of the results. Reference will be made below to this aspect of the case. So far as the hybrid series are concerned, I shall restrict my- self, in this preliminary discussion, to a comparison of the variability of the F, and F, generations, in respect to certain characters. In table 4 the standard deviations for five of these characters have been given. These are tail length (rela- tive), foot length, ear length, width of tail stripe, and depth of foot pigmentation. Since two different crosses are under con- sideration, and the two sexes have been treated separately, there are twenty pairs of figures to be compared, in our endeavor to ascertain the relative variability of the two hybrid generations. Owing to the fact that the absolute measurements for foot length and ear length are closely correlated with those for body length, and since the variability of the various series differs con- siderably in respect to body length, I have computed the net variability for foot and ear length (see below). To sum up the outcome of these computations, out of twenty pairs of comparable figures, that for the F, generation is greater in 8 cases and less in 4 cases, while the two are equal in 8 others. (I have considered two figures as equal when the difference be-- tween them is less than the probable error of that difference.) It is not, however, certain that the parent races in either of these crosses differ significantly in respect to ear length, while in one of the crosses (Carlotta-Calistoga) it is also questionable whether the wild stocks differ significantly in respect to foot VARIATION AND MENDELIAN INHERITANCE 381 pigmentation. Considering then, the fourteen remaining fig- ures, representing characters in respect to which the parent races differ unmistakably, we have— F, greater than F, in 8 cases F, less than F, in 2 cases F, equal to F; in 4 cases. Let us add that in two of the instances in the first of these groups, the differences are scarcely larger than their probable errors, while in only a few of the entire fourteen do they attain anything approaching statistical certainty. I think it is plain, therefore, that the exponents of the ‘mul- tiple factor’ hypothesis will derive rather cold comfort from the figures which I have to offer, even though the table as a whole may show a slightly preponderant increase of variability in the second hybrid generation. Of course, one answer is obvious. For studies such as these I have not chosen ‘favorable’ material. In these wild races, it may be contended, the number of factor mutations for each character has been so great that segregation cannot be expected to manifest itself appreciably in these small series. Such arguments are as unanswerable as they are unconvincing. Experience warns me that another objection is likely to be made to the validity of these results, although it is an objection which I believe to be utterly irrelevant when brought in this connection. It will be pointed out that each of the parent races with which I am dealing is in reality not pure, but is a mixture of genetically distinct strains. As evidence of this will be cited the wide range of variability within each race, and the fact that, for two characters at least, I have shown these variations to be hereditary. Even if all this is granted, however, and the con- tentions of the pure-line school be admitted in full, it still seems to me inevitable that we should, on Mendelian principles, have an increased variability in the F, generation of hybrids between two such mixed races. For there would in general be more fac- torial differences between representatives of two geographical races than between two individuals of the same race. The F, generation from such a cross would present more heterozygosis THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 30, No. 3 382 FRANCIS B. SUMNER than either parent race, taken by itself, and the chances of the segregation of extreme types would be much better in the F, than in the F; series. All this seems so obvious that it is difficult to understand how such an objection could be raised seriously. Passing to a very interesting incidental result of these studies, certain striking differences have been observed between the sexes. The most certain of these relate to the feet and the pelvic bones. When mice of the same size are compared, it is found in all but one of the eight local collections that the aver- age foot length of the males is greater than that of the females. In most cases this difference is statistically significant, whereas in the single exceptional case there is a practical equality between the figures. Also, in all of the four races whose skeletons have thus far been measured, the average length of the innominate bone (pelvis) is greater in the females, this difference, in three of the cases, being large in proportion to its probable error. It is of interest to note that the two differences just men- tioned are of opposite sign. We may profitably consider the bearing of these relations upon certain prevalent ideas regard- ing the origin of secondary sexual characters. Although many theories have been put forth in this field, it is my understand- ing that secondary sexual differences, in the higher vertebrates at least, are now commonly supposed to be due largely to the action of internal secretions or ‘hormones,’ produced by the gonads or by certain cells of these. To simplify the situation, we might assume, in each sex, the existence of a single charac- teristic hormone which determined all of the secondary sexual characters. In the present instance. the male hormone would, among other things, stimulate the growth of the feet and inhibit that of the pelvic bones. The female hormone, on the con- trary, might be supposed to stimulate the growth of the pelvic bones and inhibit the growth of the feet. Now, according to the foregoing hypothesis, such a state of affairs would inevitably bring about in each sex a negative cor- relation between feet and pelvic bones. For in each sex there would surely be wide individual differences in the amount of VARIATION AND MENDELIAN INHERITANCE 383 the hormone formed. As a consequence, we should have more and less masculine males, as well as more and less feminine females. This last, of course, is a quite obvious fact in many species, including man. Granting these differences in the potency of an agent, assumed to modify two characters in oppo- site directions, a negative correlation between these last would necessarily result when the animals of either sex were treated statistically. The interesting fact to be reported here is that precisely the opposite relation is found to obtain. Foot length and the length of the pelvis are found to be positively correlated, with a high degree of probability." Accordingly, if these two sec- ondary sexual differences are conditioned by ‘hormones’ at all (which does not, in itself, seem improbable), there must be at least two such hormones, which vary independently of one another. It may be permissible to call attention to the similarity be- tween this situation as regards the sexes and a condition already discussed in considering our geographic races. The case was mentioned of two characters (indeed, these same two charac- ters) which were correlated within the single race, but which nevertheless were found to have varied in opposite directions, when certain races were compared with one another. Whether such an agreement in behavior in these two cases has any general significance [ cannot even conjecture. CONSIDERATION OF DATA IN DETAIL The ensuing section will consist, for the most part, of a dis- cussion of some figures and tables which present certain portions of my data in graphic or in summarized form. It is upon these that the generalized statements in the first section were chiefly based. 18 [T will mention here that the evidence for the reality of such a correlation is somewhat stronger than the figures comprized in table 3. (See discussion below.) Measurement of the bones of four more races, now available to me, should render this point decisive. 384 FRANCIS B. SUMNER The map of California (fig. 1) shows the position of the eight stations at which my most extensive collections have been made. Two of these stations (La Jolla and Berkeley) were chosen largely as matters of convenience, but the other points were CARLOTTA <=>) >. UI CASS ER © FORT BRAGG) = = Rest = (se © CALISTOGA = | S@BAYs DUNCAN MILLS. Se a ae No nn 5© BER WY r= Sey, «© VICTORVILLE » / = 7 | DISTRIBUTION MAP | MUSEUM OF VERTEBRATE ZOOLOGY UNIVERSITY OF CALIFORNIA Fig. 1 Map of California, showing the eight stations at which representative collections of mice have been trapped. selected with reference to definite geographical and biological conditions. Mice of the species under consideration (Pero- myscus maniculatus) inhabiting four of these stations (Eureka, Carlotta, Fort Bragg, Duncan Mills) have been assigned by © VARIATION AND MENDELIAN INHERITANCE 385 recent systematists to the subspecies rubidus Osgood. Those from Calistoga, Berkeley, and La Jolla are assigned to gambeli (Baird), while ones from Victorville are assigned to the charac- teristically desert race sonoriensis (Le Conte). In table 1 we find the mean values given for most of the char- acters which have been subjected to measurement.