pherdtd prep ese. 3 is ; re aay : 3 fi o bee ie : sis, ik ~ eg " : a is . ap 4 oy esse ye : oS eg rete ws Crt, / Redetes SEE ae he = ata ke Eee seats G - arate ‘ artis : ’ : yes z: ists zi ess or - es 5 A est ° ot Gy: 2434 “ey it * 33 i iste : sia cde . stats spacasss : : ier, preie ies Hiethis tore Sages paster are Pat eab Nees * é; f ; ie 1 : ; 2 Let =)" - . r : ts) oe * pure res ; : 5 ey? repent bereys+ as te a gre eree - ate wd Cael 4 uy eet EES = = = Ee a Wi ASE Sink Rae ef reeset eta een sy! Ne i! ira betes SS _ ii We ( * THE JOURNAL OF EXPERIMENTAL ZOOLOGY EDITED BY Wi.iiaMm E.. CastLE FRANK R. LILuie Harvard University University of Chicugo Epwin G. ConxKLIN JacquEs LorB Princeton University Rockefeller Institute CHARLES B. DAVENPORT Tuomas H. MorGan Carnegie Institution Columbia University HORACE JAYNE GrorcE H. PARKER The Wistar Institute Harvard University HERBERT 8. JENNINGS Epmunp B. WILson, Johns Hopkins University Columbia University and Ross G. HARRISON, Yale University Managing Editor VOLUME 13 1912 THE WISTAR INSTITUTE OF ANATOMY AND BIOLOGY PHILADELPHIA, PA. Ge wad COMPOSED AND PRINTED AT THE WAVERLY PRESS By tee WiniramMs & WiLKINS COMPANY BautTimore, U.S. A. CONTENTS 1912 Nos l JUL C. B. Davenport. Sex-limited inheritance in poultry. Eight colored DERE eee ees ee ae eee ee ERR Acre ails Sch yom. f 2 ibis Socks 6 27 GrorGce ALFRED BarITsELL. Experiments on the reproduction of the hypo- trichous Infusoria. I. Conjugation between closely related individuals of Stylonychia pustulata. Twenty figures (one plate)................. 47 T. H. Moraan AND Evetn Carrett. Data for the study of sex-linked in- hermtance mUDrosophila. -: ws. os os -c s eas ome oe eee ie eee 79 C.M. Cuttp. Studies on the dynamics of morphogenesis and inheritance in experimental reproduction. IV. Certain dynamic factors in the regulatory morphogenesis of Planaria dorotocephala in relation to the axial gradient. Forty-six figures....... 02... 002- 8.10302 se seen nn heen 103 No. 2 AUGUST 20 2 Raymonp Peart. The mode of inheritance of fecundity in the domestic fowlilbaree eT OUTEShi..55 5.1.9 cas - sero ce hase hohe alates theme ta ne cert een 153 W. C. Auten. An experimental analysis of the relation between physiologi- cal states and rheotaxis in Isopoda. Ten figures..................+..-. 269 ii oe lv CONTENTS No. 3 OCTOBER 5 Epmunp B. Witson. Studies on chromosomes. VIII. Observations on the maturation-phenomena in certain Hemiptera and other forms, with considerations on synapsis and reduction. Nine plates..... eee bite ols 345 WayLanp M. Cuester. Wound closure and polarity in the tentacle of Metridium marginatum. Wight figures:...---¢.2- 0.25. -.9.4.- 3-2 eee 451 Max Morse. Artificial parthenogenesis and hybridization in the eggs of Centaintinvertebrates:..s-. Gee eee Reece weir oer eee rere 471 No. 4 NOVEMBER 20 Epwarp B. Metes. Contributions to the general physiology of smooth and striated muscle. ‘Twentby, te uress sees ayes te rete y epee nee bal eral Tete eetete les arene 497 A. R. Moors. Concerning negative phototropism in Daphnia pulex. One Jacours Lozs. The comparative efficiency of weak and strong bases in artificial parthenogenesis wi. <5 e2 ceente ete ests All males striped and black-red; all 92 9 un-striped and brown-red 3 F, (black-red o X brown-red@) X_ black-red 2 ‘Gametes {s : \s Fi St) Ss s All males and 50 per cent femalesstriped; others unstriped OP C. B. DAVENPORT results of this mating are shown in table 6. The numbers of offspring obtained were small. To the right is given the approx- imate numerical result. 4. THE NEW CASE OF A SEX-LIMITED CHARACTER IN POULTRY 1. Material employed In my experiments I used poultry belonging to the Brown Leg- horn and the Dark Brahma races—two races that have very dis- similar plumage colorations and exhibit a sex-dimorphism in pattern and coloration. They are, consequently, well adapted to testing again the behavior of sex-limited characters in fowl. The points of difference considered in the two races are indicated in table 7 both for the males and the females. TABLE 6 osfos er) REMARKS 1 White Rock < Brown Leghorn Gametes te Le Be 0 F, BeLC BCe Both sexes barred; the males, only, - splashed with Brown Leghorn- 2 Brown Leghorn x White Rock fL B Gametes ‘ae Fy BL L The males only barred; the femalés either black with orange- laced hackle or ap- proximately Brown Leghorn 3 White Rock < Fi(White Rock @ Brown Leghorn 9 Gaetan ie BC or Be Be —cor —C Offspring ieee Ree : BBCe BCe Half oo white; half barred; 92@¢@_ half white, half barred SEX-LIMITED INHERITANCE IN POULTRY 11 TABLE 7 Sex-dimor phic racial characteristics | CHARACTERS BROWN LEGHORN DARK BRAHMA | A. Male (Figs. 1 and 2) Lacing (marginal coloration) of, hackle and saddle.......... | Red White (dominant) Upper wing coverts (wing bow) Red White, trace of red in | middle zone B. Female (Figs. 3 and 4) Wacimiorothacklers es. eens | Golden yellow Silver white Back and upper wing......... | Light brown, finely stippled with darker, Gray, with three con- centric loops of black | (penciling) 2. Matings and results The F, generation. In 1910 I mated (Pen 1009) a Brown Leg- horn cock? (No. 14123) to various Dark Brahma hens that had _ been bred by me and were descended from Nos. 121 ¢ and 122¢ described by me in a former publication (Davenport, ’06). The following mothers produced offspring that grew to maturity so that their permanent plumage could be described: Nos. 5835, 7549, 7859, 7869, 8001, 11160. Moreover, I mated in Pen 1015 a Dark Brahma cock (No. 11161) to a Brown Leghorn hen of the same origin as the cock of Pen 1009. About fifty-four chicks were reared from these two pens and their adult plumage color studied. The distribution of plumage colors in the two sexes and the two sets of experiments is set forth in table 8. Table 8 gives a definite answer to the question of the method of inheritance of the characteristics considered. In the reciprocal crosses of Pens 1009 and 1015 the 27 males are all alike; but the females differ according as the Brown Leghorn or the Dark Brahma is used as father and in their hackle color they resemble, in both cases, the father (figs. 5 and 6, 7 and 8). 2 Bred by and purchased from H. W. Smith, Islip, Long Island, N. Y. 12 C. B. DAVENPORT - TABLE 8 Frequency of specified characters in the offspring in the given matings | MatInG Matine 1009. Fi (Brown Lecuorn o’ X Dark BrauMa 9) | 1015. Fy (Dark Basses x MOTHERS’ NUMBERS OES CHARACTERISTICS Suir | TOTAL | LEGHORN) 5837 | 7549 | 7859 | 7869 | 8001 | 11160 | 14129 Males with lacing white. a2 9 | fh SA ORD 2 1 This | 2+ Males with lacing red....| 0 Of @ 0 0 0 0 0 Males with wing bar white) 0 0 0 0 0 0 0 0 Males with wingbarred...| 2 9 a 2 2 1 BL | 6 Females with lacing white| 0 0 0 0 0 0 0 4 Females withlacing golden 2 if 3 i 2 4 23 0 | The male offspring, although all remarkably alike, are like the males of neither of the races involved: for they have the white lacing that is characteristic of the Dark Brahma and the red upper wing coverts that are characteristic of the Brown Leghorn— thus their characteristics are clearly derived from the germ-plasm of both parents. Can the same be said of the characteristics of the female offspring? So far as concerns those characters that are not sexually dimorphic such inheritance from both parental germ-plasms is clear, for the pullets of the reciprocal crosses are alike in having pea combs instead of single combs and in having slight in place of heavy booting. Ordinary somatic characters follow the ordinary laws of equivalence in reciprocal crosses. Even in the sexually dimorphic character of pattern of the feathers of the breast, back, and upper wing there is only a slight difference in the reciprocal crosses. Thus, when the Dark Brahma is taken as father, while the pullets (fig. 8) all have the gray back- ground on the feather correlated with the white hackle, the broad concentric loops, on the contrary, shown in fig. 4, are replaced by finer loops which are more or less discontinuous and show a transition to the condition of stippling characteristic of the Brown Leghorn hen. The pattern is quite the same in the pullets (fig. 7) derived from the Brown Leghorn father. The pattern is truly intermediate. Also, in both crosses, the amount of red on the wing in the pullets is intermediate (figs. 7, 8) although SEX-LIMITED INHERITANCE IN POULTRY 13 the general ground color is determined almost exclusively by the father’s germ plasm. The shafting, likewise, is intermediate, inasmuch as it is sometimes absent (though usually present to an intermediate degree) in pullets from a Brown Leghorn sire and is usually absent, but sometimes (as in fig. 7) shows, in pullets from a Dark Brahma sire. Even though we make every allowance for the circumstance that the pullets figured are only about eight months old—an age at which the sharp contrasts of the adults are lacking—still the conclusion can not, I think, be avoided, that not all sex-dimorphic characters are sex-limited. The F, generation. In the spring of the following year (1911) I mated together the (Brown Leghorn ¢ X Dark Brahma 9?) hybrids in Pens 1112 and 1115; and the (Dark Brahma ¢ xX Brown Leghorn ¢) hybrids were mated in Pen 1126. The results from these two sets of matings are given in table 9. Expectation in the mating of a male F; (Brown Leghorn 7 X Dark Brahma ¢) by a female of the same origin may easily be deduced by the formulae used in the earlier part of this paper. Table 10 gives the formula for inheritance of lacing; W, white and w, absence of white. | This formula indicates that half of the father’s sperm have and half lack the W factor (inhibitor of formation of red pigment). By hypothesis only half of the eggs have the sex-chromosomes and these eggs all carry absence of white (i.e. golden) hackle. Con- sequently, white x golden (= white) and no white, or red, X golden (= red) combinations for hackle color should be equally common in the males; actually there are eight white hackles to four red. Though the numbers are small and the accord is pro- portionately not close yet the absolute departure from equality is small and the result is: of the expected order—both types of hackle color occur in approximately the ratio of 1:1. Likewise, since, by hypothesis, the sex-limited characters of the pullets are derived solely from the sperm there will be two sorts of pullets, equally numerous, viz., white-laced and golden-laced. Actually these occur in the proportions of 10 to 7. Again the proportions are of the expected order. Cc. B. DAVENPORT 14 ‘OOUVUIUMIOP JOoj1od UIT 07 ONp ssayyqnop ‘Ivq SUIM UO , pot O[}}I] B ,, YJIM OS¥O OU SopNyoUT y G G 0 G T I I 0 I I I! 0 0 (0) 0 G 0 G | oeser | coset TIVLOL ' j SUdH#WOAN | SUGHLOW | 6 NuoHoayT NMOUg X . VWHVUG MUVC WW “YSTT ONILV]A | OT al | 1Vaou G 0 I 0 | | ae LEFT SOT9T | G&PST SYUMAGNON SSUAHLOW 0 | 62PST | 9OLFT | 0 0 YO NYOHDAT NMOUET 6 VWHVUg MUVq X Tat ‘CIT ONILVW I #6 a 6 WIGVL GILTT | OF89T ONYOHDATNMOUG “WT “ZI[] ONILVIN Lg! Te OS 0 SOGIT | PLIOT | FSPST 610ST SUGHNON SUAHLOW 6 VNHVUgG MHVC X SLIJIDIDYI pIzVUr]-Las ay? [O WOUMLaUaB &@ yy ay] UWL UWOLTWNGLLYSYP AY, ], “**"“Weplos Sut -OB] YJIM Soreua,y OO OO “OVI M SUT -OB] YYIA Soyvulo gy Veseess par aeq -SUIM YIM So[vy “on Tyan req -SUIM YJIM So[vyy Ria intr orale “pod SULOV] YYIM Sov ee oat SULOV] YIM Sole yy SOILSINALOVUVHO SEX-LIMITED INHERITANCE IN POULTRY 15 TABLE 10 F, (Brown Leghorn & X Dark Brahma 2 ) X F, (Brown Leghorn 3 X Dark Brahma 9) Inheritance of hackle lacing Gametes in m \w F {Ww W 50 per cent of the : \ ww w males with white hackle; 50 per cent with red; same with the females Returning to the males, the inheritance of the color of the wing- bar has to be considered. By hypothesis, the sperms of the hy- brids carry in equal numbers, red-bar and no réd (white) bar: while the eggs all carry red-bar. Two kinds of males with the same external appearance are, accordingly, to be expected; namely, duplex and simplex red-bar. All offspring recorded had red wing-bar, but in some cases less strongly developed than in the pure Brown Leghorn race. In all cases, then, the results agree with hypothesis. The matings of hybrids of the reciprocal cross (viz., Dark | Brahma ¢ and Brown Leghorn ¢) yielded few offspring, only two ‘males and four females; but the results are rot without interest. By hypothesis, in hybrid males half of the sperms have the deter- miner for white lacing and the remainder for red lacing; also half of the sperms have the determiner for red wing-bar and half lack it. All hybrid pullets carry the determiner for white lacing in half of their eggs and for white wing-bar in half of their eggs. It follows that of male offspring all should have white lacing, and half should have red and half white wing-bar. Of the female offspring, on the other hand, half should have red and half white lacing. The formulae are as given in table 11. Actually each class is represented in exactly the expected proportions. 