“ It must be explained, however, that the mean values here given have, for most characters, been ‘corrected’ in such a way as to be com- parable with one another. This was rendered necessary by the fact that the mean body length (total length minus tail length) differed considerably among the various collections, owing largely to the inclusion of differing numbers of immature individuals. Since most of the characters here considered are rather strongly correlated with body length, their mean values in these different sets would obviously have not been directly comparable. The figures (or most of them) have accordingly been corrected for each series in such a way as to give their most probable value had the mean body length of the series in question been 90 mm. This was accomplished, I need hardly say, by the use of the so-called regression coefficient.!?7 As a matter of fact, the cor- rections which were applied were in most cases small in com- parison with the differences between the various races. They were largest in the Fort Bragg series, which contained a greater proportion of immature animals than did any of the others. 14 Weight and skull width have been omitted, for reasons which need not here be discussed. Foot-pigmentation has only recently been included among the characters measured, so that figures are not yet available for these races. Body length is not included for reasons stated in the next paragraph. 18 All animals below 80 mm. in body length were, however, arbitrarily excluded. 16 In a more complete presentation of these results, I shall include the original averages, but this does not seem necessary for the present. 17 The correction is obtained by the equation xz = r = y, in which z represents oy the difference sought between the corrected and obtained values for a given character, r the correlation between the character in question and body length, o and oy the standard deviations for this character and for body length, re- spectively, and y the difference between the standard value (90 mm.) and the actual mean body length of the lot in question. Of course, all these standard deviations and correlation coefficients had to be first computed for each race and sex separately. ‘gol108 BI[OL BT pus Ao[oysog oY} Ul 99TUI Jo soquinu JoZ1B] Jo aroy UOISNpOUL oy} 07 Burmo ‘roded gyEp~ AW ul pourezUos osoyy WOAJ UoAOYIp ATJYSIS oq OF PUNOJ oq [IM Soindy osoyy JO UlBYA9/) | 90 OF F8' PG 90° OF 6° F% 80° OF G8’ FZ GO OF 0 °Sz OL OF £9 Pc 20°0F L9°¥z 90° OF 19°S¢ GOOF TL°Sc 80'0F 0°91 90°OF 28°ST 60°0*F 62°ST 90 OF 28 °ST ST cE le Ostet 80°0F GL 80°0- O08 ST 90° OF €6 ST OL’ OF €0'8T 90° OF 88 ZT 60° 0 ZI SI 90°O-F 99° LT PL OF 60°07 Il OF “49 LT LOO c¥ LI HLON@T T1048 ywoOWwda SIAT¢Ud 60° OF €F'9z IL OF €6°Sz 60° OF 88°92 10° OF SP'9% OL'OF 18 °Sz OL OF G8'SZ IL OF LE°8¢ 60° OF £0°8z 98 0-F o& OF (Oz (Oa Oats ‘OF ‘OF iOcts ‘OF (ORF ‘OF ‘OF Raa vy OF 69,075 cy OF 66 8G 6b LG 19°GE 80°GE G8 9€ TE Ss cr LE 09 8E bE OF GE 68 6G GP 66 OF 96 1h IG GP 90° OF 90° OF 80°07 G0'0*F 10°0* 90'0F 80° 0F 90° OF 60° OF 80° 0F OL OF L0°0* 60° OF 20°0-F L100 90° OF 0€ “ZT\90'O-F GG LI) V0 OF 88° L1/F0 OF GL LT!90 OF OF 91/20 OF OL 9T\S0' OF 96 9T|90 OF GL 91/90 OF v9 91/20 OF €2° 91/20 OF 80° 21/90’ 0F G6 9T/S0 OF FI L120 OF €€ L190 OF 02° LT/S0' OF 99° L170 OF 99 61 1661 E106 GO 0G €9 61 18°61 92 °6I SL 06 0G 0G 18°06 96 02 6916 80°16 GG 1G cP OF0E 18 bh OF6Z 18 GEOFF €8 CS OF 9E 'F8 Ly OF ES 18 66 OF 08 18 Lv 0-09 G8 FP OFE9'S8 cG OF TS 16 6F OF69 F6 oh OF 3 OF OF Deka 09 0-F LaOrt 69 OF Sra0 89 OF LoeOTs 6G 0FL0 001/99 OF SF OFOL OOTITS OF 99 OF 8Z 0T|Z9 OF Gr OF FO FOTOP OF 90° 12/2S OF ZE €0T/SS OF SP 1Z|8E OFSP POL|Cr OF 82° €L 06° GL 92° SZ 62°92 88° EZ £6 GL IT 22 18°92 L9°G8 86 °F8 89°06 [9°26 06 £6 FS &6 Te 6) OL €6, 69 8L vg L9 fo) e % Ot % OF “*OT[TALOPOT A oy NOL BT “+ + Kopoysoq ** BSOYSI[BO ueound ‘SSvig W104 “+ Bq, 01189 ByIIN, GVUdaALuaA AdIULS TIV.L Uva Lood ING) Had TIVL @LOTOSav TIV.L ugqaWwon “wu 06 {0 yjbua) fipog pavpunjs 0} poonpas ,‘saon4, 1090] 1YH1a UL SLazID.DYD ULD}.L99 fo SaN]DA WHIP] tT ATAViL 6 iv) 3 VARIATION AND MENDELIAN INHERITANCE 387 For relative tail length, relative width of the tail stripe, and number of vertebrae no corrections have been introduced, since the first of these characters is correlated feebly and the last two probably not all with body length. Another way of comparing the mean values of these various characters in series of animals which differ in size has been employed by me in earlier papers. This is to divide each series into groups containing individuals of approximately the same size, and to compare the means of the corresponding size groups of different series. The mean difference between the two series under comparison may readily be computed according to a simple formula.'® Such mean differences I have not yet calcu- lated, however, for the present material. Figure 2 is based upon the corrected averages referred to above, the mean of the figures for the two sexes (not weighted) being employed for each race. The differences between these various racial means are plotted to scale along the vertical lines. The bone measurements have thus far been taken for only four of the eight races, though the various bones have already been prepared for measurement. It will be noted that for no two of the characters considered is the arrangement exactly the same. The nearest approach to agreement is found between the scales for tail and foot length, the chief difference being that for the latter character there are fewer distinguishable grades. The order for tail stripe is the same as that for tail length with this important exception, that Berkeley has been transferred from near the bottom of the scale to a point above the middle. Skull length in the four races for which figures are available follows nearly the same order as tail and foot length. In respect to the number of caudal vertebrae, Eureka stands at the top of the scale, as was true for all of the four characters just considered. The differences among the other four sta- tions are, however, of doubtful significance in the case of the vertebrae. 18 Journal of Experimental Zodlogy, April, 1915, p. 346. SUMNER FRANCIS B. 388 ‘SUOI}IO[[OD [BIO] OY} JO Noy ATMO 10F opvUT Udeq IV} SNY} OAVY SPUSMIOINSBOUT [VJO[OYG 4X0} oY} Ul poure[dxo sv ‘sIsvq UOUIUIOD B OF podNpal Udoeq [[B OAVY SUBOU OYJ, “SLOJOVIVYO JUILOYIP WAYS JO Son[vVA UBT OY} 07 YOodsSeI UT ,‘S9DBAI, [BOO] OY} JO UOIYepLIS Burmoys survisviq Z ‘BIg HLONI1 THIS SIAT3d JVYSILYIA WONVI IdIYLS WVL HLONST YV3 HLONI1 1004 HLONIT VL vwaun3 ya) AIV3NNIS AdayIe TUAYOLIIA B9L! FTHAYOLIIA oF BOL JTIAOLIIA enaaethena 37330 S26! | (.$)Sua!40U05,) 4 JVVAMO LIA FF 92°61 ATUAYOLIIA "GZ ynyo12iA Bove uae wr VI JUIAYOLIIA 4 Cupqued.) AI 1aN38 FF 2°62 vitor v1 aprers VDOLSIIVD 4PS6'6) (noun : vM0r v1 8 ¢'sz Ana wean vor v1 $20.02 ; (yqueB,) & w70r v1 WO01SI1¥9 “OL VOOLSIIVO vOOLSITV9 99vua 1u04 Ad ay70 yuaun 3un3 | ona smIW NVONNO STW NVONNG WP O'CE FTMAYOLOIA (SAP IQN4,) $11lW NYONNO Hr 8°C8 Sovu@ 1y04 VLLOTWYD vwaund 4 VU 1yOI EIN vauna wnnor V1 0@zt v1 LOTUv9 (59 1993,,) novus LuOd Ir 9°68 (snpias,) va gh S'C6 csopiqna) VL LOTUS £eé VARIATION AND MENDELIAN INHERITANCE 389 As to the two remaining characters (length of the ear and of the pelvis) the arrangement of the stations follows a quite dif- ferent order in the two cases, while for neither is the order like that for any of the characters previously considered. It must be said, however, that ear length is a rather erratic character in its behavior, some rather perplexing and contradictory results hay- ing been obtained in the course of these studies. The outstand- ing fact here is that the two extremes for ear length are dis- played by the Berkeley and the La Jolla animals, both of which ‘races’ are commonly assigned to the single subspecies gambeli. These differ in average ear length by about 1? mm., the difference being undoubtedly a real one, characteristic of the mice of the two localities. As to the innominate bone, the difference be- tween the two extremes here shown is doubtless statistically significant, despite its small magnitude. The lesser differences are not so certain. Two other modes of portraying these racial differences are shown in figures 3 and 4 and in figure 5, respectively. The former type of chart has been employed by me in many pre- vious papers dealing with variation and heredity in mice. In the case of the present material, I have plotted graphs of this sort for most of the characters here discussed, though only one -of these (that for foot length) has been included in this paper. As I have frequently had occasion to explain previously, each of the ‘curves’ in this figure results from connecting the mean values of this character for the various size-groups into which the individuals of a given race have been divided. Thus, ani- mals of the same size are represented by corresponding points on the various ‘curves,’ and the comparisons between the races are strictly legitimate. The essential agreement between figures 3 and 4 and the scale for foot length in figure 2 is obvious. Figure 5 consists of histograms based upon the individual fre- quencies of the various values for two characters in the eight local collections. These, of course, have the advantage of show- ing the total range and variability of the respective characters within each race. Such a method of representation would be inappropriate for most of the other characters under discussion, 390 FRANCIS B. SUMNER since the variability of these is so largely dependent upon the variability in the size of the animals concerned. As has already been noted, the order of arrangement is much the same for the two characters portrayed. The only difference concerns the relative positions of the three ‘gambeli’ races. Eureka (rubidus) Carlotta 99 ———— : Fort Bragg 99 omen Ve 22.0 DuncanMills » -+---+--+--. 6,10 = Ls} i) ° LaJolla(gambeli) Berkeley «€ — — — . H~RHORUDMYVABLO—-NHNORUDNSGO-HOAUDIB®OO—PHOhUAD Calisto ga «e ae Et / M A Life: 5 VictorvilleGonoriensi) ---------- 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 Fig. 3. Graphs showing foot length in the eight races (males). Body length is represented on the horizontal axis, foot length on the vertical one. The fig- ures along the various ‘curves’ indicate the number of individuals in the re- spective size groups. Each ‘curve’ connects the group-means for a single local ‘race.’ oS ° These three, however, taken together, occupy in both cases the same position, relative to the other races. The broad over- lapping of adjacent races and the large degree of distinctness of the extremes are well shown in this figure. Passing to a consideration of the correlations among the various characters, it will be seen from table 2 that all of the VARIATION AND MENDELIAN INHERITANCE ; 391 92 93 94 95 96 97 98 99 100 101 102 103 Fig. 4 Foot length of females (see fig. 3). FEMALES a o D co) co o © tw io) o o To) © a+ re) a. a ast a Ma N a Lo] Uy — 77 2) ry fo) ro) SRQLOHEMA-CMMREHtQATSAQROEHIagromonon a x a a © 392 FRANCIS B. SUMNER CARLOTTA 108 CASES Mean 10425 141 CASES Mean 104,01 CARLOTTA | | : : | BERKELEY 85 CASES Mean 35.93 LA JOLLA BERKELEY t LA JOLLA (31 CASES Mean 32.31 VICTORVILLE ViCTORVILLE 136 CASES . 133 CASES Mean 81.29 Mean 2842 34 $8 100 102 104 106 108 NO N2 4 16 18 (20 223 25 27 29 31 33 35 37.39 4) 3 45 47 49 Si SS SS ST 65669 193 8 TT IS BI BS AS BT BO. 9S 69 3910 15 105 107 109 WN NS HS M7 9 (2-20 22 24 2 7B WO 32 34 36 38 40 42 44 46 48 50 52 SK SS Fig. 5 Histograms showing distribution frequencies for the various values of relative tail length (percentage of body length) and relative width of the tail stripe (percentage of circumference) in the eight local collections (sexes com- bined). The broken lines connect the means of the various series. 393 VARIATION AND MENDELIAN INHERITANCE *10J po}yUNOI’ sny} 91¥ S10110 ajqeqoid oy} ut sorouvdososip yuoreddy ‘sueutoeds paSseuep Ul S}USUIOINSBOUT UTe}I00 JO Ov] 0} Sumo ‘o[qe} oy} Ur poze 1s IoqUINU oY} ULYY Lo][VUUs yeYMouTOS A[JUENbadJ SI poseq SI UOL}V[9L109 UOATS B YOM UOdn s[eNprAIpuUr Jo Jequanu oy, 1 810°0+ CoO sir O92 Os 4|. 008 OF (218 On- 1265 04> eZ Ores etOScOne |eSLGn\is oreo eben 60°0F 21°0+ | 80°0 7e0— | 29°:0+ | 80+ | I180+ | S8O0tT | IS 0+! WOT] 2 |6 \ sts eee ss QrTTAZOIOTA 10°0 92°0— | 80°:0F O1'0— | 82°0+ | 92°0+ | 22°0+ | 880+ |] 810+ | 0+] 84 |LJ : } 800 920+ | 80°0 Gz'0— | 880+ | gsso+ | 28°0+ | 29°0+ | WOT | e901 | 02 Sa age ee seen oy 20°OF I1'0- | 20:0 91'0— | es'0t+ | 180+ | zeot| zoot+ | 220+] 6F0+ | SOI |P gL'OF 60°0— | a1'0F go-0+ | zeo+ | 280+ | 60+ | woot | wot] wot] FH [d)............ goromsog 60°0F ¢0°0+ | 60°0F sz'0— | Ts0+ | ssot+| F0+ | 6zot | orot| gsot] gs |e 60°0F 10°0+ | 60'0F 9T'0— HOt | 60°04 |= 200% |oam el 9 rae eee 80°0F 00°0 | 80°0F 20°0- es'ot+ | oot | spot] 29 |e ; OL: OF 080+ | OF: OF 1z'0— igo+ | or0+| e2:0+| oF |d)...... ae en “Ssyji eoun OL'OF OL'0+ | 60°0F ¥z'0— ezo+ | Teo+ | zgo+ | Te s i ae IL'0F 00°0 | OL0F ZI‘0+ 3 ee0e | SEO: |r C-rs ea onl a: a eee eae 60°0F $0°0— | 80°0F 0z°0+ 6r'0+ | ogo+ | g9'0+ | 6¢ |© OL'OF OL'0+ | 60°0F Fe°0- SiOer | 82 Ot | ore Owe ele elles gree eres 60°0F 90°0— | 80°0F 9z°0— zoo+ | og'o+ | T9'0+ | ¢9 | 60°0F g0°0— | s0'0F ge'0— | scot | cot | rz0+| atot| ezot| egot+| 69 [4)........ a aa 80°0* 60°0+ | 20°:0F so'0— | s9o+ | z2:0+ | szot+| t¢0+ | wot] sot] 28 |e UdIuLs INGO add TIOWS doOnds SIATad ava cLooda (aia Tosayv) uga TIVL GNV AGOd TIVL GNV AGOd aNv A4qdog any AGdo@ aNv 4qa0a@ aNv Aqdog@ anv A4Gqo@ eens -WON anv 40d0& slajop.pyo 132470 pup yjbua] hpog uaamjag suo01jM7]a110) % ATAVL 394 FRANCIS B. SUMNER absolute values which have been determined are strongly and positively correlated with body length.!® The correlation is highest in respect to the three skeletal characters which are included. Of the two relative characters, the ratio of tail to body is seen to be negatively correlated with body length. That is, larger mice have proportionally slightly shorter tails. On the other hand, the relative width of the tail stripe (ratio to circumference of tail) does not appear to be correlated signifi- cantly with the general size of the animal. Much more instructive from our point of view are the correla- tions of the various characters, other than body length, with one another (table 3). That between tail stripe and relative tail length has been computed in the entire undivided populations, irrespective of size. Of the sixteen different figures (the races and sexes being treated separately), it will be seen that six are positive and ten negative, the mean for the entire series being slightly negative. Since, however, the probable errors for these single coefficients are, for the most part, nearly or quite as great as the coefficient themselves, it is quite unlikely that the preponderance of the negative values is significant. We may fairly assume, therefore, that the two characters are not appre- ciably correlated. I have likewise thrown together (though not in the present table), the data from the four ‘races’ taken north of San Fran- cisco Bay, and treated the entire lot as a single population, the sexes, however, being dealt with separately. Deviations trom the grand averages were employed in the computations. The stations represented are Calistoga, Duncan Mills, Fort Bragg, and Carlotta.2° Fairly high positive coefficients now appear between the two characters last mentioned (+0.366 for the 230 males and +0.351 for the 176 females): The significance of this fact has been discussed in the preliminary section of the present paper. 19 Exception should be made of the number of caudal vertebrae. 20 Since the Carlotta and Eureka collections are nearly identical in their mean characters, I have not included the latter. The four sets used were trapped and measured during the same season and are possibly more nearly comparable, on this account, than ones taken in different years. VARIATION AND MENDELIAN INHERITANCE 395 Two characters which are correlated with a high degree of cer- tainty are relative tail length and the number of caudal verie- brae. It seems, on first thought, curious that relatively longer tails should tend to have a larger number of vertebrae, while the absolute length of this appendage should play no appreciable part in the matter. I have not tested directly the correlation between absolute tail length and the number of caudal verte- brae, but the fact that the longer animals of my series (having, as a consequence, longer tails) do not have more vertebrae than the shorter ones, renders improbable the existence of such a cor- relation. It must be pointed out here that the slight differences met with in the number of the caudal vertebrae have little part in determining the differences in tail length, whether between races or individuals. These depend mainly upon the size, rather than the number of the separate bones. In determining correlations between the various other pairs of characters, a different procedure has been adopted, owing to the fact that these characters are all strongly correlated with body length, and therefore, in a population of mixed size, necessarily are correlated with one another. For this, as well as for other pur- poses, I have divided up the animals of eachlocal collection into groups of individuals differing by less than 2mm. Correlations have been determined for each size-group containing ten or more individuals, and the means of these coefficients employed.”! Positive coefficients of probable significance have been ob- tained for tail and foot and for tail and skull. The correlation between tail and ear length is far less certain, while none appears to exist between the tail and the pelvis. On the other hand, the foot and the pelvis seem to be correlated with a consider- able degree of probability. It is worth adding that the existence of a positive correlation between tail length and foot length, as well as between foot length and that of the pelvis, is made yet more probable from an inspection of certain series of mice which were measured before my methods were fully standardized, and which have therefore not been included in the present compu- tations. 