16 Cc. B. DAVENPORT TABLE 11 F, (Dark Brahma 3 X Brown Leghorn 2) X F, (Dark Brahma # X Brown Leghorn 2 ) For hackle lacing, W estou ge REMARKS Gametes....... { iB Mi F {| WW W PE hiya et ae \ Ww W Expectation; all oo and 50 per cent females with white lacing; 50 per cent 2 2 with red lacing For wing bar R, evident in males only Gametes....... { * ‘i F if Rr | Ome Soe \ rr | | Expectation; 50 per cent males with red wing bar and 50 per cent with white CONCLUSIONS The foregoing observations thus accord with the hypothesis that the male carries two sex-chromosomes and the female one; and that the determiners for certain secondary sex-characters are centered in the sex-chromosomes. We have still to consider the question: Are all sex-limited characters carried in the same chromosome? This does not imply that the sex-limited charac- ters of one variety shall always appear together. This isobviously opposed to the facts, for in the first hybrid generation a white hackle (Dark Brahma) and red wing-bar (Brown Leghorn) appear together; and must always do so when the determiners for these dominant traits are in the same zygote. In F, from Brown Leg- horn grandfathers two kinds of males are possible on the hy- pothesis that the two sex-limited determiners are united, as in the same chromosome: namely, with both red hackle and red wing bar (consequently, quite like the pure bred Brown Leghorn) and with red hackle and white wing-bar (the hybrid type) ; and these, SEX-LIMITED INHERITANCE IN POULTRY 17 and they only, were realized. Thus the combination white hackle—white wing-bar (pure Dark Brahma type) did not occur. In F, from Dark Brahma grandfathers two kinds of males are possible, viz., those like the pure bred Dark Brahma and hybrids. Of the two males that matured one belonged to the Dark Brahma type and one to the hybrid type but the Brown Leghorn combi- nation did not reappear. Both sets of experiments thus speak strongly for the conclusion that all sex-limited characters are linked together in the sex-chromosome. Finally, our study throws light on the question whether all secondary sex-characters are sex-limited—in the sense of being carried in the sex-chromosome. For if they are we should expect the female hybrid to inherit such characters from her father’s race only. But as we have seen, this is not the case. The de- tails of penciling, stippling, and shafting show clear evidence of blending of conditions from both parental races, in the first hy- brid generation. Figs. 7 and 8 show how like the feather pattern is in the pullets of reciprocal crosses. This leads us, provision- ally, to classify sex-dimorphic characters into two classes: a, characters whose development is controlled primarily by deter- miners located in. the sex-chromosomes and, b, characters whose development is specially influenced or modified, probably by secretions of the sex-glands. January 19, 1912 LITERATURE CITED Bateson, W. 1909 Mendel’s principles of heredity. Cambridge, England, 396 pp. Davenport, C. B. 1911 Another case of sex-limited heredity in poultry. Proc. Soc. for Exper. Biology and Medicine, vol. 9, pp. 19, 20. Dec. 20, 1911. DoncastER, L. anpD Raynor, G. H. 1906 Breeding experiments with Lepidop- tera. Proc. Zool. Soc., London, vol. 1, p. 125. DuruaM, F. M. anp Marryat, D. C. E. 1909 Inheritance of sex in canaries. Rept. Evol. Com., vol. 4, pp. 57-60. GoopaLr, H. D. 1909 Sex and its relation to the barring factor in poultry. Science, vol. 29, 1004, 1005, June 25. THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 13, NO. 1 18 Cc. B. DAVENPORT GoopaLE, H. D. 1910 Breeding experiments in poultry. Proc. Soc. for Exper. Biology and Medicine, vol. 7, pp. 178-179. 1911 Sex-limited inheritance and sexual dimorphism in poultry. Sci- ence, vol. 33, pp. 939-940, June 16. Guyer, M. F. 1909 The spermatogenesis of the domestic chicken (Gallus dom.). Anat. Anz., Bd. 34, pp. 573-580; 2 pls. Hagepoorn, A. L. 1909 Mendelian inheritance of sex. Arch. f. Entw.-Mech. d. Org. Bd. 28, pp. 1-34. Morean, T. H. 1910 The method of inheritance of two sex-limited characters in the same animal. Proc. Soc. Exper. Biol. and Medicine, vol. 8, pp. 17-19, October 19. 1911 The application of the conception of pure lines to sex-limited inheritance and to sexual dimorphism. Amer. Nat., vol. 45, pp. 65-78, February. 1911la An attempt to analyze the constitution of the chromosomes on the basis of sex-limited inheritance in Drosophila. Jour. Exp. Zodl., vol. 11, pp. 365-412. One plate. PEARL, R. anp Surrace, F.M. 1910 Studies on hybrid poultry. Ann. Report, Maine Agric. Exper. Station; pp. 84-116; May. 1910a On the inheritance of the barred color pattern in poultry. Arch. f. Entw-mech. d. Org., Bd. 30, 1 Th., pp. 45- 61. 1910b Further data regarding the sex limited inheritance of the barred color pattern in poultry. Science, vol. 32, pp. 870-874, December. PuNNETT, R. C. anp Bateson, W. 1908 The heredity of sex. Science, N. S., vol. 27, pp. 785-787, May 15. Sprutman, W. J. 1909 Spurious allelomorphism: results of recent investigations. Amer. Nat., vol. 42, pp. 610-615. 1909 Barring in barred Plymouth Rocks. Poultry, vol. 5, nos. 7, 8. Srurtevant, A. H. 1911 Another sex-limited character in fowls. Science, vol. 33, pp. 337-338, March 3. WILson, E. B. 1911 Studies in chromosomes, VII. A review of the chromo- somes of Nezara; with some more general considerations. Jour Morph., vol. 22, pp. 71-110, March. 1911 a The sex chromosomes. Arch. f. mikros. Anat., Bd. 77, pp. 249-271. SEX LIMITED INHERITANCE IN POULTRY PLATE 1 Cc. B. DAVENPORT 1 Male Brown Leghorn 19 THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 13, NO, 1 PLATE 2 SEX LIMITED INHERITANCE IN POULTRY Cc. B. DAVENPORT 2 Male Dark Brahma; above and at right a feather from the fifth, the third, and the first row of wing coverts 20 SEX LIMITED INHERITANCE IN POULTRY PLATE 8 C B. DAVENPORT? 3 Female Jungle Fowl to show type of coloration of female Brown Leghorn 21 PLATE 4 SEX LIMITED INHERITANCE IN POULTRY Cc. B. DAVENPORT 4 Female Dark Brahma te ine SEX LIMITED INHERITANCE IN POULTRY PLATE 5 Cc. B. DAVENPORT 5 Fi male hybrid between Dark Brahma and Brown Leghorn¢ 23 PL ATE SEX LIMITED INHERITANCE IN POULTRY Cc. B. DAVENPORT 6 F; male hybrid between Brown Leghorn 7 and Dark Brahmag 24 SEX LIMITED INHERITANCE IN POULTRY PLATE 7 Cc. B. DAVENPORT 7 F, female hybrid between Brown Leghorn 4 and Dark Brahmag 25 PLATE 8 SEX LIMITED INHERITANCE IN POULTRY Cc. B. DAVENPORT 8 Fi female hybrid between Dark Brahma # and Brown Leghorng 26 HEREDITY OF BODY COLOR IN DROSOPHILA T. H. MORGAN From the Zoélogical Laboratory, Columbia University FOUR COLORED FIGURES—PLATE | Cultures of the fruit fly, Drosophila ampelophila, have given rise to three mutations in the color of the body and wings. The origin of these new types has been briefly described in a pre- liminary note, and some of the main facts connected with their inheritance have been given there, but the principal data on which the statements rest have been reserved until the present time. The results include 81,070 counts. It may be asked what advantage is there in doing the experiments on so large a scale. Why would not a few cases with suitable tests show the mode of inheritance of the factors involved? The answer is two-fold. The question of the relative viability can only be determined in this way, and for further work with these body colors it is necessary to know what réle this condition plays in the numerical results. In the second place it seemed worth while to illustrate on a large scale the phenomenon of sex-linked inheritance. It is an impressive fact, for instance, to find in the F, generation (out of black female by brown male) 6124 black females, 3015 black males and 2472 brown males. Not a single brown female in 11,000 grandchildren. While in the reciprocal cross there are present 2191 black females, 1987 black males, 1532 brown females and 1448 brown males. Such results cannot fail, I think, to impress those who take a sceptical atti- tude toward the modern study of heredity. The black and the yellow mutants arose directly and inde- pendently from inbred, normally colored, or gray flies. The brown flies were produced by crossing and extracting. They also arose independently in cultures related to the black flies 27 28 T. H. MORGAN but were at first supposed to be a particular kind of yellow fly which was recognized, however, as different from the ordinary yellow and had been maintained in a separate culture. When, however, in the second generation from black and yellow these same flies appeared, their relation to the other colors was appar- ent. Forms like these, that represent a type due to absences, derivable through combination of primary mutations may be said to arise by permutation. In making the counts I have been assisted by Miss E. M. Wallace, Miss Eleth Cattell and Mr. C. B. Bridges. In practi- cally all cases I have checked up each count after the separation had been made, so that the results stand for the agreement of two observers. Only in the case of separation of yellows and browns could any disagreement arise, and while I cannot claim for this case that the separation has been exact, I think that it is very nearly so. Description of the mutants The color differences between the normal, wild, or gray fly and the mutants is shown by the accompanying plate, figures 1 to 4. The detailed comparisons are as follows: The wild fly. The upper surface of the thorax is olive yellow, the olive shade being very faint. As the flies get older the-color deepens. Some of the wild flies have a black trident and two lateral black streaks on the upper surface of the thorax (not shown in the present case), but many flies do not show this marking. J have made a long series of selection experiments with this marking and have produced one race that never shows the trident, and another race that has a dark well-developed trident. How far the character is a fluctuating one and how far due to genotypic difference need not be discussed here. The abdomen of the female is banded with lemon yellow and black. In the male there are only two black bands as a rule; the end of the abdomen is black. The legs are colored like the thorax but somewhat lighter. The hairs on the body are black. The wings are very transparent, blue gray or smoky. The veins appear dark, but under the microscope are seen to be HEREDITY OF BODY COLOR IN DROSOPHILA 29 brownish yellow. There are no dark bands along the sides of the veins. The upper surface of the head of the wild fly, and also of the mutants is colored like the thorax. 'The under surface of the abdomen is more yellow in the male, than in the female. The yellow fly. The upper surface of the thorax is yellow ochre in color and lighter than that of the wild fly. It is inter- esting to observe that the dark marking or shield is always absent from the yellow flies. The light bands on the abdomen are of the same color as the thorax, 1. e., pure yellow ochre, and lighter than those of the wild fly. The dark bands are brown. The legs are the same color as the thorax. The veins of the wings are yellow like the thorax. The interspaces (or background) are of a transparent golden yellow and strongly contrast with the color of the wings of the wild fly. The hairs are brown, instead of black as in the wild fly. The black fly. The upper surface of the thorax is the same general color as that of the wild fly, but darker in the sense of being browner. The black trident is always present and con- spicuous, and the two lateral markings are also conspicuous. The trident is not only well developed, but appears to be longer and narrower than that of the normal. It is one of the most striking features of the black mutant. The light bands of the abdomen are darker than those of the normal, but not so dark as is the thorax. The dark bands are very black. The legs are blacker than the legs of the wild fly, especially the more distal parts. The hairs on the body are black. The veins of the wings are very black. On each side of each vein there is a dark (semitransparent) band. The interspaces between these bands are gray, but darker than the gray of the wing of the wild fly. While the black fly is blacker than the wild fly in nearly all of its parts, so that a heap of them is very dark compared with a heap of the normal flies, the most striking character that dis- tinguishes them from the other types is the dark wings with the dark bands on the sides of the veins. The brown fly. The upper surface of the thorax is brown. The brown color deepens as the fly gets older, and the color THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 13, No. 1 30 T. H. MORGAN shown in the figure is that of an older fly. When first hatched the brown is more yellowish in shade, and not so easily dis- tinguished from that of the older yellow flies. Most of the brown flies, especially the older ones, show a brown shield on the thorax, in form like that on the thorax of the black fly. The light bands of the abdomen are light brown and differ therefore from the yellow bands of the yellow fly. The dark bands are, brown. The legs alsoshow some brown. ‘The veins of the wings are brown like the thorax. On each side of each vein is a brown band. It is by means of these bands