21'The various methods here used have been discussed in a previous paper (Journal of Experimental Zodlogy, April, 1915). SUMNER FRANCIS B. 396 noc cree nn eee OZ 'O+||99F|F1E 0+ 662/602 O+|/SZZ FST ‘O+ ze | Pe O+1/8% | 10°'0+ Gp | 61 O+||FF | 78°0+ Ge | ST'O+|/&2 | Z0°0+ 82 | 98'O+I/8Z | 80°0+ ze | 80'0—||2c | 90°0+ Tg | 90°0+/|08 | €1°0+ 99 | 91°0+||8¢ | 610+ Bi 4 g | stataa = | T1nx8 & |aNVLooa|| & aNV TIVL J 5 0€ Lg agqanon | 82 0— I 0- 80°0— SIA1Tad anv TIVL OL 61 taiU (Urs FL O— ce 0+ aba Oa Zo‘ O+ 200+ ava anv TIVL ol‘ O+|/9S bE O+||8S FI 'O+||19 88'0+|/68 0€ 18 °0+||2S 8% 0+ é1 a4 Gl VG O& 09 uqdanoNn | oh 0+ 68 0— S0°0= ce 0+ €6'0+ ce O+ €6'0+ €1'0+ 880+ ve 0+ ch 0+ GZ O+|\sF | e8°0+ 80°0+1/22 | T8°0+ LOooOa aNvV TIVL ugaqnon ava -ALUaA aNv INGO add TIVL 910'0—- 80° OF 80° 0F 60° 0F L0°0-F GL OF 60° 0+ 60° 0 80° 0F OL OF OL’ OF HEX USE 60° 0F OL’ OF 60° 0F 0) Oa 80° 0 JA O= 60°0+ Ol’ 0+ 10°0+ vE0= AAV aw 0° 0-— 90°0-- TS Vase 10°0— T0"0- AG Ua $1 0— vI'O+ 80°0— L0°0— AdIULs TIVL GNV ING) Uadd TIVE (9x93 aas) y}buaz fipog aus ay} € HIAV.L sppnprrpur Buown suoynja.too yuasaidas sainbyf ay} ‘suunjos om} qsuyf ayy 7da90xa 70 UT Spree Meee tee tiatage. ofbie «Bs oF Ske aa ee rapes eeceseseseses *QTAIOIOIA % O PaO Ne SELES ele ee SAITO (p eT % OF een cea oaeh aepr ee ane A OTOM LOG > OF i hel ER 27 L023 16) SCCM heise riche shone cans | UTAT RULE ONCE RD OF BOISE OSD Cae ar Fell | 4104 SD OF g sian eo) eieiiahs ries ae eee ieee AO ORTB CY % OF ND OF ee er Cl oe ES EE TCC CODON 5: YET | ND Or *SL9JIDIDYD SNOLLDA U9AMNI9Q SUW01}D19LL0) VARIATION AND MENDELIAN INHERITANCE 397 Passing now to the hybridization of races, I shall not discuss this at any great length in the present paper, owing to the fact that two other important series remain to be measured, and I plan to publish a more complete account when these additional data are available. I have not, therefore, included a table giving the mean values of the various characters in the different crosses. Regarding these mean values, I may say that, in respect to characters in which the parent races differ, the figures for both the F, and F; generations, are intermediate, though not always midway between those of the parents. And with one exception (foot length in the Eureka-Victorville cross), the mean values agree pretty closely in the two hybrid generations. In this exceptional instance there is a perceptible reduction in the mean foot length in the second generation. It is important to know to what degree these hybrids have been modified by captivity, and particularly whether any of the relations to be discussed below are attributable to this cause. In addition to frequent sterility and a greater tendency toward adiposity, the chief modifications to be observed in many of the cage-bred mice consist in: 1) reduction in general body size; 2) reduction in the relative length of the appendages and in the width of the tail stripe, and, 3), for some characters, at least, an increase in variability. Now, in both of the crosses to be considered, the mean body length is somewhat less in the F, generation than in the F,, and in one of these (Eureka-Victorville) there has occurred a small reduction, both relative and absolute, in the mean length of the toot. There seems to be evidence, therefore, that the F2 gen- eration is somewhat less normal than the F;. So far as this fact has any bearing on the comparative variability of the two hy- brid generations, it must be stated that the probable effect of an increase of abnormality would be an increase of variability. As regards tail Jength, at least, I have definite evidence of this in comparing the standard deviations of wild and cage-bred mice.” The figures for the latter are much larger. Thus, if abnormality 22 T have not yet computed standard deviations for most of the characters in cage-bred animals of the pure races. THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 30, No. 3 398 FRANCIS B. SUMNER due to captivity be really a factor in accounting for the relations to be discussed presently, its mode of operation has been to increase rather than to decrease the appearance of segregation. It should be said, too, that the F; generation, in the Carlotta- Calistoga cross, was visibly more normal than in the Eureka- Victorville one. Indeed, in the former, save for the fact that most of the animals were appreciably smaller than wild ones, there was rarely any indication of abnormality in either hybrid generation. It is in — i> BS . aS Series 339, 368. Pieces A, B,C Geni isoeechfentwalted)| 22 2 9 | 2 ee iar ably ernba bts oO cians Wie To 5: (cece ered ioe Pde ory . 5) . 1G 4 1 2 76 6 TG 34] 14 26 26 Series 321. Pieces B, C (fig. 1), IB 6.) 12 6 26 50 50 each, from well fed ani- IIB 20 | 36 10 24 8 mals: I,10-l1lmm. II, 16-18 IC 6 8 86 mm. IIc 4] 14 2 16 62 IA 60 | 40 IIA 90 | 10 Series 201,277. Pieces A, B,C, IB 20 80 D (fig. 2), 10 each, from well IIB 60 | 40 fed animals: I, 5-6 mm. II, IC 10 | 40 10 30 10 16 mm. TI¢ 90 | 10 ID 60 | 30 IID | 100 nutritive material for the development of a new head, but there are various reasons for believing that this is not the case: first, head-frequency is lowest in pieces from levels near the mouth where the amount of nutritive material in the pieces is greatest; second, it has been shown elsewhere that head-fre- quency may be altered experimentally in both directions with- out altering the amount of nutritive material (Child, ’16; Behre, 408 Cc. M. CHILD 18), and further data bearing on this point are presented below; third, normal heads and often biaxial heads may develop on pieces so short that the whole substance is used up in the development of the head or heads. Taking all the facts into account, it is evident that while the amount of nutritive material may play some part in determining head-frequency, it is not the primary factor in determining the differences between large and small animals. HEAD-FREQUENCY IN RELATION TO NUTRITIVE CONDITION It has been shown that starvation in P. dorotocephala is ac- companied by various changes in physiological condition. The susceptibility of body wall and ectoderm (probably also of paren- chyma) increases from the beginning of starvation, that of the alimentary tract in the later stages (Child, ’15 a, chap. VII, 19 ¢). Carbon dioxide production and oxygen consumption decrease in the early stages of starvation, undoubtedly because of the decrease in activity of the alimentary tract in the absence of food, but later increase (Child, ’19a; Hyman, 719 a, and in advanced stages of starvation the animal may have a much higher rate of respiration than at the beginning of starvation, and this rate is still further increased by renewed feeding. As regards rate of respiration and susceptibility, the starving animal apparently becomes physiolog- ically younger, and with renewed feeding may again begin growth and progressive development from a stage physiologically earlier than that at the beginning of starvation. It is of interest, in the light of these facts, to determine the effect of starvation upon head-frequency, and data along this line are given in table 2. Table 2 shows in all cases that the head-frequency in pieces from starved animals is less than in fed animals, even when of the same size. Certain points in the various series demand brief notice. So far as actual difference in percentage is concerned, it will be ob- served that the effect of starvation is in general most conspicuous in the most anterior pieces (A) and in series 327 in the long pieces (X). The differences in these cases are entirely or almost en- HEAD-FREQUENCY IN PLANARIA 409 tirely between normal and teratophthalmic, and it has been found that very slight changes in physiological condition will determine the development of teratophthalmic instead of normal heads or vice versa. It is probable, therefore, that the large differences TABLE 2 Head-frequency in relation to nutritive condition Eeeogaa logs [be e | a | 2 S28] gel Bl ere ieee oe ty lea | oy lade 4 4 iS] iS] < < a Series 368. Pieces A, B, C (fig. IA| 10 90 1), 50 each I1A| 54 | 42 4 I, from animals 7-8 mm., IB 2 96 2 starved 17 days IIB 2 96 2 II, from animals 7-8 mm., well IC 6 88 6 fed (CELE ee state QP 76 6 LA) 20 72 2 2 4 IIA | 38 58 4 Series 327. Pieces A, B, C, X || IIIA]! 30 66 ft (fig. 1), 50 each IB 6 6 4 82 2 I, from animals 6-7 mm., IIB 2 2 94 2 starved 19 days IIIB 6 30 4 14 42 II, from animals 6-7 mm., LG: 92 8 starved 18 days, fed once IDK 100 III, from animals 6-7 mm., || IIIC | t 96 heavily fed IX | 30 70 IIx | 30 70 DEEXA -70 18 12 Series 320. Pieces B, C (fig. 1), | 50 each lk AWE? Qe slOls| Guna 2d I, from animals 14-16 mm., IIB 4 46 6 30 14 starved 40 days Lé 2 4 86 8 II, from animals 14-16 mm., IIc 2 6 2 22 68 well fed Series 638. Pieces A, B, C (fig. IA | 22 70 2 2 2 2 1), 50 each -IJA| 88 8 4 I, from animals starved 93 days, IB 30 12 10 32 16 reduced from 25 mm. to 7-8 IIB 2 60 8 12 18 mm. LC. 4 4 10 82 II from animals20mm., well fed Ti¢ 10 2 16 72 410 C. M. CHILD in percentage in normal and teratophthalmic in the anterior pieces have no greater significance physiologically than smaller differences in percentages between headless and anophthalmic for example. In series 368, 327, and 320 starved and fed animals and pieces are of equal size, but in series 638, well-fed large animals are compared with animals originally of still larger size, but reduced by starvation to a small fraction of the original size. In all cases the comparative results are the same, the head- frequency being lower in the starved lot. In series 327 the ani- mals of lot II are fed once after eighteen days of starvation and are compared with animals starved nineteen days (I) and heavily fed animals of the same size (III). Here the single feeding of lot II apparently increases head-frequency in A, but not else- where. The head-frequencies are low, even in the fed animals (IIT) of this series, except in the long pieces (X), because of the small size of the animals, but the difference between starved (1) and fed (III) is very great in B and X. That lack of available nutritive material is not the primary factor in determining the decrease in head-frequency in starvation is indicated by various facts: for example, the slight effect of the single feeding in series 327 II upon head-frequency indicates that some other factor than nutritive material is chiefly concerned. These pieces were well filled with food, at least during at least the earlier stages of regen- eration. Moreover, all the series show that starvation and feed- ing do not alter essentially the relation between head-frequency and body-level, although there can be no doubt that in starva- tion nutritive reserves are exhausted first and resorption of the alimentary tract proceeds most rapidly in the anterior body re- gions. In other words, these regions are most starved, but still show a much higher head-frequency than the less starved regions near the mouth, and even after heavy feeding there is less surplus nutritive substance in the anterior body region, because of the alimentary tract is less extensively developed there than in re- gions about the mouth, yet anterior pieces show the highest head-frequency. In short, it is evident from these data on head- frequency in relation to starvation as well as from those on size HEAD-FREQUENCY IN PLANARIA 411 in the preceding section that amount of available nutritive ma- terial is not the primary factor in determining head-frequency. Further discussion is postponed to the final section of the paper. HEAD-FREQUENCY IN RELATION TO MOTOR ACTIVITY OF PIECES These experiments consist essentially in determination of head- frequency in pieces which are repeatedly stimulated to motor activity during a longer or shorter time beginning immediately after section (II) as compared with that of smaller pieces which are left undisturbed (I). The results for two series are given in table 3. In both cases the pieces B and C (fig. 1) are used be- cause such pieces have neither an extremely high nor an extremely low head-frequency and change in both directions is readily possible. In A pieces and in pieces from the posterior zooids the head-frequency is usually so high that only decrease appears clearly in experiment. The methods of stimulation were various: currents of water were used to loosen the pieces from the glass, some motor activity usually following; individual pieces were loosened and turned over with the aid of a camel’s-hair brush and usually attempted sooner or later to turn back again. Gentle stroking or slight pressure with a camel’s-hair brush was also found to be effective. In series 302 such stimulation was re- peated at least every hour for some eight hours after section, after which the pieces were left undisturbed overnight. On each day following until the new heads and eyes were distinctly visible the pieces were stimulated at least once an hour from 8.30 a.m. to 6 p.m. and at least twice between 8 and 11 p.m. In order to equalize oxygen supply as far as possible the water on the un- stimulated lots (I) was stirred gently at the times when the other lots were stimulated, the stirring being usually not sufficient to induce motor activity. In series 498 the stimulation of II was repeated every five to ten minutes during the first two hours after section, once an hour during the second two hours, and again every five to ten minutes during the third two hours. After this the pieces remained undisturbed. It was known at the time of this experiment that the continuation of the stimulation during several days as in THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 30, NO. 3 412 C. M. CHILD series 302 was wholly unnecessary, since under ordinary condi- tions it is determined within five or six hours after section whether a piece shall give rise to a head or not (Child, ’14 d). In table 3 the column ‘teratophthalmic’ is divided into two columns, a and 6b, which represent two different degrees of ter- atophthalmia. Column a includes the cases in which the two pigment spots of the eyes are distinct, but differ in size or are TABLE 3 Head-frequency in relation to motor activity of pieces TERATOPH- 2 : ° 4 = THALMIC Oe 2 2 r 2 : a b : 5 s : : 3 m4 Z a < < a Series 302. Pieces B, C (fig. 1), from animals 16-18 mm., starved 8 days IB 8 8 26 58 I, undisturbed, water occa- IIB 10 | 14] 12 32 32 sionally stirred Ic 4 4 12 80 II, stimulated to motor ac- LLC 4 2 ae 2 18 64 2 tivity as often as possible — during 6 days after section Series 498. Pieces V, C (fig. 1), from well-fed animals 16-18 mm. I, undisturbed REO A 88) 44 6a) ae IIB] 20 | 48 | 32 II, stimulated to motor ac- ge rae Nec ? ze ba Ses E JAKE) 8 | 36 | 32 14 10 tivity frequently during 6 hours after section somewhat asymmetrical in position. Column b includes those cases in which the two pigment spots are more or less approxi- mated to each other and united by a band of pigment or partly fused. The forms of column a are somewhat nearer normal than those of column b. The head-frequency data for these two series in table 3 show that in every case the stimulated lots (II) have a distinctly higher head-frequency, or more specifically, approach more closely to normality. The difference appears as clearly in columns a and HEAD-FREQUENCY IN PLANARIA 413 b under the head ‘teratophthalmic’ as in the other columns. In series 498 with well-fed animals the level of head-frequency is higher in all pieces than in series 302 with animals starved eight days and only moderately fed for several weeks previously, but the differences between I and II are distinct in every case. Moreover, the short period of stimulation in series 498 is appar- ently about as effective in increasing head-frequency as the long period in series 302. According to the evidence from head deter- mination (Child, ’14 d), this is to be expected, but it is of interest to find the data confirming expectation. DISCUSSION Extensive investigation of the conditions determining and af- fecting head formation in isolated pieces of Planaria has led to XxX Figure 3 certain conclusions concerning the physiological relations of the new head to other parts of the piece. It is a well-known fact that regulatory development at the anterior end of a piece begins with the formation of a head, what- ever the level from which the piece is taken, and that the parts which normally lie between the head and the level represented by the piece are formed later by reorganization of regions posterior to the new head and never develop unless at least a rudimentary head is formed first (Child, ’11¢). In short, the head seems to arise as something more or less independent physiologically of other parts of the piece, while the further reogranization of the anterior regions of the piece occurs only under the influence of the new developing head. Moreover, it has been shown that the relation between the cells from which the new head develops (fig. 3, 2) and the rest of 414 Cc. M. CHILD the piece (fig. 3, y) is In a sense antagonistic, in that conditions which stimulate or accelerate the physiological activity of x in relation to y, or conversely, decrease the activity of y in relation to x increase head-frequency, while conditions which increase the activity of y in relation to x or decrease the activity of x in rela- tion to y decrease head-frequency (Child, 714d, 716). These re- lations have been expressed in the formula head-frequency Gabel: This formula is merely a brief expression of relations indicated by the experimental data. The physiological situa- tion is apparently as follows: In an isolated piece certain cells («) along the cut surface are isolated by the cut from all the correl- ative factors that formerly reached them from parts anterior to the cut and are also stimulated by the wound. These correl- ative factors represented a large part of the factors which deter- mined the differentiation and behavior of these cells as a part of the body, and in their absence the cells tend to lose this differenti- ation and to become physiologically younger. But the relation of the cells 2 to the regions posterior to them (y) have not been altered by the cut, and any correlative factors which reach x from y must tend to prevent or retard its dedifferentiation and independent development. Consequently, the result in any par- ticular piece will vary as one or the other of these factors has the ascendancy. For the present we may express this relation in terms of the activity of the regions x and y._ If the activity (and probably the energy-liberating activity is primarily concerned) of x is sufficiently intense as compared with that of y, the cells of x will be in large measure independent of y and will dedifferentiate and develop anew in spite of y, and the product of this develop- - ment will be a normal head; i.e., the primary developmental reaction of planarian protoplasm will occur (Child, *11 ¢, ’15b, pp. 96-102). If on the other hand, the activity of y in relation to that of x is sufficiently intense to inhibit or retard these pro- cesses to some extent, the development of the head will be retarded and its structure will range from a slight degree of teratophthalmia to extreme anophthalmia, or in the extreme case the development of a head is completely inhibited and the piece remains acephalic. HEAD-FREQUENCY IN PLANARIA . 