that one can most readily separate under a lens the yellow from the brown fly. The interspaces between the bands are gray-brown, and more trans- parent than the bands. The hairs are also brown. In separating the flies into the color groups, (after slightly etherizing) there is never any difficulty in distinguishing the gray from the black and from the yellow, provided they have not just hatched; but the yellow flies and the brown offer greater difficulties, especially when flies of different age are mixed to- gether, and when small flies are also present, due to starving the larvae. I do not feel certain that the separation of these two groups has been always perfect, but the errors are not great enough I think to vitiate seriously the classification. In some cases I have kept the flies alive for several days in order to verify my first separation, and have found occasionally that one or two flies have been put into the wrong group. The errors have been in both directions, and counterbalance to some extent. I have followed the rule of classifiying flies as brown only when they were certainly brown, as shown best by the broad brown bands along the veins of the wings, and this has lead, I fear, to the inclusion of a few brown flies in the yellow group. I have also tested my ability to separate these two groups by breeding doubtful flies, which, in general, I would have placed in the yellow category. If the fly in question is a yellow female and is bred to a black male, all the female offspring should be gray (because only the female producing sperm carries the factor for black) and all the males should be yellow. If the fly in question were a brown female, and were bred to a black male, HEREDITY OF BODY COLOR IN DROSOPHILA 31 all the female offspring should be black and all the males brown. In general I have found my separation to be correct. The difficulty has arisen apparently in most cases with heterozygous yellow females that contain only one dose of yellow instead of two, as does the ordinary pure yellow. In such cases two classes of females appear when crossed to black males, namely, gray and black, In regard to the distribution in the wild fly of the products of the three color factors, that go to produce its color, it is diffi- cult to speak with certainty, but from comparison with corre- sponding regions in the mutants when one or another of the color factors is absent it appears that the black regions are due to the black factor, but the brown may be present and overlaid by the black. At least it may be said that the black regions in the gray fly and in the black fly are brown in the brown fly, but of course it is possible that when the black develops the brown may not develop, or the black may even be a further stage of development of the brown pigment. The yellow of the wild fly also seems to replace the brown of the brown fly at least when yellow is absent the color of the yellow regions is brown. Possibly, as I have suggested, yellow when present inhibits brown, for otherwise it is difficult to see how the yellow fly should be lighter in color than the brown fly. The black flies are large and vigorous. There is no difficulty in breeding them or in crossing. The yellows are generally smaller (though not always) and are more delicate. They get stuck very easily to the moist sides of the culture bottles, and, being unable to free themselves, perish. They are more difficult to breed and to cross. The brown flies although generally large are weak. They get stuck to the food and to the walls of the bottles and die. Otherwise they seem healthy. On the whole the mutants are weaker than the normal flies, but the loss of the yellow factor that produces the black flies is less injurious then the loss of the black factor that produces the yellow fly. I have at times thought that the loss of both of the factors produces the weakest fly in the series—the brown fly— but this is difficult to prove. 32 T. H. MORGAN Formulas The color of the wild fly appears to be due to the presence of three factors, Black (B), Yellow (Y). and Brown(Br). For brevity this color is spoken of as Gray, which corresponds nearly to the color of the semi-transparent wings. If the black factor (B) is absent (b) the color of the fly is yellow (Y), more especi- ally the wings. The yellow fly is therefore bYBr. Where the factor for yellow (Y) is absent (y) the fly is black, more especially the wings. The black fly is therefore ByBr. When both black and yellow are absent the fly is brown, more espe- cially the wings. The brown fly is, therefore, byBr. The brown fly can always be produced by crossing yellow and black and inbreeding the F,’s which give by recombination some Browns in the second, F:, generation. Of course, the same result would follow if both yellow and black were lost from the gray fly at the same time, but this is unlikely since the black and the yellow factors lie in different parts of the hereditary complex. Of these three color factors that of black is sex-linked; the yellow factor is not sex-linked, and is contained in all gametes both in the male and in the female of gray and yellow flies. One must be careful to observe that while the factor for black is sex-linked the black fly, bred to gray does not show sex- linked (sex-lmited) inheritance; while the yellow fly bred to normal shows sex-linked inheritance. This will be clear from an examination of the analyses given below. All possible crosses have been made between these mutants, and, these may now be taken up in order. Wild (gray) by black When the female wild flies (Gray) are mated to male black flies all of the offspring are gray. These F, gray flies are darker than the wild flies, i. e., they are to some extent intermediate in color between gray and black. It is true that there is much variation in these hybrids, and some flies can not always with certainty be distinguished from wild flies, but most of them are undoubtedly distinctly darker, especially the wings. How this HEREDITY OF BODY COLOR IN DROSOPHILA 33 could occur may not appear clear at first sight, for both the wild gray females and the gray-black hybrids contain two doses of black—the only difference between the two is the absence of one dose of yellow, in the hybrids. If, in the absence’ of this dose of yellow, the black has a better chance to show itself more positively, we can account for the intermediate char- acter of the hybrid. If such is the case,the yellow factor is to some extent a partial inhibitor of black. The same explanation applied to the males is as follows: the wild gray male has only one dose of black (since black is sex-linked). It has two doses of yellow. The hybrid has also one dose of black (B) but only one dose of yellow. It differs, therefore, from the wild male in having one dose of yellow instead of two. The darker color of the hybrid male would, in consequence, be due to relatively less yellow than that present in the wild male. The explanation is the same therefore for both sexes, but it involves the assumption that the color of the wild female which is the same as that of the male is due to the presence of a double dose of B and Y(BY, BY) while the color of the wild male is due to one dose of black and two of yellow (BY, Y). Removal of one Y from the wild female makes her darker; and similarly the removal of one Y from the wild male makes him darker also. The numerical results for the F, generation are as follows: Gray: Oc. 2a ee eee ego eeees O5I é ; skeen Graiy Usha ag ae ee eee 4861 a eae GMatr. «| Black’) 9a! 4. ee 1280 (Black: -o?s:-4 <4 4 Chae eens alps eee 1385 The expectation, as shown by the analysis below, calls for three grays to one black. There are 9914 grays and 2665 blacks. The blacks fall considerably bélow expectation, yet the black flies are a vigorous strain and appear in the cultures to breed as well as the grays. It will be noted that while the gray males run some 200 flies behind the females, the black males exceed the females by 100 flies. Since these counts are from more complex crosses involving white eyes and short wings as well as red eyes and long wings, 34 T. H. MORGAN and since the former characters are associated at times with diminishing returns, I give in the next table some results where only black and gray color are involved: (Grey? Oi eeg oa a Ae . ee 867 oes Bs GOs, of SGV) Cte: el oir eee en an 5 See eee 811 Ce OY RIC Net art BUC ots dhe enn yy oe ae 201 (Bele tot Sih ahs Pe Oe ea re eee 180 There are 1678 grays and 381 blacks. The ratio is approxi- mately four to one which is not very different from the pre- ceding ratio. The analysis of this cross is as follows: Gray @ BYBrX — BYBrX Black & ByBrX — byBr Gray ° BYBrX — ByBrX Gray o' BYBrX — byBr BYBrX — ByBrA BYBrX — ByBrX — byBr — bYBr BR, Gametes of F, F, Gray 2 3 Black ¢ 1 Gray o' 3 Black o 1 The reciprocal cross, gray male by black female gave the following results: GRAY; 2 Qa. eile 7 a ane 1465 a Gray ©. 2 Graya opens. 2 chao. 1a oe Cee 1453 Bo PEGE! See o VeBlnckes Of byte. te ee 344 WEBlacle sop aie GAs iy Aan ee ae 299 The expectation is here also three grays to one black. The total for the grays is 2918, and for the blacks 643, which is about 4i to 1. In the grays, the males and females are nearly equal, while in the blacks, the females exceed the males by a fair mar- gin. Since these numbers also are derived from mixed counts (as before) I give below a count involving only the two colors in question: RG CMOR f.) Stan os oe Os a ash Wee 836 Gray 92 ROTEL Y SRG ERs. Y calowic,s 2 cus ee eee 875 Perey: Cis ee oH BIR Cho eee 23 ST. 2 eee ee 209 Fi: 2) AG ae EE ROME ke a 8 148 HEREDITY OF BODY COLOR IN DROSOPHILA 35 There are 1711 grays and 357 blacks which is very nearly 5 to Ie Wild (gray) by yellow The results of this cross have been already published (1911), but may be given here for the sake of completeness. In both cases the parents were alike except for body color: iS Grayvee Oe ers Se een aes 525 IVE r 6: ( GebyYcv= eee oes =o eave Sate eee Ae 340 ANTS ((P¥elllow: ciare 42 sa ee ee 194 The sum of the males is here nearly equal to the females as called for in the expectation (see below). It will be noted, however, that the gray males in F, greatly exceed the yellow males, and that the gray males are considerably more than half (268) the gray females. It may seem that the discrepancy in the yellow males is not due to their viability alone, but rather that the gray-bearing male-producing spermatozoa are more likely to fertilize the eggs than are the BalOw bearing sperm. The analysis is as follows: Gray 9° BYBrX — BYBrX Yellow o« bYBrX — bYBr F Gray 92 - BYBrX _pbYBrX : ' Gray «| BYBrX — bYBr F Gray @ 2 : Gray co 1 Yellow o 1} The reciprocal cross is as follows: | Gray | DSR eee err eee 346 Q 397 Guay:<. CO ea ieee er Cee 259 aby Gic! = ig 282 a Yellow! 9-0 eet Ae ates... 2 ame 226 | Yellow Sige e ue), aaa 230 The excess of females in F, is noticeable. in F, is equal numbers for all classes. the yellows. The expectation The grays run ahead of There is a very noticeable deficiency in the gray 36 T. H. MORGAN males compared with the gray females. The analysis is as follows: Yellow @ bYBrX —bYBry 9 Gray o BYBrX—DbYBr @ Gray @ bYBrX — BYBrX 9 B, Yay Cree bYBrYX—bYBr F Yellow 92 1 Yellow o@ 1 Gray 9? 1 Gray o1 Black by yellow Black females mated to yellow males give gray females and males. The data for F, are: (Gina Oke Be ee os eo a 5147 Gray (COE eaten. et A Ve eae ee 2451 Bre ot bel {G Os VPIBIRI CIS Oona wees reheat Lok oe a ey? 1591 Be ey 2 SS MiG, NeBlaclie ouieeuae to tle ca 750 Vell orcas ice hits Le eee ee 1957 BrOWDINGUS ete rae a ee ee ee 28 eto) The expectation is for the females, gray 6, black 2; and for the males, gray 3, black 1, yellow 3, brown 1. There are nearly 374 more gray females than 3 times the black females. There are also 500 more gray males than yellow males. Both excessive classes correspond to the F, classes. There are more black females than two times the black males. The analysis is as follows: Black @ .ByBrX — ByBrX 9 Yellow «* bYBrX—DbYBr @& Gray @ ByBrX—bYBrX @ = Gueueed Boise << BYBrxX — ByBrX — bYBrX — byBrX 9 Gametes of Fy BY Brx — ByBrX —IpY Br == byBr roa Gray @ 6 Crave moins Black @ 2 Black ot 1 Yellow ¢@ 3 Brown o 1 HEREDITY OF BODY COLOR IN DROSOPHILA aU The reciprocal cross yellow females by black males gives gray females and yellow males. The numerical data follow: (Gray PaO. Anco Semen, ep cee 2442 Gray) (ictudec foe ern eae Nore 1893 AD) EVO) Gat uMONE Rares OTME hat ree, een ee 1893 . mf Gy Or RR PB etch asia te ceed anemone soe oer tae 723 Peo tas EY ah eae Veclion Cor tae ene C0.) oul 1547 Mellow: | oh pak eee ee te nea 1548 BROW a Ory nee ee are eee ar eee 44] BROW MCUs tee are een 428 The expectation for the females and males alike is gray 3, yellow 3, black 1, brown 1. There are about 900 more gray females than yellow females; while the gray males are only 350 more numerous than the yellow males. The gray females and yellow males are the F, classes. The yellow females are as numerous as the yellow males. The black males are about a third of the gray males and more than this ratio in regard to the yellow males. The blacks run well ahead of the browns, the females being more than four times as numerous. The analysis is as follows: BL BX Yellow 9 bYBrX — $yBr Black & ByBrX — byBr Gray @ byBrX — ByBrX Bi Wellantcm bVEox — by be py Bek bY Bex = By Br By Bee Gametes of F, byBrX —bYBrX — byBr —bYBr 2 Brown @ 1 Brown of 1 Yellow 2 3 Yellow o& 3 Black 9@ 1 #£Black o' 1 Graynor | Grayous F, 38 T. H. MORGAN Wild (gray) by brown When the normal (Gray) females were mated with brown males all the offspring were gray. The numerical results for F, and F, were as follows: (Garany | 1 Qin hs Sree st. Bert cece cea 2618 Litany) cia mee aids Te Sete 1342 = tG? 266%) Blake” Oe Pas ae ces Forget 765 G9 byBro = 10 S358 = Ble Grae BAY Coes 351 Vellow:ot 2c en ee eeenicth dae oe 985 . BrOWIN Ors wei crease tree ays toe 300 The expectation is six gray females to two black females, and for the males 3 gray, 3 yellow, one black, one brown. The normal females are almost exactly twice the number of normal males. The normal males exceed three times the black males by nearly 300; and the black males exceed the brown males 50 flies. The yellow males are somewhat more than three times as numerous as the brown males, but less than three times as numerous as the black males. The yellow males, which should be as numerous as the normal males are about 350 fewer. The sum total of all the females is 3383 and of the males 2988. The males are about 400 flies fewer than the females. Despite these differences the numbers accord fairly well with the expectation, at least the classes stand in the same general relation that the analysis calls for. The analysis follows: Gray @ YBBrxX — YBBrX Brown o& ybBrX — ybBr Gray 9 YBBrX — ybBrxX Bi, Gray o&@ YBBrX — ybBrX vBBrX — YBBrX — ybBrX — YbBrX 9 Gametes of F, yBBrX — YBBrX — ybBr —YbBr ¢ F, Black @ 2 Black of 1 Gray @ 6 Gray o 3 Yellow o& 3 Brown o 1 The reciprocal cross, brown females by gray males gave gray females and yellow males. The numerical data for the F; and HEREDITY OF BODY COLOR IN DROSOPHILA 39 F, generation are as follows. The counts of the F, flies are taken from another similar cross. (Gira yet ROU FAs Tee ess Ae! 406 VeGiray~ PncUets. semen Beis es Re 171 Backs yiOi aie os aor ens: ls lene 74 G @ 124 Bl BIGkAS ot pare tes ace aetna S 3552 OD) Br uC ar ee Sanity ello: (G7 Sere ys vei ars. 3 0 oc oe 162 : MellowiG@icwtre re tn ee oy nee 190 IBTO WI) BORG IE ers war eines ae net 37 BROW rote cee ee 51 Owing to the failure of many of the brown females to breed the numerical results are small. The expectation for the females is gray 3, black 1, yellow 3, brown 1, and for the males, gray 3, black 1, yellow 3, brown 1. It will be observed that the gray females greatly exceed the yellow females while the yellow males exceed slightly the gray males. The F; females are gray and the males yellow.!. The analysis follows: Brown 9 ybBrX — ybBrX Gray o YBBrX — YbBr Cray 2 ybBrxX — YBBrX By, Yellow o@ ybBrX — YbBr i YbBrX — ybBrX — YBBrX — yBBrxX @ Couvergs cy YbBrX — ybBrX — YbBr —ybBr ¢ Gray 2 3 Gray of 3 FP Black Vo) 1 Black 1 : Yellow 93 Yellow 73 Brown @ 1 Brown o 1 Black by brown When black females are mated to brown males all the off- spring are black. The numerical data for F; and F, are as fol- lows: Black oh eee ee ee 6124 9 BCE De ee oy lacie) aie ec ke Soe 3015 \B ot 274 Brown: ere ernie. «meee 2472 1Tn another experiment the gray males equalled the gray females. 40 T, H. MORGAN The expectation is that the black females shall be as numer- ous as the sum of the two classes of males; there are about 600 too few males owing largely to a deficit in the brown class which runs about 450 behind the black males. The analysis follows: Black 92 yBBrX — yBBrX Brown & ybBrX — ybBr Black &@ yBBrX — ybBrX F, and gametes of F; BCL Ol Bee hee Black 9 2 F, Black of 1 Brown o 1 The reciprocal cross, brown females and black males gives black females and brown males. The numerical data are as follows: (Black: "O00. ../:4 ee ites ema Boe Black © Gis ..t\- chin. ho eee 1987 BS apes ‘aoe 5 BEOWil QO ieegah' ss) sos ge 6 ee ee 1532 2 Brown? ote) i. eee 1448 The expectation is equality throughout; but the Browns run, class for class, about 600 behind the Blacks. The analysis gives: Brown 2 ybBrX — ybBrX Q Black &@ yBBrX—ybBr @¢ Black @. ybBrX — yBBrxX @ By Brown & ybBrX—ybBr o@ Brown 92 1 r Brown o& 1 ; Black ¢ 1 Black co 1 Yellow by brown Yellow females by brown males give yellow females and males. The numerical data are: HEREDITY OF BODY COLOR IN DROSOPHILA 41 (Pe ellowgy Ors Weer se /)4, Pale. 2295 - cel ty ee Ort Ou Wie Mlowarotat 4 meee aya. sc eee 2232 UN) cae e Se GOH Mal) Brown On) AU ean ek Edin ie $30 Browne's so \aeaeet se tee teens 758 The expectation is three yellows to one brown, and the num- bers approximate to this relation. When the difficulties of separating these two classes is taken into account the agree- ment is remarkably close. The analysis follows: Yellow 2 YbBrX — YbBrX Brown Moe Fis — ybBr Yellow Q -YbBrX — ybBrX Q Fy Yellow o YbBrX — ybBrX — ybBr — YbBr Yellow @ 3 F Yellow oc 3 Y Brown @ 1 Brown o 1 The reciprocal cross, brown females by yellow males, gives yellow females and males. The numerical data are: Yellow Oh. eer aeee: 1181 Pe Apy eor20.. || “Yellow ots = epee meee oe ere 1409 Bryer by Src Be of" *) Brown") eye eee ree 571 Browniiquite eo Cee ee 520 The expectation is again three yellows to one brown;:and this is fairly well realized. There are more yellow males than yellow females, and slightly more brown females than brown males. The analysis is as follows: Brown @ ybBrX — ybBrX Yellow #@ YbBrX — YbBr Yellow 9 ybBrX — YbBrX By Wollow ctybBEX 5 | WbORrx— bee aa Yellow @ 3 F Yellow o& 3 3 Brown @ 1 Brown o 1 42 T, H. MORGAN DISCUSSION The color of the body of the wild fly appears from the experimental data to be due to at least three factors viz., yel- low, black, brown. It has been shown in a former paper that the red eye of the wild fly is also due to the presence of three factors viz., vermilion, pink, orange. In both series one at least of the three factors is sex-linked; the factor for black in the one and for pink in the other. In crossing both series give almost parallel results. In the eye-color series the factor for orange is always present, either simplex or duplex. In my former paper I could not determine whether it is sex-linked or not, because it had never dropped out, but since then I have obtained a new mutation in which orange has dropped out, and, by suitable experiments, it has been shown that this factor also is sex-linked. It appears then that in the eye color series there are two sex-linked factors, P and O, and one not sex-linked, V. In the present series brown occupies a similar position in the symbolism used to that of the orange factor in the eye color series, but on the basis of this similarity it would not be justi- fiable to conclude that the brown factor is sex-linked. In this connection I may record that during the summer of 1910 there appeared for a time, in one of my cultures, flies that had almost no color in the body although the eyes were red. A few pigment granules brownish in color were scattered over the abdomen. The flies resembled in some respects flies that had just emerged from the pupa case. The flies were extremely weak and died after a few days without progeny. Whether they represent the loss of the brown factor, or of the color producer can not be stated. Since they appeared in cul- tures of gray flies the latter interpretation seems more probable. Whether the comparison drawn above between the eye color series and the body color series has any real significance can, of course, be only a matter of conjecture. It should be pointed out that any eye color may be combined with any body color, and I have been unable to detect any correlated effect of these two combinations upon each other, except such effect as is due to color contrast. , HEREDITY OF BODY COLOR IN DROSOPHILA 43 One can not work with these body colors without being im- pressed by the similarities between the brown and the black flies on the one hand, and the yellow and the gray on the other. Brown and black lack the yellow factor, and if this, as I suppose, acts to some extent as an inhibitor the resemblance is manifest; while conversely the presence of the yellow factor in the yellow and in the normal fly makes clear their resemblance. One is tempted to surmise that black and brown may both be stages of the same chemical reaction, which surmise would be more probable if it could be shown that both factors are contained in the X chromosome, but this relation in itself could not be used as an argument to urge their dependent chemical nature. % PLATE 1 EXPLANATION OF FIGURES 1 Normal or gray female (the outer marginal vein is slightly exaggerated in the figure). (3 2 A black female. 3 A brown female. . x 4 A yellow female. The contrast between the black, yellow, and brown flies is well brought out in the figures. HEREDITY OF BODY COLOR IN DROSOPHILA PLATE 1 T. H. MORGAN PHOTO-CHROMOTYPE, PHILA)PA 4 THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 13, No. 1 E. M. WALLACE, DEL. 45 EXPERIMENTS ON THE REPRODUCTION OF THE HYPOTRICHOUS INFUSORIA I. CONJUGATION BETWEEN CLOSELY RELATED INDIVIDUALS OF STYLONYCHIA PUSTULATA GEORGE ALFRED BAITSELL From the Sheffield Biological Laboratory, Yale University TWENTY FIGURES (ONE PLATE) CONTENTS Vis — SLUM RC (6 LOCA KO Ee: cat Se ae eee a EE Ui nk IR RB OS Sia 47 INES AAWIGITOVOYO Sore 5 tea @ cle Deena ee IAL a Sk ee Ae Le Caine oi TONE, ~ IM EERIG TRIE Lopate =: Sse Co eae ane RRR Re errser co! 8 EC AEM eo 53 IV. Morphology and physiology of the non-conjugants..................... 55 Aen iyimea MVaberialens feck. o's sss woe ba a Ee ae eee 55 Bee brep arecsma penta | it 8 ei. dale als cnc hocallyt 2 Rae ee See a eee 61 V. Morphology and physiology of the conjugants.......... Le cath Er a 62 ANS. \baiab ayes nehiere | es Cee eee ema Ae oe Lela i ae ee 62 Bebrepareaumnabenialiae yt... .'. sa ove on 2 5 Pe Te > oe eee 64 VI. Morphology and physiology of the ex-conjugants...................... 65 PAT alii yala PUTA GOT ei. ini 3 id)a <3 + 2a sacs 9 Le ee ae 65 Pere pareGia heal: so) co. is os. oc. deacsia ss Rae ee ee eee 67. CAPES plattaC om eatOM ef ss. cep. 2 ced acinar Cee ee 67 MUSE DIscussionsanduconelusions. .... +... #11. 4 Ss ee eee ee 68 WIENER Generali sitmmiainyi rte ois ic. o's. 2's Hk veo bo ee ee io ee eee 72 LCDS 5 Abies (CHUTE ae CLR Oe OE ROM ne cr! on 5 5 74 I. INTRODUCTION Maupas (88, ’89), from his investigations on the phenomenon of conjugation in Stylonychia pustulata, believed that conjuga- tion, in order to be fertile, must take place when the organisms were at a certain stage of maturity which, in this species, occurred between the 130thand 170th generations. Conjugation could occur between gametes after this so-called period of sexual maturity, but in such eases the union was infertile, and the ex-conjugants would invariably die soon after separation. Provided conjugation did 47 THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 13, No. 1 4S GEORGE ALFRED BAITSELL not occur before the 170th generation senile degeneration set in, resulting in an abnormal diminution in size of the organisms and other pathological conditions and culminating in a complete dis- appearance of the micronuclei. ‘This senile degeneration reached its climax at the 316th generation at which point all the non-con- jugants would die. This author also held that conjugation in order to be fertile must occur between gametes of diverse ances- try, and that the phenomenon would not take place except when the animals were in a medium in which there was a scarcity of food. Maupas believed that these requirements were necessary for fertile syzygies not only in 8. pustulata but also, with some differ- ences in time of sexual maturity, for all Infusoria and that senile degeneration would always occur provided conjugation did not take place during the so-called period of maturity. He upheld then the view that rejuvenation by means of conjugation was nec- essary for the continued existence of the organisms. In case it did not occur the life cycle would end aftera fairly definite number of generations. Joukowsky (’98) found, as had been previously announced by Biitschli (’76), that. in Paramaecium putrinum fertile syzygies could occur between individuals only a wery few generations re- moved from conjugation. In a number of instances only seven or eight generations elapsed before fertile unions occurred between descendants of a single isolated ex-conjugant. This investigator also kept Pleurotricha through 458 generations in eight months without the occurrence of either conjugation or degeneration. Calkins (’02a) in his study of the life history of Paramaecium caudatum found that fertile syzygies occurred at the 350th, 410th, 467th and 500th generations, thus showing that, in this form at least, the so called period of sexual maturity really has no signifi- eance. He also showed that conjugation between closely related gametes was as generally fertile as in