415 With this brief statement of the general conception of head- formation in pieces, we may turn to the interpretation of the data recorded above. As regards the relation between head- frequency and size, it was pointed out above that the rate of oxi- dation is unquestionably higher in the small than in the large animals. In isolated pieces of a smaller animal the tissue throughout is physiologically younger and more active, conse- quently the region x in its reaction to the absence of the parts in front and to the wound does not undergo so great an increase in rate in relation to y as in a piece of a larger older animal, and is therefore less independent of y in its development; i.e., shows a lower head-frequency than in the pieces from larger, older animals. From this point of view the lower head-frequency in the younger animals is essentially a consequence of their higher rate of metabolism or oxidation. As regards the starved animals, it has been pointed out that lack of available nutritive material cannot be the primary factor in determining the lower head-frequency. The facts noted above concerning changes in susceptibility, CO. production and oxygen consumption during starvation indicate that the decrease in CO: production and oxygen consumption in the earlier stages are merely the result of the decrease in activity of the alimentary tract in the absence of food and that the rate of oxidation in ectoderm and body wall—probably also in the parenchyma— increases from the beginning of starvation. These are the re- gions chiefly concerned in the determination of head-frequency in pieces, and if the conclusions concerning rate of oxidation are correct, it is evident that the lower head-frequency in starving as compared with fed animals is due to difference in the relation rate x Pately of the same sort as in young animals as compared with old. In other words, the rate of energy-liberating metabolism being higher in the piece as a whole (fig. 3, y), the changes in the region x do not increase its rate over that of y so far as in well- fed or old animals. Consequently, in starving animals dedif- ferentiation and new development of x is less independent of correlative factors in y than in fed animals, and head-frequency is therefore lower. 416 C. M. CHILD In the pieces stimulated to motor activity, head-frequency is increased as compared with those at rest. The interpretation suggested for this fact is that the cells at the anterior end which are concerned in head-formation (fig. 3, x), are stimulated to a greater degree than the rest of the piece by the forward move- ment and that their independence of the correlative factors in y is therefore greater than in the pieces at rest. In other words, rate x undergoes increase in relation to rate y in this case and the result is increase in head-frequency. It is clear, then, first, that the three aenaeiie te factors, age as indicated by size, nutrition, and motor activity, influence head- frequency; second, that this influence consists in increasing or decreasing the relative number of pieces in a lot which produce either heads or heads of a particular degree of development. As regards the various degrees of head-development, it is a fact of considerable interest that the different forms of head produced remain the same as regards their structure, whether such factors as size and region of piece (Child, ’11 b, 714 b, ’14 d), physiologi- cal age, nutrition, and motor activity or an external chemical agent such as KNC (Child, ’16) or a physical agent such as tem- perature (Behre, ’18) are concerned. It is impossible to escape the conclusion that the characteristics of the series of head forms from normal to acephalic are determined by the specific consti- tution of the protoplasm and that the effects of the various physi- ological conditions in altering head-frequency are essentially quantitative and non-specific. In the light of all these and var- ious other facts, the quantitative interpretation presented above has gradually taken form and thus far no facts have been ob- served which are in conflict with it. In conclusion, it may be pointed out that the tabulated data confirm earlier work on head-frequency in relation to length of piece and region of body (Child, ’11 b, 716). In table 1 the pieces of series 201 and 277 are longer in relation to total length of body than the pieces of other series of the same table and the head- frequency is higher in these longer pieces. Similarly, the X pieces of series 327 in table 2, being longer than the A pieces of the same series, show a higher head-frequency than these, al- HEAD-FREQUENCY IN PLANARIA 417 though the anterior ends of both are at the same level of the body. The tables also show, as do the earlier data, that in pieces of equal length head-frequency decreases with increasing distance from the head of the animal, back to the level of the posterior zooid (Child, ’11d). Series 201 and 277 of table 1, the only series of this paper in which the posterior zooid is included, show an increase in head-frequency in the region of the posterior zooid. These relations between head-frequency and relative length of piece and region of body have been discussed and interpreted in the same terms as the relations between head-frequency and physiological condition considered in this paper (Child, ’11 b, 14b, 714d) and recent work on carbon-dioxide production in pieces (Robbins and Child, ’20) has added further evidence in support of the conclusions drawn from the earlier investigations. SUMMARY 1. Head-frequency in the regeneration of pieces is lower in physiologically younger (smaller) than in physiologically older (larger) animals. 2. Head-frequency is lower in pieces from starved than in pieces from well-fed animals, even when the two are of the same size. ; 3. Head-frequency is higher in pieces which are frequently stimulated to motor activity during at least several hours after section than in pieces remaining undisturbed. 4. The range of head forms is the same in relation both to physiological conditions and to external chemical and physical agents, and the changes produced are changes in the frequency of the different forms. This non-specific effect of both physiologi- cal and external factors indicates that the action of these factors is essentially quantitative. It is shown that the quantitative interpretation of changes in head-frequency previously advanced serves for the facts presented in this paper. 418 C. M. CHILD BIBLIOGRAPHY ALLEN, G. D. 1919 Quantitative studies on the rate of respiratory metabolism in Planaria. II., Amer. Jour. Physiol., vol. 49. Beure, E. H. 1918 An experimental study of acclimation to temperature in CuHILD, C. CHILD, C. Planaria dorotocephala. Biog. Bull.. vol. 35. M. 1911 a Experimental control of morphogenesis in the regulation of Planaria. Biol. Bull., vol. 20. 1911 b Studies on the dynamics of morphogenesis and inheritance in experimental reproduction. I. Jour. Exp. Zool., vol. 10. 1911 c¢ Studies, ete. II. Jour. Exp. Zool., vol. 11. 1911 d Studies, etc. III. Jour. Exp. Zool., vol. 11. 1914a Starvation, rejuvenescence and acclimation in Planaria doroto- cephala. Arch. f. Entwickelungsmech., Bd. 37. 1914b Studies, ete. VII. Jour. Exp. Zool., vol.:16. 1914 Asexual breeding and prevention of senescence in Planaria velata. Biol. Bull., vol. 26. 1914d Studies, ete. VIII. Jour. Exp. Zool., vol. 17. 1915 a Senescence and rejuvenescence. Chicago. 1915 b Individuality in organisms. Chicago. 1916 Studies, ete. IX. Jour. Exp. Zool., vol. 21. 1919a A comparative study of carbon dioxide production during star- vation in Planaria. Amer. Jour. Physiol., vol. 48. 1919 b The effect of cyanides on carbon dioxide production and on susceptibility to lack of oxygen in Planaria dorotocephala. Amer. Jour. Physiol., vol. 48. 1919 c Susceptibility to lack of oxygen during starvation in Planaria. Amer. Jour. Physiol., vol. 49. M., anp McKiz, E. V. M. The central nervous system in teratoph- thalmic and teratomorphic forms of Planaria dorotocephala. Biol. Bull., vol. 22. Hyman, L. H. 1919a Physiological studies on Planaria. I. Oxygen consump- ROBBINS, tion in relation to feeding and starvation. Amer. Jour. Physiol., vol. 49. 1919 b Physiological studies on Planaria. II. Oxygen consumption in relation to regeneration. Amer. Jour. Physiol., vol. 50. 1919 ¢ Physiological studies on Planaria. III. Oxygen consumption in relation to age (size). Biol. Bull., vol. 37. Harriet L., anp Curip, C. M. 1920 Carbon dioxide production in relation to regeneration in Planaria dorotocephala. Biol. Bull., vol. 38. THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 30, No. 4 MAY, 1920 Resumen por el autor, E. C. McDowell. Estacion de Evolucién Experimental, Cold Spring Harbor, L. I. Herencia de las cerdas en Drosophila. III. Correlacién. Los coeficientes de correlacién y las lineas directas de regresi6én que indican el grado de semejanza entre los grados de cerdas de los padres y progenie en una raza de Drosophila, seleccionada durante cuarenta y nueve generaciones con el propdsito de au- mentar el ntiimero de cerdas, indican que en las primeras genera- ciones seleccionadas, los padres de grado elevado produjeron mas progenie dela misma clase que los de grado inferior; en las ultimas generaciones no sucedié esto. Estos hallazgos estén comple- tamente de acuerdo con los efectos de la seleccién sobre los med- ios de la raza; en las primeras generaciones estos medios se ele- varon por seleccién, mientras que en las generaciones ulteriores la selecci6n no produce efecto alguno. Para probar finalmente el plasma germinativo presente en la raza seleccionada, se sus- pendié la seleccién a partir de la cuaranta y nueve generaci6n, cridndose un numero elevado de moscas descendientes de un- mismo par de la generaci6n cuaranta y nueve bajo las mismas con- diciones ambientes. Las correlaciones entre los padres y la progenie en las generationes cincuenta y dos y cincuenta y tres (en las que se hicieron mas de 31.000 numeraciones de cerdas) indican la completa ausencia de tendencia alguna en las moscas de grado elevado hacia la produccién de descendientes de un grado mas elevado que el producido por los padres de grado mas bajo. El plasma generativo de la raza parece haberse trans- formado en uniforme por la seleccién e “‘inbreeding;” la pri- mera ha reducido la cantidad de diferencias genéticas entre las células germinales; no hay prueba alguna de que diferencia gené- tica alguna no presente al principio de los experimentos haya sido operativa. Translation by José F. Nonidez Carnegie Institution of Washington AUTHOR’S ABSTRACT OF THIS PAPER ISSUED BY THE BIBLIOGRAPHIC SERVICE, MARCH 20 BRISTLE INHERITANCE IN DROSOPHILA! III. CORRELATION EDWIN CARLETON MacDOWELL Station for Experimental Evolution, Cold Spring Harbor, Long Island EIGHT FIGURES CONTENTS MEG OGUC ELON ape Mer cese cee: < h ene Se ed, REL aee ae SLD 2Y* OE eT AO 419 TEvOUStreSsUltss a yue. eae eh: sets. |.