the cases where the gametes were of diverse ancestry. Scarcity of food was shown to be not a sufficient factor to induce conjugation and finally, while some evi- dences of degeneration were noted in the animals of the cultures at times, such as low division rate, dimished size, and occurrence REPRODUCTION OF THE HYPOTRICHOUS INFUSORIA 49 of monsters, the micronuclei were always present. The investi- gations on this form, however, gave evidence that the life cycle was essentially a limited one but that various chemical stimuli could be substituted for conjugation to ‘rejuvenate’ the animals and in this way the life history could be prolonged to a great de- gree. Eventually, however, the organisms reached a stage where it was found impossible to stimulate them further and the cultures died out. Work by Woodruff (’05) on a number of the hypotrichous in- fusoria, Popoff (07) on the Stylonychia mytilus, and Gregory (09)on Tillina magna all gave additional support to the view that the life cycle of these forms is a limited one. Both Woodruff and Gregory were able to prolong the life of the cultures somewhat by artificial stimulation. Enriques (05) from his investigations on a number of spe- cies of the Infusoria, including 8. pustulata, came to the conclu- sion that the degeneration which he observed in some cases was not a so-called senile degeneration due to ‘protoplasmic’ old age but a degeneration caused by the toxic influences of certain bacterial poisons in the culture media. Recent investigations by Woodruff (’08, 712) with Paramaecium aurelia which he has kept on a ‘varied environment’medium! have shown that this organism can be bred indefinitely without conju-_ gation or artificial stimulation, thus furnishing conclusive evi- dence of the unlimited power of reproduction of an infusorian without the need of ‘rejuvenesence’ when suitable culture condi- tions are supplied. Jennings (10), also working with Paramaecium, found no evi- dence to support Maupas’ view of the infertility of conjugation occurring between closely related gametes. 1 Woodruff (08) thus described the ‘varied environment’ medium: ‘‘It was found that Paramaecium can exist in nearly any infusion which may be made from materials collected in ponds and swamps, and accordingly, in the hope of supplying as far as possible all the elements which may be encountered in the usual habitat of the organism, water was taken from ponds, laboratory aquaria, ete., together with its animal and plant life. In other words, no definite method was employed in se- lecting the material, but it was simply collected at random from many sources, thoroughly boiled, and then used.” 10) GEORGE ALFRED BAITSELL To summarize briefly, then it has been shown by a number of investigators working on several species of Infusoria that Maupas was incorrect in asserting (1) that the Infusoria have a definite life cycle in which can be distinguished a period of sexual maturity during which time conjugation will be fertile and if conjugation does not take place, a resulting period of ‘senile degeneration’ finally terminating with the death of the organism after a certain number of generations, (2) that conjugation, in order to be fertile, must be between gametes of diverse ancestry. Finally, the work ~ of Woodruff with Paramaecium aurelia has shown that it is possible to breed this organism indefinitely without conjugation or artificial stimulation. Inasmuch as Stylonychia pustulata has been shown to be a spe- cies well adapted to the treatment demanded in laboratory cul- tures and also because some of Maupas’ conclusions which have been shown not to hold for other of the Infusoria have not been re-investigated in the species which he used, it was decided to make use of this Species in these experiments. In the present paper is presented a study of the following points: 1. The life history of Stylonychia pustulata when bred on a ‘constant’ culture medium of beef extract and a culture medium of hay infusion. 2. The morphological and physiological changes during the life of the organism when bred on these culture media. 3. The relation of conjugation to the environment of the or- ganisms. 4. The effect of conjugation between closely related individu- als which have had an identical environmental history. The writer is glad of the opportunity to express his great indebt- edness to Professor Lorande L. Woodruff not only for suggesting the problem but also for his advice and helpful criticism during the entire course of the work. REPRODUCTION OF THE HYPOTRICHOUS INFUSORIA 51 Il. METHODS In order to study the life history of a number of the Protozoa, Maupas isolated an individual in a favorable culture medium and in the course of a few days, when a large number of animals had arisen by fission, he counted the individuals, computed and re- corded the number of divisions that had taken place, and isolated one of the animals in another dish with fresh culture medium to continue the culture. In some cases over 900 individuals were al- lowed to accumulate before the isolation tock place. Biitschli later pointed out that there were two sources of error present in such a method, namely, the difficulty of securing an accurate count of so many organisms on a preparation at one time and the assump- tion that the rate of division had been the same for all the animals in the cultures, which might or might not be true depending en- tirely upon the physiological condition of the different individu- als of the culture. Calkins (’02) in his work with Paramaecium caudatum modified ~ Maupas method in order to overcome these difficulties. Instead of using a large receptacle and allowing the animals to accumulate in large numbers before isolation, he kept them on glass slides having a central depression holding about five drops of the culture medium, and from these slides he isolated an individual every one or two days so that only a few were present at the time of isolation. This method of isolation, besides giving the exact number of gener- ations through which the culture has passed, also prevents con- jugation occurring in the slides of the main lines whichare isolated daily. This method very little modified has since been used by a number of investigators, and notably by Woodruff in his work with Paramaecium aurelia. The cultures discussed in this paper have all been conducted by this same method and.as it has now become well-known only a brief description of the process will be given. The animals to be studied were isolated on glass slides having a central depression large enough to hold five drops of water. The slides to prevent evaporation were kept in moist chambers made from large Stender dishes with ground glass covers, and each hay- ing a capacity of eight slides. In starting a culture, an animal was a2, GEORGE ALFRED BAITSELL isolated on a depression slide in some of the culture medium and placed in the moist chamber. When it had produced four individ- uals by division, these were isolated and thus the four main lines of a culture were started. These four lines were examined daily, the number of divisions observed and recorded and one animal isolated from each to continue the main lines. The slides from which the isolation took place were saved as stock in order to re- plenish the main lines in case any of the individuals isolated were lost through accident. In this work three days stock was kept which, with the main lines, made in all sixteen slides under obser- vation inaculture. For isolation, capillary pipets were made use of, a separate one being reserved for each culture. In all the iso- lation work the greatest care was taken to prevent contamination of the cultures. All pipets and slides that were used were care- fully sterilized immediately before using. A dissecting micro- | scope with a 10 multiple lens was used in all the isolation work. Permanent preparations of the individuals from the cultures were made from time to time. The method used was that of Calkins (02a) and Woodruff (’05).? Two culture media were employed; a ‘constant’ medium which consisted of a 0.025 per cent. solution of Liebig’s extract of meat, and a hay infusionmedium. There are at least two essential con- ditions for a ‘constant’ culture medium. It must contain the elements necessary for the maintenance of protoplasm and it must be a medium in which bacteria will develop readily, in order to pro- vide food for the animals. For a complete list to date of the publications, in which the results of the inves- tigation referred to have appeared, see the bibliography at the end of this paper. Throughout this paper numbers in parentheses refer to titles in the bibliography. INHERITANCE OF FECUNDITY 15a ‘pedigree lines’ or strains of birds which breed true, generation after generation, to definite degrees of fecundity. - Some of these lines breed true to a high condition or degree of the character fecundity; others to a low state or degree of this character. Definite as these results are they give no clue as to how fecund- dity is inherited; what the mechanism is. Plate (43) has recently said: ‘Das Ziel der Erblichkeitsforschung muss die Aufstellung von ‘Erbformeln’ fiir alle untersuchten Merkmale sein.” This expresses the case precisely. To determine the ‘Erbformeln’ of fowls with respect to fecundity has been the goal towards which every part of the present investigation has been directed and urged. It is believed that a first approximation to the solution of the problem has now been reached. While there remain ob- scure points still to be cleared up, yet the results now in hand appear to indicate pretty clearly the general character of the mechanism of the inheritance of fecundity, and to show what lines further investigation of the problem may most profitably take. It is the purpose of this paper to present an account of the results men- tioned. In doing this it will be necessary to bring forward evi- dence of several distinct sorts, anatomical and physiological as well as genetic. Only by approaching this problem of the inher- itance of fecundity from allangles has it been possible to gain that understanding of the character itself which, in this instance certainly, is absolutely essential to a correct, interpretation of any results respecting its inheritance. BIOLOGICAL ANALYSIS OF THE CHARACTER FECUNDITY At the outstart it will be well to understand clearly what is meant by the term fecundity as here used. In a former paper (34) the terms ‘fecundity’ and ‘fertility’ were defined as follows, and have been used as there defined throughout the course of the investigation: We would suggest that the term ‘fecundity’ be used only to designate the innate potential reproductive capacity of the individual organism, as denoted by its ability to form and separate from the body mature germ cells. ‘Fecundity in the female will depend upon the production of ova and in the male upon the production of spermatozoa. In mam- 156 RAYMOND PEARL mals it will obviously be very difficult, if not impossible, to get reliable quantitative data regarding pure fecundity. On the other hand we would suggest that the term ‘fertility’ be used to designate the total actual reproductive capacity of pairs of organisms, male and female, as expressed by their ability when mated together to produce (i.e., bring to birth) individual offspring. Fertility, according to this view, depends upon and includes fecundity, but also a great number of other factors in addition. Clearly it is fertility rather than fecundity which is measured in statistics of birth of mammals. Taking fecundity as above defined it is obviously a character depending upon the interaction of several factors. In the first place the number of ova separated from the body by a hen must depend, in part at least, upon an anatomical basis, namely, the number of ova present in the ovary and available for discharge. Further there must be involved a series of physiological factors. The mere presence of an anatomically normal reproductive sys- tem, including a normal ovary with a full complement of ova, and a normal oviduct, is not enough to insure that a hen shall lay eggs, that is, exhibit actual as well as potential fecundity. While comparatively very rare, cases do occur in which a bird possesses a perfect ovary and perfect oviduct and is in all other respects entirely normal and healthy, yet never lays even a single egg in her life time. Such cases as these prove (a) that what we may call the anatomical factor is not alone sufficient to insure that poten- tial fecundity shall become actual, and (6) that the anatomical and physiological factors are distinct, in the sense that the normal existence of one in an individual does not necessarily imply the co-existence of the other in the same individual. A ease of this kind is found in hen no. 8051 batched March 29, 1909, and killed for autopsy record August 24, 1911. This bird had the secondary sexual characters of the female perfectly developed, and was entirely normal in other respects (body weight, 2366 grams). This bird never laid an egg during its life. The ovary was normal (fig. 