(o% Ut tal Severe so, PRR ENS: k . Se ES ee 420 Gravee eniorn ieee ARR eee 24's 28 SUR seom taney cr Otel oy ali AOR Sg ARUN pal 421 INTC HOC Sip et Nig ee RE TCT: cok ce cee RN Ee RED SIAC Pcs | SAME Cote f 423 VESTUG Se ee co erase, PPAR SEIIRS. soln eM cri nee Mn Saree | 2. .” MORRO I YS 432 CONClUISIONSs PAAR 3 (Santee toch es DEED AE te INAS SA a 440 NMevigdatar| CeLection SUSPENGeG: cx. Oi «sack dagseed « srsestomme oaysre.ove Sadek whe! obs yo 442 CRIM OMT cote hse tale pers: «0s 5-5 caus ten Ges Pee. oe. oie SE AS 443 IVESUIESR Eee eee, Seniesa Se ee eee oe SENOS OR 445 Concluslonss = See seh ee ea. oa ret aR At. Ok ee ted 448 MRaileskysh. femal ess pts hey) ste otra. ti occa ed ARR inte einai Cd ees 448 DIScussiongomlibersbureny- | nese cet. oe ek ee aa eee 451 SHEER CIEE ae lig el oc Le A A 8 AO Ws SO a MRI EN 458 BT LO ERA Natt cts Mes te pe eee PAH IAT oe ache, «hah hmm Ramee, REY oe) Abr 459 INTRODUCTION This paper is a coordinate part of the second paper in this series (MacDowell, ’17 b) and would have followed it immedi- ately had not the war intervened. The preceding report gave the details of the experiments and presented results in the form of means, standard deviations, and frequency distributions; this paper analyzes the same primary data by means of correlation tables. New data are presented, involving over 31,000 bristle counts, from four additional generations raised without selection as a final test of the germinal constitution of the selected race. 1 Acknowledgment should be made of the cooperation of J. Gowen, J. Krafka and E. M. Vicari in the calculation of the constants in this paper, and espe- cially of Miss Vicari’s part in constructing the figures. 419 420 EDWIN CARLETON MacDOWELL A summary of these results has already been presented (Mac- Dowell, ’17.a). Due to the lapse of time since the appearance of the second report, a brief résumé of the general results seems needed to introduce the new calculations. PREVIOUS RESULTS Extra dorsocentral bristles, in a certain race of Drosophila melanogaster, were found to act as a simple Mendelian char- acter when crossed to normal wild flies. The number of these extra bristles varied, thus affording material for the employ- ment of artificial selection. Starting from one pair of flies with extra bristles, and making brother by sister matings throughout, selections for increased bristle numbers were made for forty- nine generations. After the early generations, this selection did not modify the means of the race, atlhough the high limit of variation was far from being reached. Selections for de- creased numbers of extra bristles at the beginning established a low race; similar selections from the later generations of the high-selected race were unsuccessful. Yet after a cross with normal these same generations of the high-selected race became immediately as amenable to low selection as were the unselected flies at the beginning. These general results were interpreted as being due to genetic differences among the germ cells of the original extra-bristled flies. These differences were quite inde- pendent of the single factor that controlled the appearance of any extra bristles. Selection reduced the number of these dif- ferences by reducing the amount of heterozygosis; crossing with normals increased these differences by increasing the amount of heterozygosis. The numbers of extra bristles that appear on a fly are influenced by external conditions as well as by genetic factors. The average number of extra bristles can be largely controlled by the amount of food a brood has a chance to eat before pupating; this amount depends upon the amount of food present and upon its attractiveness. This external influence naturally acts as a blind to the relationship between the grade and the genetic constitution of an individual, but it is obvious that this was not a complete blind, as there must have been BRISTLE INHERITANCE IN DROSOPHILA 421 some connection between the degree of bristling and the genetic constitution before selection had started and again after a cross with normals. It has been shown that only in the early generations are especially high parental averages associated with exceptionally high filial averages. Since the offspring can never be raised under an environment identical with that of the parents, the failure to find any correspondence between the means of the parents and offspring in later generations may be due to the dif- ferent environments in successive generations. However, the highest-grade flies may have been produced by the highest- grade parents, however well the environment may have con- cealed such a relationship between the means when different generations were compared. The whole frequency distribution of the offspring may be centered about quite a different mean from that of the parents, yet the relative positions of parents and offspring in their own distributions may be the same. The study of the relationship between parents and offspring in indi- vidual families offers a different line of attacking the problem. Such an approach goes directly to the heart of the question that selection seeks to answer empirically: are there genetic dif- ferences between flies with different numbers of extra bristles and, if so, do such differences arise continuously? If the grades of the parents and offspring bear any direct relationship to each other, selection can progress; if such a relationship appears at first and then disappears, one may conclude that selection will not continue to be successful. The demonstration that these are the facts of the relationships between parents and offspring in the various selected generations will strengthen considerably the conclusions drawn from the results of selection based on the means. CORRELATION The obvious method of investigating the relationship be- tween the parents and offspring, is that of the correlation table, with the coefficients of correlation and regression calculated from it. This method affords a clear description of large masses 422 EDWIN CARLETON MacDOWELL of data that would otherwise be difficult, if not impossible, to summarize. The relationship of every fly to its parents has its effect on the results. However, correlation tables afford only superficial descriptions, and accordingly great care is necessary in determining their meaning. A positive correlation coefficient indicates that higher parents had higher-grade offspring; a negative coefficient indicates that higher-grade parents had lower-grade offspring. But from any single coefficient little can be said of the underlying genetic significance. In any par- ticular generation the high-grade parents may have had, on the whole, better conditions than the low-grade parents and conse- quently produce higher-grade offspring; this would give a posi- tive coefficient. The reverse might be equally possible in an- other generation; the lower-grade parents might, by chance, have found better conditions and so have produced higher- grade offspring than the high parents; in such a case the correla- tion would be negative. Although a single plus or minus coeffi- cient would not bear much evidence, a series of one or the other would indicate that real genetic phenomena were involved. Even such a series cannot be authoritative unless the experi- mental procedure has been the same for all families included. This is one of the greatest difficulties with results based alone on mathematical treatment. In many such cases the material has lacked homogeneity. In this respect the bristle data have one great advantage, being derived entirely from experimental procedure; their origin and the nature of the families put together are fully known. The selection of low-grade flies tends to isolate the small ones and therefore to carry on any conditions that tend to make small flies, such as weakness or disease. Such weakness, apart from any germinal cause, would tend to make the offspring of these low-selected parents lower than the offspring of parents selected for grades not associated with weakness. If such differently selected lines be united in one correlation table, even though they are all the result of brother and sister mat- ings and are all the same number of generations away from common ancestors, the finding of a positive coefficient would BRISTLE INHERITANCE IN DROSOPHILA 423 not prove that genetic differences exist between high and low flies. Thus there are two non-genetic factors that may occasion positive correlation coefficients: environment and lack of vigor due to the continued selection of small flies. Methods In the correlation tables nothing but the progeny of one original pair of flies has been included. As indicated, the return-selected lines probably tend to increase the amount of correlation unduly. However, one of these lines has been included in the tables, namely, the line started from the 16th generation of the high race; two generations of the second re- turn line, started from the 27th generation, have also been in- cluded. The later generations of this last line consisted of such obviously inferior flies that they were omitted. The line of low-grade selection started at the beginning has also been included. The essential genetic difference between this low race and the two return-selected races has been pointed out. There is no suggestion that the low-selected race at the begin- ning had any lack of vigor; the means were immediately influ- enced and there was no question as to the distinctness of the means from those of the high race. On the other hand, the first few generations of the return-selected races showed no influence at all of the selection, but there did gradually appear a slight relative lowering of the means, which may well be due to accu- mulated weakness. The inclusion of these two sorts of low- selected lines will make the comparison of the correlation in the corresponding periods more fair. In the tables the families were arranged according to the mean bristle grades of the parents. Since in most cases the parents were of the same, or within one bristle of the same grade, the averages were only slightly different from the actual grades. Sons and daughters were tabulated separately. The tables themselves are too numerous for publication in a journal. The correlation coefficients, their probable errors and the ratios of the errors to the coefficients are given in table 1. EDWIN CARLETON MacDOWELL 424 60'T | ¥890°0F6090'0— | LLLL°P | L0G ¥9°0 | SOSO OFPZEO O— | LOTE'S | 19Z L 879t' 9} OL £6 40'F | PISO OFTLZI'0+ | 690h'F | 629 80'T 1€80' OF 8¢80' 0+ | 9890'€ | €8¢ 8 CG0G' 7 | GZ GG 6'€ | 9220 OF O60T' 0+ | 99FE'F | GZ8 89°F | 8820 OF8PET' 0+ | O9T0'E | ShZ L 80PIT GS | 0G 1G oy! 1€80' OF SF90'0+ | 869° | 09¢ €1'c | Zv80 OF6EL0 0+ | FI8S'S | 82g 8 €L196°G | 02 0G €8°9 | F020 OFE6ZI'0+ | EIZh'E | S8FT €8°% | $0Z0'O-F0880'0+ | £99F'S | TEST L €900'S | GP 61 12°% | $120 OFTI60'0+ | ZEZE'E | GFET 9 ¥ 1220 OF L201 O+ | SFOE'S | 98ZT 6 6ItV v | CE 8T 29°F | 2420 OF8SZI'O— | 6299'E | GHB 0€'0 | 8920 0-F€800'0+ | 6FIF'S | 768 jd 89ES'F | LI 87% | 8080 09920 0— | STSS’€ | 699 €@'T 1Z80 OF S6E0 O— | OF9Z'S | IZ9 6 LIL8'V | 96 91 I'S | $620 OF LOST’ O+ | Zh8L'E | LEZ 021 620 OF SS80 0+ | F2LE°S | TFL 9 66909 | 81 ST 89'S | ZOFO OF LFZZ' 0+ | Z2ZI'F | O9E €Z°% |19F0 OFI9ZIO 0+ | FE08'S% | £62 9 €869'¢ | 9 ia! 88'S | 6040 OFFP0G O— | 92E8°F | ZIT €1°0 | S880 OF0GIO O— | 69%2'E | I8 g c08c GS |} 91 €I OL % | 20 OFZ66T' O— | SHST'P | OFF c6'S | &Sh0' OF869Z'O0— | GE8L'S | 892 g oSc8'G} OL rae G9 T 0120 OF LZFE0'0+ | 2809'S | ISFT €0°% | O1Z0'O0F9ZFO' 0+ | EIIS’S | ESF 9 9866'S | OF IT GLO | c&c0' 00800 O— | 1622 '€ | T6IT €¢ I 1660 OF 6FE0' O— | 9IE9'S | SECT 9 9829°¢ | Og OL LL 1920’ 0F¢860'0+ | O&10'F | 126 18°0 | 1220 OFO0TO' O— | O€S8"c | TZ8 v 6ZEV'S | 8z@ 6 49'0 | 6020’ OF FSTO' 0+ | L280'F | 6IZT PEF | 9820 OF9ZOT 0+ | 92LEL°% | ZEIT GI | 8r29'¢ | VE 8 78'& | LPIO OFG9CO'O+ | EZ8L°E | 896c 18'Z | 9F10' OF 8EE0' 0+ | OFLF'S | OL0E GStl'¢ | FL L GL'9 | €1Z0 OF 8EFI' 0+ | SOO F | 9SET 10°% | 6120 0Z690'0+ | 8002°2 | LZ&T 8 CE6F P| GE 9 GS IT | 9820 0-F862z'0+ | 8691'S | Z86 10°9 | 2920 OF I6ST' 0+ | S881'z | 168 L G09G'F | 96 g 99° 1Z | 1220 OF 68LF' 0+ | 2626's | LLL G6'FL | 67C0 OFEZLE'O+ | EFZS'S | TLL 9 8960 F | 92 v Z9'OL | 6020 OF SEF 0+ | 860° | 680T 18°8 | ZhZ0' 0-F9Z0Z'0+ | $960'T | 900T 9 €99F'€ | O€ € tr 9 1Z10' 0 6180°0+ | O8F6'T | 8268 92°F | E810 0F29G0'0+ | Ze9G'T | 98SE v 98¢6 I GL G S}PUSTIONJI0O UOIZBaII0D | suBey |s1equINyy S}PUBIOYZE00 UOT}BjaII0D | suBop | Sloquiny | ssuey isuvey |sisquInyy Pom oe a SS oe EHO Betts SUTLHDOVAG SNOS SLNGUVd ualpj)iya pun sjuasod fo suaqunu ay} fo sasonbs upaw ay} UO pasDpg S1oL1a a7qnqoid {a0n4 paiqui ay} fo suoynuiauab ay} 1]0 ur siajybnop pup suos ay} fo syuarayfa09 07 SLo11a fo soups pun siotsa ajqoqgoid ‘Buridsffo fiq syuasnd ‘syuarayfa09 uoynja.tioo ‘suaqunu ‘supa ‘spuasnd fo saqunu pup suvayy Tl G@TaViL 425 BRISTLE INHERITANCE IN DROSOPHILA *SJUSTIOYJOOO JO UOTZB[NI][VO VY} OJ SOI[IUIV] Moy OOT, z 8900’ 01220’ 0— 8910 OF S8ZLET O— OIFO OF ZFS O— 08€0' 0 01Z0' 0+ F0S0 OF EFL0' 0— 6840 OF 1160 0+ 0290 02660 0— 6070 0*F LETO 0— 6860 OF FZE0 O— €9F40 OF FIET O— Z640 OF 6SF0 0+ 0990 0F0E0T 0— £980 0 88¢0°0+ 0980 09290 0+ 9180 OF FESO 0— 8240 OF T190 O— F940 OFZIFO O— G8E0 OF LEST O— OFS0 OF Z69T 0+ 0080 0 2890 0+ 86£0 OF 8861 0+ 1980’ 0-F6990° 0-+ 8960 OF TSZT O—- ¢¢Z0° 0 8060° 0+ 6220 0-F E102 0+ GP8L € | Ec6ET 6206 € | S6IZ GSCE'P | PVE G60 SG | StF 0790 ¢ | 0Sz ISt6'P | 8c& G9ST 9 | C6L €068 °F | €8€ LLYI'G | SOL Z98E FP | 066 6069'S | 79z 068 F | 00G 6ELZT SG | €8P £692 °¢.| O6F 6c60'¢ | LE9 €8cL VP | LE 0202 'F | L6c IZIg vy | Ziv £669 FP | 80 6919'F | OTL IIé2'¢ | €Zé P60E 9 | S&P 92229 | 186 G8tP PF | 216 €ScE'P | ScIl 6900 012420 0— 9920 OF 9EFI O— £090 OF 6101 0— 9680 OF 80cI 'O0— TES0' 0F9S60 0— F670 0FZ000 0— 9020’ 0*FZES0' 0— 1070 0FZ90T 0— 9620 09980 0+- 92°90 0 €£00' 0+ F960 OFOTL0 0+ 8890 OF E081 0+ I8E0 OF FLZZT O— 290° OF 8F10 0+ OF€0' 0 8800° 0+ GcvO OF FSST O— 99F0 09200 0— 99€0' OF O10Z 0— 1S90°0*860' 0+ L120’ OF S08T “O+- GErO OF 6ST O+ ZOFO OF 1680’ 0+ 6220 OFZIOL O— F820 OFZ6IT’ O+ ZrO OF SIFT O+ 8699 °G | OIcET 966 OL6L°S | 69TG vy (6F WOIYBIOUS 104j8 popuedsns UOTzTVS |B) P9S0'€ | 8h 9 9g9e°2 | SI ChP8'& | 86E 9 LES6°¢ | 0& GLOL'& | GES L TE99°S | 02 S68E'°E | CIS q 1469°9 | 81 GZ86'€ | GLT Vv 918h'9 | GZ PCPS '€ | 68E 9 €9EL°9 | &Z 169° | O&L g Gc80°9 | oF TLFI'€ | 18% ¢ cross 9 | PI 8090 °F | 96¢ g Lg8h°9 | 81 GOTS'€ | IST € L919 9 | ¥E LSG9'€ | Lov q c98S°9 | OF 6S18€ | 68P v 69869 | 8@ 6602 °€ | SS¢ ¢ LELV'9 | 9E 6926 '€ | SPE 9 cs9y'9 | SI OLLE'E | 966 i] 6L469'°9 | .vI O190°€ | CrP v STOT 9°} SI 80FS € | 96T i 9013°9'| VE 0086 € | 808 89969 | 8F G9E0' F | 808 6 SP8c'°8 | FI LIVE F | S6E 6 L09L°9 | 06 9286 °€ | 208 Il | 6962°9 | PLL0'€ | SLL 6 9879 G | Ve 1Z06°@ | €S0I ¢ LEFT € | GZ ‘Sutidsyo Jo S1equInU 0} SUIPIODIB pazYysIoM 1 €¢ Gg 21G—-0¢ 67 LG 296-GE-¥FG EDWIN CARLETON MacDOWELL 426 cS 6V i VI [ lr | | = SNOL Ven GC NTE Ol Wy TU SYSLHONVG KX SLNAYVd SUN SHOVEESO®) NOILVTSYYOD oct Oe'+ Ov'+ Os'+ Saivos H@® SINSIS/IBasO9 427 BRISTLE INHERITANCE IN DROSOPHILA snjd a}¥voIpul oul] seq oY} 9AOGR sIBg cS 6P Sv “{UBOIUDIS AT[BOI4ST}B4S jou 218 uoTZIOd YoVl[q OU Y}IM sIVq {10110 eTqeqoid syI seuITy 90144 Spo00xe JUOTOYJI0 ay} JUNOUIe OY} O}BOIPUT SuBq oY} JO SUOTJIOd Yo]. prog al SNOS X SLN3YVd SINS Ss3a05D NOILV1TSHxyOD ‘s]USTOWJe09 snutul ‘MoyTaq ssoy} ‘syUsTOyJe0o ‘squoied Aq suos ‘g ‘syueied Aq Ssioqysnep 1OJ syUsTOyJooo UOTYBleLIOD ‘p | “Shy OCs 0) mu Oe m a nSNOILVYANSD i Ol S 00 O ill ea eS 6 Ol'+ m 7 a fe) mM oe | W fo) ~) a 428 EDWIN CARLETON MacDOWELL In figures 1, A and B, the correlation coefficients for the sons and daughters are presented graphicly. Bars above the baseline indicate plus coefficients, those below, minus coefficients. The breaks in the base line call attention to the generation omitted because they included too few families for the calculations of the coefficients. The solid black portions of the bars indicate the amount the coefficient exceeds three times its probable error; accordingly, bars with no black portion represent coefficients that are not statistically significant. Before discussing the actual findings, the manner of calculating the probable errors should be explained. Clearly probable errors have the greatest importance in evaluating any statistical result. In calculating correlation coefficients involving few parents and many offspring, the numbers of parents are weighted according to the numbers of their offspring. However, there are fewer parents, and the probable errors based on the numbers of offspring are therefore smaller than the facts would justify. If, on the other hand, the actual number of parents is used, the errors will be far too large. Sturtevant (’18, p. 10) has stated the case clearly; he gives the errors based upon the numbers of offspring, but does not consider that his correlation coefficients have much signifi- cance. In the present case the errors have been based on the mean squares of the numbers of parents and offspring: a 2 in which n, is the number of parents and nz, the number of offspring. This of course does not solve the difficulty, but the errors seem to be more reasonable than those obtained by other methods. The errors by both other methods have been calculated; the ones given are not enough greater than those based on the numbers of offspring to change the general appear- ance of the charts; the errors based on the numbers of parents considerably reduce the number of significant coefficients, but the same conclusions are to be drawn whatever set of errors is used. Figure 2, showing the empirical means and the regres- sion straight lines, should be studied in connection with these ND EMPIRICAL MEANS FOR EACH GENERATION eles REGRESSION STRAIGHT LINES a es) pee ’ MALES FEMALES i as Bee Pete Peer BENET TTR ns = \) li] ia a c Slips NY Xl ARC i 5 ¢ i bh m\\ mmcamys : See PSE 7 ——_e a ase | pet tte 4 INU VA elt a Ss ey Hh 2 Ce ia fa F *