1) and was of about the size proper to a fully developed pullet just reaching the point of beginning to deposit yolk rapidly in certain odcytes in preparation for laying. While counts were not made this ovary appeared to carry a nor- mal number of odcytes. In general it was anatomically normal, INHERITANCE OF FECUNDITY Lo7 but physiologically in the state of development appropriate to a five or six months old pullet just about to lay. The same was true of the oviduct. In this case the physiological factor or factors necessary to the bringing about of ovulation were simply totally lacking, in an otherwise perfectly normal bird. Some other cases demonstrating the same thing might be cited from our records, but this will suffice for present purposes. Turning now to the physiological factors involved in fecundity it would appear that there are at least two such factors or groups Fig. 1 Photograph (about twice natural size) of ovary of hen no. 8051. Note the presence of a large number of oécytes; none of which is enlarging in preparation for laying. See text for further explanation. of factors. The first of these may be designated as the ‘normal ovulation’ factor. By this is meant the complex of physiological conditions which taken together determine the laying of about such a number of eggs as represents the normal reproductive activity of the wild Gallus bankiva. Under conditions of domes- tication the activity of this normal ovulation factor will mean the production of more eggs than under wild conditions. Continued egg production involves certain definite and rather severe meta- bolic demands, which under wild conditions will not always, or 158 RAYMOND PEARL even often be met. Further, as has been especially emphasized by Herrick (18, 19, and other papers), egg laying in wild birds is simply one phase of a cyclical process. If the cycle is not dis- turbed in any way the egg production is simply the minimum required for the perpetuation of the race. If, however, the cycle is disturbed, as for example, by the eggs being removed from the nest as fast as they are laid, a very considerable increase in the total number of eggs produced will result. This, of course, is what happens under domestication. What an effect in increas- ing the actual expressed fecundity of a wild bird the simple re-— moval of eggs as fast as they are laid may have, may be illustrated by three cases from the literature. Austin (1) shows that whereas the wild Mallard duck in a state of nature lays only 12 to 18 eggs in the year, it will lay from 80 to 100 if they are removed as fast as laid and the bird is kept confined in a pen at night. Hanke (16) by regularly removing the eggs got 48 in succession from a common wryneck (Inyx torquilla*). Wenzel (53) in the same way brought a house sparrow’s productivity up to 51 eggs. With the domesticated Gallus the ‘normal ovulation’ factor may be taken as‘inducing a production of anything up to from forty to eighty eggs in a year, this production being spread over the period of from sometime in February to September or Octo- ber. In this physiological complex are involved the elaboration and deposition of yolks, the rapid growth of a few odcytes just preceding ovulation, ovulation itself, the activationof the oviduct, etc. The details of some of the processes involved have been described elsewhere (cf. Rubaschkin (44), Sonnenbrodt (48), Pearl and Curtis (33) and Pearl and Surface (37)) and do not concern us here. The essential point to be noted is that in this normal ovulation factor we are dealing with the basic physiological processes of normal ‘unimproved’ laying. ‘To make a normal lay- ing hen it is necessary to have present both the anatomical basis 3 1 give this scientific name with much hesitation, not knowing what pranks the rule of priority or other nomenclatorial disturbers of the peace may have played with it.in recent years. In any event the common name will quite sufficiently indicate what bird it is that is here under discussion. INHERITANCE OF FECUNDITY 159 discussed above and the physiological basis, which has been designated the normal ovulation factor. It is a fact well known to poultrymen, and one capable of easy observation and confirmation, that different breeds and strains of poultry differ widely in their laying capacity. In saying this the writer would not be understood to affirm that a definite degree of fecundity is a fixed and unalterable characteristic of any particular breed. The history of breeds shows very clearly that certain breeds now notably poor in laying qualities were once particularly good. One of the best examples of this is the Polish fowl. But, in spite of this, inheritable breed and strain differences in fecundity exist, and probably always have existed. Such inheritable differences are independent of feeding or any other environmental factors. Thus the strain of Cornish Indian Games with which I have worked are poor layers, regardless of how they are fed or handled. This is merely a statement of particular fact; it does not imply that there may not exist other strains of Cornish Indian Games that are good layers. The difference between this strain of Cornish Indian Games and Barred Plymouth Rocks, when kept under the same condi- tions and managed in the same way, is shown in tables | and 2, which give the frequency distributions and constants respectively, for flocks of these breeds kept at the Maine Station. The birds included in table 1 were all pullets, hatched at approximately the same time, and reared, housed, fed and cared for in all respects similarly. The Plymouth Rock distribution includes birds of both high and low fecundity strains. The low producing birds lower the mean in what is really an unfair manner, so far as concerns breed comparisons. The point is that, in the work of the Station, low-producing lines have been propagated for experi- mental purposes to a much greater extent than would be the case in purely random breeding of the Maine Station’s stock, the Barred Plymouth Rock breed. To make a perfectly just comparison between Cornish Indian Games and Barred Rocks, thé strains of the latter deliberately bred for low egg production should be excluded. It has, however, in the present case been 160 RAYMOND PEARL TABLE 1 Frequency distribution of winter egg production of the Barred Plymouth Rock and Cornish Indian Game breeds BARRED PLYMOUTH ROCKS LAYING THE | CORNISH INDIAN GAMES LAYING THE EGGS LAID IN THE SPECIFIED NUMBER OF EGGS | SPECIFIED NUMBER OF EGGS WINTER PERIOD Absolute number | Per cent of flock | Absolutenumber | Per cent of flock 0-5 43 14.4 32 | 48.5 6-11 22 7.4 8 12.1 12-17 28 9.4 9 13.6 18-23 19 6.3 6 BI 24-29 25 8.4 7 10.6 30-35 26 8.7 1 1.5 36-41 19 6.4 5) 4.5 42-47 27 9.0 48-53 16 5.4 54-59 21 7.0 60-65 14 anf: 66-71 10 3.3 12-17 9 3.0 78-83 3 1.0 84-89 3 10 90-95 0 96-101 8 Ta 102-107 10) 108-113 t ES | 114-119 2 0.7 | Tetalyy wade: 299 100.0 Cer 99.9 TABLE 2 Constants for variation in winter egg production of the Barred Plymouth Rock and Cornish Indian Game breeds bree eas nee , tga 3 ns Sait eggs | per cent ; Barred Plymouth Rock........ 36.35 + 1.04 | *26.69+0.74| 73.42 + 2.92 Cornish Indian Game.......... | 11.64 + 0.88 | 10.61 +0.62/ 91.15 + 8.73 Ditieren cess. tise, Ah aaa k eee +94.71 + 1.36 |+16.08 + 0.97 —17.73 + 9.21 Barred Plymouth Rock: | | All High Lines in 1908-09!..... | 54.16 | All High Lines in 1909-10'....., 47.57 . All High Lines in 1910-111.....| 50.58 | 1 Figures taken from Pearl (28). INHERITANCE OF FECUNDITY 161 deemed best to take the whole flock of Barred Rock pullets for the laying year 1910-11, without any selection. ‘The comparison is sufficiently striking even on this basis. From tables 1 and 2 it will be noted that: 1. The mean winter production of the Cornish Indian Games is less than one-third that of the general flock of Barred Plymouth Rocks, under uniform environmental conditions. 2. The winter production of the Games is considerably less than a fourth of that of the high producing lines of the Barred Rocks. 3. The variabilities in both cases are high, but relatively not significantly different. It is of interest to note that the observed coefficients of variation for winter production here given are of the same order of magnitude as the mean coefficients for the lay- ing of the four winter months, November, December, January and February. ‘Taking the mean of the coefficients of variation for these four months as given by Pearl and Surface (37, table 5, p. 96) we get 95.15. The inferiority in egg production of the Cornish Indian Games is most strikingly shown by the integral curves from. table 1. In table 3 the integral curves are given (in inversed form) for the winter production of Barred Rock and Cornish fowls. The data of table 3 are shown graphically in figure 2. This diagram is to be read in the following manner. The percentages of the flock laying a specified number of eggs are plotted on the abscissal axis. The different egg productions are plotted as ordinates. From the diagram it appears (for example) that whereas 47 out of every 100 birds in the Barred Rock flock each produced 35 or more eggs in the winter period, only 4 and a fraction birds out of every 100 in the Cornish Indian Game flock were able to produce as many eggs as this—35—in the same period. Now in individuals which are high layers, and have this charac- teristic in hereditary form, there must be involved some further physiological factor in addition to the normal ovulation factor alréady discussed. An analysis of extensive statistics has shown (36, 37) that high fecundity represents essentially an addition of two definite seasonal, laying cycles to the basic, normal reproduc- 162 RAYMOND PEARL TABLE 3 Showing the percentage of the whole flock producing in the winter period more than certain specified numbers of eggs, in the case (a) of Barred Plymouth Rocks and (b) of Cornish Indian Games SIN ATED PE AGE OF THE FLOCK PRODUCES BARR Y MOUTH THE INDICATED RCENTAG OCK P UCES ED PLYMOUT | CORNISH INDIAN GAME IN THE WINTER PERIOD ROCK GKOMAINOTe CLESs. fanaa martes cere 85.6 51.5 [2VOr-MMORELE LOSi = fos coe eee eee oie 78 .2 39.4 TSMOT MONETE LES Raa: re ene oem ate eens 68.8 25.8 DAY OF MOLEC LES. Hee keds eee eee re | 62.5 Gia SOWOrSIMONE ELSES acdsee eeu ere « | 54.1 6.1 SHLOLr MONE CLES 4s ahr, eich ciate eee, or 45.4 AD TOY MONE CLLSH cs sere (bee eae 39.0 ASL OF MOLE. CL ESh. se o../, ahs pee ee ee 30.0 HA JOTAINONELE DOSS Sy. 2 here. agian emer Setar 24.6 GOVOTAMIOETELOStece seer eee ee: LAO 66/0 MONG CLESs. 5 ot jaca. > ei tociee onee 12.9 (2NOT MMOLE CLES rnc... aa iene eee 9.6 TRVOPMNOTE CLEANS i sc blag ane eee ee | 6.6 S84) OL MOre CLESoasi ke, Melaraee hake ee 5 OOVOTPMOLEIELESs 5 505 tesa ee eee a ee 4 OGFOLMMOLELC GOS. Sic nce cue ee eee oe 2 LOZ OFAMOreve LOS. nF ets oes ee OTR 2. LOSSOT MOLKe CLES se iidee oa at hee eee 0 WAS OR ANIOTEAE POSS. 5. nN ctomieat . ER Ee: 0 20. OP MORE YC POSH a.) caret eta eee 0 Soo Soe SSeS oS See ts jor) tion cycle. These added periods of productivity are what may be called (cf. 37, 28, 30) the winter cycle and the summer cycle. The winter cycle is the more important of these. It is the best practical measure of relative fecundity which we have and has been used as the chief unit of fecundity in these studies. It con- stitutes a distinct and definite entity in fecundity curves. The existence of this added fecundity, in high laying birds must de- pend upon some additional physiological factor or mechanism besides that which suffices for the normal reproductive egg pro- duction. Given the basic anatomical and physiological factors the bird only lays a large number of eggs if an additional factor is present. As to the nature of this physiological mechanism we can only speculate. It probably involves fundamentally such matters as ‘SoUIBY) UBIPU] YSTULOQ—oUl] UeyOIg ‘“SyooY YNowATG peleg—ouly pyog ‘sjayjnd ower uerpuy ystus0D pu yooy YyyNouA|g poeiieg Jo uorponpord Bo ToJUIM JO SoAINO [BASOYUT FZ “BT 163 JYOW YO SIDPFT S0 YAIGWIN CFLVIIGN/ INIDNTOYS LNII Y IAS 0 O/ OZ OL Of OS oF OL og Ob S YILN/IM S Ss) IIT Ss i) NOILINIOLS INHERITANCE OF FECUNDITY 8 O47 164 RAYMOND PEARL more perfect metabolism, including the distribution of substance and energy to the ovary, on which very heavy demands are laid in a high fecundity record. Immediately it involves a control of the process by which the supply of odcytes on the ovary in the final stages of rapid growth by yolk deposition is kept at a rela- tively high level for long periods of time. Sonnenbrodt’s (48) work suggests that the interstitial cells of the ovary may be connected with the process. Thus he says (loc. cit., p. 421): ‘Bei dlteren Hiihnern findet man die Zwischenzellen immer noch, und besonders in der Nahe der Gefiisse. Sie legen heir gruppen-und nesterweise zwischen den Follikeln und vor allem auch in den Stielen der grésseren Follikel,t immer dort, wo beson- ders starke Blutzufubr giinstige Ernahrungsbedingungen bietet.”’ It is quite conceivable that the presence of numerous inter- stitial cells on the stalks of the follicles of rapidly growing oécytes is a cause of the rapid growth rather than an effect, as Sonnen- brodt suggests. The whole subject of the intimate physiology of the ovary needs more study. Whatever the precise nature of the factor under discussion, which is a matter for future investigation, the main points which appear clear at present are that: (a) high fecundity represents a defi- nite addition to the normal egg production sufficient in amount for purposes of reproduction. This added fecundity has been shown (cf. 23, 30) to be definitely inherited in certain cases at least and may be regarded as dependent on or determined by some physio- logical factor or complex of factors not present in birds which exhibit a low degree of fecundity.* This physiological complex may be designated as the ‘excess production’ factor in fecundity. We may next consider in greater detail these factors influencing fecundity, taking first 4 My italics.—R. P. 6 Throughout this discussion it is presumed that the reader will understand with- out repeated specific statements that attention was paid to environmental factors in the experimental work. That is, when the statement is made that one bird or set of birds exhibits high fecundity and another low fecundity it is to be understood that both sets were hatched, reared, fed and cared for in all respects in as nearly precisely the same way as is possible, considering that fowls are, in some degree, free agents and cannot be absolutely controlled. The extent both in time and INHERITANCE OF FECUNDITY 165 The anatomical basis of fecundity Since, as already pointed out, egg production obviously depends in part upon the presence of ova in a normal ovary, a question which demands consideration is the following: To what extent are observed variations in fecundity (i.e., in the number of eggs laid) to be referred to anatomical differences? In other words, does the ovary of a high producing hen, with for example, a winter record of from 75 to 115 eggs, contain a larger number of odcytes than does the ovary of a hen which is a poor producer, laying no eggs in the winter period and perhaps but 10 or 15 eggs in the year? To get light upon this question the observations to be described have been made. The object was to arrive at as accurate a relative judgment as possible regarding the number of odcytes in the ovaries of different individual birds. It is, of course, impossible practically to determine accurately the total absolute number of odcytes in the ovary. What ean be done, is to count the number of odcytes which are visible to the unaided eye. While such results do not tell us, nor enable us to estimate with great accuracy, the total number of odcytes in the ovary, they do nevertheless throw interesting and useful light on the question raised above. The counts of the visible odcytes for a number of birds are given in table 4. These counts were made at my suggestion by my assistant, Miss Maynie R. Curtis, to whose painstaking care and skill in carrying through the tedious business of counting it is a pleasure to acknowledge gratefully my indebtedness. Prof. W. F. Schoppe of the University of Maine is carrying this work for- ward and later we hope to be able to publish more extensive data. space, and the manifoldness in respect to method, of the experiments upon which this discussion is based are so great and the checks on this point have been so numerous as to make it quite certain that the results are not influenced by a differ- ential effect of the environment, arising from individual preferences of birds for particular sorts of food, or other similar peculiarities of behavior. When a result is Stated to be due to inheritance the reader may assume, even though a specific statement is not made to that effect, that careful, critical consideration has been given to possible environmental influences. 166 RAYMOND PEARL So far as I am aware the counts here given are the first attempt yet made at anything more than the roughest sort of a guess at the number of eggs in a bird’s ovary. While these counts do not give the total numbers they do establish minimum values. A given ovary certainly does not carry any less than the number of visible ova. A word should be said as to the method of making the counts, and the meaning of the subdivisions of the table. The counts were made in some cases on fresh, and in other cases on pre- served ovaries. There was found to be little difference in the two methods, as regards the ease and accuracy of counting. In making the counts small pieces of ovary were cut off, and teased apart with needles under water and the visible odcytes on the small fragments counted. In delimiting boundaries where a number of small odcytes were closely packed together, a hand lens was used. No o6cyte was counted, however, which could not be seen with the unaided eye. In other words the lens was not used to find oécytes which might otherwise be missed, but merely to ald in the dissecting of the material. ; In the odcyte counts given in the table it will be noted that these are grouped into four categories. The first class includes ruptured follicles from which the ova have been discharged. A ruptured follicle which is large at the moment the ovum leaves it gradually shrinks in size and is more or less completely absorbed. On the ovary of a hen which has laid, however, there will always be found a certain number of these discharged follicles not yet absorbed. When such follicles get very small it is exceedingly difficult to distinguish them from small odcytes (1.e., undischarged follicles). Undoubtedly there are errors in classification in this respect in the counts, but for present purposes this is not a matter of great importance. If the eye were sharp enough it might per- haps be possible to distinguish a ruptured follicle for every egg which has ever been laid, since it is doubtful if the absorption is ever so complete as to leave absolutely no scar. It is of interest to note that in the counts there is a reasonably close relation between the follicle count and the record of eggs laid. INHERITANCE OF FECUNDITY 167 The odcytes proper are divided in the counting into three classes: those 1 em. or over in diameter, those between 1 mm. and 1 em. in diameter, and those less than 1 mm. in diameter. The first of these classes includes the large yolks nearly ready to leave the ovary and pass into the oviduct. They are in process of rapid enlargement by the deposition of yolk. The next class includes those odcytes in which yolk deposition is started but is proceeding at a slow rate. It is from this class that the first class of rapidly growing yolks is constantly being recruited. Finally the ‘‘under 1 mm.” class represents the make-up of the bulk of the ovary. It will be understood that these size classes are only roughly delimited, the diameter of each odcyte having been estimated but not carefully measured. Columns in the table are devoted to ‘Total number of eggs laid in life’? and ‘‘Winter production.” The first of these has no particular significance since obviously it depends on when the bird was killed in order to make the odcyte count. Winter production, however, represents a definite entity in fecundity as already pointed out above (p. 162).° Winter production records are directly comparable with one another. It is the inheritance of this fecundity unit that is primarily being studied in these investigations. From this table a number of points are to be noted. In the first place it is clear that the number of visible o6cytes in the ovary of a hen is very large; much larger, I think, than has generally been supposed. While to be sure there are for the most part only vague statements respecting this point in the literature, usually these statements are to the effect that the bird’s ovary contains ‘several hundred’ ova. The only direct statement as’ to the actual number of odcytes in a hen’s ovary which I have been able to find is given by Matthews Duncan (8) on the very dubious authority of Geyelin (11) to the following effect (loc. cit., p. 36): “Tt has been ascertained that the ovarium of a fowl is composed of 600 ovula or eggs; therefore, a hen during the whole of her life 6 For general ‘discussion of ‘“‘winter production’’ as a unit of fecundity, see (28), (30), (34), (87), (88). It comprises the egg production up to March 1 of the laying year. RAYMOND PEARL 168 ‘pol[P] UdYM UOTJIPUOD SUTAR] UT JOU SplT ¢ ‘uoTyonpoad [e404 ayy A][eorjovid quoseidet soinSy 94 J, “OT6T ‘[ 9quUIsAON 03 dn spaoded oy} ynq ‘ajT] IOJ Sp10901 OY} JOU aIB SVSBd 9SeY} UT Sp1o0d001 S59 oY, “ApNys Juesord oY} UT VSN 1OJ S[ENPIATPUT pojsou-dea} osoy} OW VALS Ajpury AoA OYM ‘AJISIOATUL) ]]OUIOD Jo ooIy ‘WY souve ord 0} poyqoepul WB J SplIG SUTMOT[OJ 9914} BY} PUB STYy IOV z : ‘uoTydaosqe jo ssadoid ul Os § SOpNyoul sy, + 98G | GEG See ele at | j TL Arvenuesr | é uoy BouTMsy | | Day GO| PAL 9¢ | ¢ 6 eee ky ALENUB Hy Udy BoUTNY) OT 0006 | GZST | OL G 0G “SOL | Fel | TI. 02 Ye OT, ‘TE = Youre SOIT | TL | ST OSGI.| cen | 291 99 | 7S | Sl. | OL. ‘ZI Amp |) 60.‘1 dy owen ‘puy ystur0g | zg | FT GhIZ | ZC0Z | 08 pene aie0 Gee a Ole Cone Sq ete AGT ee -GOs: Wlemmcmnt UIOYseT OFM | SE8E) ST TOLT | 9Z9T | G2 blebs ac0 OL -| OL. “€I dequieosq | 60, ‘12 A®IN uIOYsoT OYA | SFE | ZT GOIE | GAZE | SOL | T | 24IG | ZE. | L6E | OF. “GT Jequiessq | 60, ‘8c A¥IN utoysey oFIYM | 290% | TT Gere | OPIS | Tes | 2 | G4 | | S6E-| OF. 0% tequieceq | 60, ‘ST Av gUloysaT OFIUM | 9FGE | OT [cciesOrel-\Grl | 12k | Te 3\0= | ez Ol. ‘2 Ame) 60, ‘OS = Your yoo “Ald powvg | Sap | 6, QLST | SGPT | 26 Giaedley Ge sae Gae SG lgec leslie wal yore, | OT, “61 ARN yoy “ATq poreg | 0108) 8 WOVS-|-6202 | 67 | 9 |-21 08 SS eal 2 youre | OL. oun yooy “ATd poreg | 6008 | 2 POIL | SZIL | oF les Recs Om =a 20 Th. Your | OL. “G oune yoo “ATd powed sl0s 9 QOkZ | TZIS | 6% Le aaeGi? aalee Ee Tslect, dy | O01, ‘860 THdy yooy “ATq porieg | L9ET | ¢ PLIT | 9601 | 89 BRE GU, PlrGacesse yy TI. ‘FI Vor OL. ‘Z ounf yor “AT peweg | S008 | F PGs GeSemecdae (| GaSe ALO Leet Ol ore | Ol, ‘T oun yoo “Ald pore | 0808 | & 9991 | 96ST | I¢ aCe akO Oe bees Your | OL. “G ounf yooy “Ald poreg | L108 | % SZZ1 | GFIT | €¢ Gp ans Or TI. ‘8% OLBIN, |S Olaed oun f yooy “ATq poweg | 1208 | 1 5c oho Repel | ae | ae Se | | PE TEES LE Ce eed | Be lobe | ear Sa cia Bob less BE | oe | Se | 3 | 4 ‘ON | ‘ON 3 B a 3 a a 2 ma | zB g Seal i a 3 > eI aqaliy aLya ONIHOLVH AO ALYG aqaqgugd aura | asvo Belgie Gerstein ogee rata Bw | & | : le ot Bo | | | | | S | splug W1p}.199 fo havao ay] U2 sajhiaeo a)qisia fo saqunu ay) buinoys ¥ WTAVL INHERITANCE OF FECUNDITY 169 eannot possibly lay more eggs than 600, which in a natural course are distributed over nine years in the following proportion.” This statement is followed by an utterly preposterous and pre- sumably entirely imaginary table from Geyelin, supposed to show the laying of hens at different ages. How far from the truth the table is is indicated by the fact that according to it the pullet year is the least productive of any of a hen’s life, save only for the ninth year when the last remnants of the original 600 eggs are being tardily and, one must suppose, sorrowfully ejacu- lated!’ As a matter of fact repeated trap-nest and other tests in all parts of the world have shown again and again that, on the average, the pullet year is the most productive of a hen’s life. From the figures given in table 4 it is furthermore apparent that the absolute number of odcytes in the hen’s ovary is very much larger than the number of eggs which any hen ever lays. A record of 200 eggs in the year is a high record of fecundity for the domestic fowl, though in exceptional cases it may go even a hun- dred eggs higher than this (ef. 29). But even a 200-egg record is only a little more than a tenth of the average total number of visible odcytes in a bird’s ovary, to say nothing of the probably much larger number of odcytes invisible to the unaided eye, but capable of growth and development. In other words it is quite evident from these figures that the potential ‘anatomical’ fecund- ity is very much higher than the actually realized fecundity This is true even if we suppose the bird to be allowed to live until it dies a natural death. Experience shows that birds which make a high fecundity record in the first year of their life, generally _ speaking, never do so thereafter. In general an examination of what long period records are available in the statistics of this Station, and also in the literature, indicates that probably only relatively few birds of the American or Asiatic breeds at least, would lay many more than 400 to 500 eggs in their natural life time, if they were allowed to live it out. Records of ‘1000-egg’ birds are in existence, but such birds are rare. 7 It is difficult to understand how so acute an investigator as F. H. A. Marshall could have been so imposed upon by this wonderful table of Geyelin’s as to repub- lish it in his valuable and interesting book on the “‘ Physiology of Reproduction.”’ THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 18, No. 2 170 RAYMOND PEARL One of the longest continuous egg records of an individual bird, which may be considered accurate, with which I am acquainted is that given by Handrik (15) (for a Leghorn). This bird was hatched in 1901. Its egg record was as follows: Calendar year Eggs laid MN QOZ | Vata cere 2 ese ro CONN oe es ae OUR RT RRL gl 105 MOOS Bhs Cats SA Ry Sa eas pes. er ake eee ee aveainee SORRIR SS 163 Or Bia Re ats en mee eat ONG. Ero, Sieg ar RT a fe 138 DOOD a sks es eet este ee RE cay Daal aR ae 159 QO GE me Bere deo Se Ree ice SAO cp Ea ee ae 160 BOOT Ful BLISS ERR SEs REE eam 8 CEG. S89 ea ne 8 133 TOS s pig ed he Te eR, pt ee ra ys 8 Beh ccna ee eae 111 WP Ot ales cv, csc las hee eee Pe ice c,h te Ree ee een le 969 Average! per Vea 5... Ge Asai na eas = fat. ois Cet eee ae ee ee 1383 Heier (17) gives a four-year record for a Braekel hen, which is distinctly higher than would usually be obtained over so long a period. The figures are as follows: Laying Year Eggs laid BESTS Gate sae EPs ee ec ces eat a | OE I 2k Ree Nar AO Ae 153 SECON NA ee. ystiy as Ae ee aks. +. 54.50 eggs 26.50 eggs C. With1 2 indicated to be of class 6 = fl,L. . Flile. + Q Progeny Winter Production: Over 30 Under 30 Zero Observed. eat weet heer 0 2 0 JONAH Ts popiaSee nia oka oe cis 0 2 0 Mean winter production of 2 9 in indicated class....... meds te 4.00 eggs All Q Progeny Winter Production: Over 30 Under 30 Zero Obsenvedinns a.0452j.o ses ee 11 14 2 HE NeCte dan me titan 12.5 UO es 0 Mean winter production........ 62.64 eggs 18.85 eggs O eggs The families are small in this case. From both these pure Barred Rock and the cross matings in which ~ 564 entered, how- ever, there can be no doubt that he is a class 7 male. The two zero birds are to be reckoned as ‘somatic zeros’ rather than game- tic. Both began laying at the very beginning of the spring period, and made records which indicated to one familiar with this sort of material that they belonged genetically in the ‘Under 30’ class and only by accident failed to lay some eggs during the winter period. INHERITANCE OF FECUNDITY 193 B.P.R. ¢ D&8. Indicated constitution = fl,L. . flLn. This bird was purchased in January, 1908, from Gardner & Dunning, a then well-known firm of Barred Rock breeders of Auburn, N. Y. Nothing was known of this bird’s previous his- tory or pedigree. The bird was hatched in the spring of 1907, and used in our breeding pens in 1908 and 1909. In 1908 he failed to get any adult daughters. This, however, was not the fault of the bird, but of the conditions under which the breeding had to be done that year (cf. Pearl and Surface 35). From the records of the daughters of 2 58 obtained in 1909 and exhibited below it appears clear that he was a class 7 male. The breeding history is as follows: . Matings: A. With9 2 9 indicated to be of class 1 = fIyL2 . Fil. Q Progeny Winter Production: Over 30 Under 30 Zero Observed Nt ater ton. sane st 10 13 1 WENeched.. cP ok 12 12 0 Mean winter egg production of 2 9 in indicated class....... 52.22 eggs 17.25 eggs 0 eggs B. With4 9 @ indicated to be of class 6 = fll. . Fhl. Q@ Progeny / Winter Production: Over 30 Under 30 Zero Obsenvedietey, Meant avi seo 0 5 0) HG ECLE MRR. Seo eo 0 5 0 Mean winter egg production of 2 2 in indicated class....... 15.80 eggs All 9 Progeny Winter Production: Over 30 Under 30 Zero ObReEVedE Ne ce ita ance 10 1S)5 1 LE DOCCE eee racy roe eer i 12 17 0 Mean winter production....... 52.22 eggs 16.82 eggs 0 eggs The single zero bird here (¢ F158) cannot fairly be regarded as a non-conformable case because of the following history. She was hatched March 30, 1909. She never laid an egg and died May 238, 1910. Autopsy showed the ovary and oviduct to be in an infantile condition. The ovary weighed-1 gram and the oviduct 194 RAYMOND PEARL 2 grams. The ovary showed no odcytes enlarged by yolk deposi- tion or enlarging. There was no evidence that the ovary had ever shown the slightest trace of functional activity. But a normal bird hatched in March will exhibit signs of ovarian activity before May of the following year, even though she belongs genet- ically to the ‘Zero’ class in respect to winter production and does not lay. While the autopsy showed no obvious lesion of ovary or oviduct, this by no means proves that there may not have been present some deep-seated functional derangement. BPA. Guorés -Indicatedsconstitution = fljas wlio: This bird was used in the breeding season of 1910, having been hatched in 1909. He proved not to be all that might be desired as a breeder, being somewhat lacking in vigor of constitution. Partly on this account, he got comparatively few adult daughters, as indicated in the following breeding history. Matings: A. With5 9 @ indicated to be of class 1 = fliLe2 . Flile. Q Progeny Winter Production: Over 30 Under 30 Zero ObDServicds ein. cce eee ae 43 63 0 GPE Cle ne nee eee hee 5.6 5.5 0 Mean winter production of 2 2 mancdicatediclass a. me ease 47.50 eggs 15.67 eggs B. With 2 92 @ indicated to be of class 2 = fl, L2 . Flale. Q@ Progeny Winter Production: Over 30 Under 30 Zero Observed nest. hes 4 0 0 LECTED Is cin Seeker 4 0 0 Mean winter production of 2 @ inandicatediclasstassm.seces-e 49.25 eggs C. With 1 2 indicated to be of class 3 = flail, . Pll. Q Progeny Winter Production: Over 30 Under 30 Zero Observed. -cheae. giuretn ss 2 1 1 Be DECUCO ee ae Ee 2 2 0 Mean winter production of 2 9 imun dicate diclasssa.sce ae eiee 55.50 eggs 16.00 eggs 0 eggs INHERITANCE OF FECUNDITY 195 All Q Progeny Winter Production: Over 30 Under 30 Zero Observed tee wots kis ok Dee 103 fe 1 TGS RACHAG TE Sal cs tac Cette Nene AIO Tero} 7.6 0 Mean winter production........ 49.80 eggs 15.71 eggs 0 eggs The zero bird here is an exception for which no apparent ex- planation is forthcoming. She was not pathological. She was however a June hatched bird. Unfortunately she was not bred, and therefore it is not possible to be sure of her gametic constitu- tion. In spite of the fact that the total number of progeny here is small, there is little doubt of the correctness of the classification. The mean productions for birds in the ‘Over 30’ class in the several matings are comparatively a little lower than those of the progeny of other class 7 males. It is interesting to speculate as to whether this may be connected with the lack of great vigor on the part of the sire. No data are available from which to get critical evidence on this point. B.P.R. 2 56. Indicated constitution = fl,L. . fl,Le. This bird was purchased in January, 1908, from Mr. C. H. Welles of Stratford, Conn. It came from a strain of Barred Rocks well known in the show-room, but not specially bred for egg pro- duction. This fact is of interest in connection with the breeding history of the bird, which indicates clearly that he was homo- zygous with respect to L2. The result shows, in other words, that a male Barred Rock from a strain bred purely for the fancy may still carry in pure form the factor for high egg production. This male bird (56) was bred two seasons (1908 and 1909). The first year he got but very few adult daughters, owing to the unfavorable conditions under which all the breeding had to be done in 1908 (cf. Pearl and Surface 35). In 1909 the results were better. The adult daughters from both seasons are taken together in the following breeding history. 196 RAYMOND PEARL Matings: A. With5 9 2 indicated to be of class2 = fli, Ls, . FIyl. Q Progeny Winter Production: Over 30 Under 30 Zero Observed) sha rk ely a ete: 7 0 0 TGA VOU eed alt die Gites ta piaordsd A ih 0 0 Mean winter production of 2 Q IM Imditcated class. + ssceie eee 54.57 eggs B. With4 9 @ indicated to be of class3 = fIils . Fils. Q Progeny Winter Production: Over 30 Under 30 Zero Observedsa sy.) ooo ee eee 9 4 0 ICD CCLCU pan en Se eee 6.5 6.5 0 Mean winter production of 2 2 Imyndicatedeclasswarn nen ee 56.89 eggs 19.50 eggs C. With 2 9 9 indicated to be of class 6 = fl,Le2 . Fhils. Q Progeny Winter Production: Over 30 Under 30 Zero Observeds yess Hehe dee 0 3 0 PPEDCGECO Pes men 8 Ac cradles 0 S 0 Mean winter production of 2 @ IMpIMGIC ATE CUCLASSs eee eae , 13.67 eggs All 2 Progeny Winter Production: Over 30 Under 30 Zero ODSservedisra: tata Anal Gas eee 16 7 0 iriNAWeGh chose seouchoal on dobs 13.5 GEO) D. Mean winter production........ 55.87 eggs 17.00 eggs m0) The agreement between observation and expectation here is satisfactory, excepting the case of the class 3 females. There the deviation from the expected half is wide, but the numbers in- volved are small. The behavior of ~ 56 with class 2 and class 6 females gives clear indication of his gametic constitution. B.P.R. 7 563. Indicated constitution = fl,Le . fl,L. This bird was hatched in 1909 and used as a breeder in 1910. He was an exceptionally fine, vigorous bird. The breeding his- tory is as follows: 3 INHERITANCE OF FECUNDITY 197 Matings: A. With 6 @ @ indicated to be of class 1 = flyL2 . Flile. Q Progeny Winter Production: Over 30 Under 30 Zero ODSErvedin.c facts os, cs 0b ee 11 11 1 TEA CACAO Ronis celia Cae Te} 11S 0 Mean winter production of 2 9 imemndicatediclasstes sm «404.02 64.09 eggs 17.91 eggs 0 eggs B. With 5 @ @ indicated to be of class 2 = fl, Le . FIle. Q Progeny Winter Production: - Over 30 Under 30 Zero Obsenviedteetec suas acaes.: 18 1 0 TEVA hic ani #60 So trol o 6 on BORE 19 0 0 Mean winter production of 2 @ in indicated class. ...5....... 63.56 eggs. 1.00 eggs All 9 Progeny Winter Production: Over 30 Under 30 Zero Obesrvic diem nage ach 29 12 1 EIDE CLEC See ee Te Nes oe ee 80.6 11.6 0 Mean winter production........ 63.76 eggs 16.50 eggs 0 eggs Aside from the two outstanding exceptions the agreement between observation and expectation is excellent. From the records available there is no evident explanation for the two excep- tions (the ‘Zero’ bird in the A matings, and the ‘Under 30’ bird in the B matings). Neither of the birds were bred, and hence no help is to be had from the progeny in explaining them. It is reasonable to suppose that the observed records for these birds are somatic fluctuations, but this cannot be demonstrated now. This ease illustrates’ an unavoidable difficulty which attends that method of work which first collects data at random and without any theoretical guide, and then later undertakes their analysis. If one had been carrying on the breeding in the present case under the guidance of the hypothesis as to the mechanism of the inherit- ance of fecundity now under discussion, obviously many matings which actually were not carried out would have been made to test out somatically exceptional individuals and so learn inet gametic constitution. 198 RAYMOND PEARL B.P.R. 2 D31. Indicated constitution = fl L, . fll. This rather remarkable bird was used as a breeder for three successive years, and then retired merely because no more of his progeny were needed, and not for any evident diminution of vigor on his part. This bird was first bred as a cockerel in the spring of 1908 (hatched in 1907). All that was known of his ancestry was that he was the son of a hen that had laid 200 or more eggs in her pullet year. Some notion of the vigor of ¢ D31 as a breeder may be gained from the fact that, taking all three seasons together and including all parts of the breeding season in each year, 89.4 per cent of all the eggs laid by hens mated with him were fertile. This is an extraordinarily high record, considering all the circumstances, and particularly the seasonal and housing conditions. So far as concerns adult daughters the breeding history of this bird is as follows: Matings: A. With9 @ 9 indicated to be of class 1 = fli Le . Flile. Q Progeny Winter Production: Over 30 Under 30 Zero Observedeee ae ata. op eee 103 114 0 IBapected: fats Rha er ee aes 11 11 0 Mean winter production of 2 9 imandicatediclass. .0-..e ee ee 48.40 eggs 15.73 eggs B. With8 92 9 indicated to be of class2 = fil, . FLils. Q@ Progeny Winter Production: Over 30 Under 30 Zero Olaseiryachosn sos cian gaceebeus 274 23 0 JOAN ACAI: Rol uda aes, Gerdes gio = a) OF 0 Mean winter production of 2 9 cae Immavrcated classusemaceseeeee 54.96 eggs 17.00 eggs C. With8 2 9 indicated to be of class 3 = flils . Flile. Q Progeny Winter Production: Over 30 Under 30 Zero Obsenvied:.s-h eee ate ae 14 15 1 Bonected ss Sn Bae ie ea eee 15 15 0 Mean winter production of 2 9 in indicated class.......:.... 41.93 eggs 12.20 eggs O eggs INHERITANCE OF FECUNDITY 199 D. With1 @ indicated to be of class 4 = fIyl. . FIjl. Q@ Progeny Winter Production: Over 30 Under 30 Zero Observed perenne ele acs 5 0 1 IBCECLCO PP IA ee rae 6 0 0 Mean winter production of 9 9 im indicated sclasss cn. cl. 4- 39.40 eggs 0 eggs All Q Progeny Winter Production: Over 30 Under 30 Zero Observed ycavale we aie ssa 57 29 2 J EB ZORA MAS Clee poe pas Se oe ee 62 26 0 Mean winter production........ 48.16 eggs 13.81 eggs 0 eggs Besides the families noted above ~ D31 got one adult daughter by each of two other females. Both these daughters had a winter record of zero eggs and were apparently pathological. In any event it was impossible to form any judgment as to their gametic constitution or that of their dams. The general agreement between observation and expectation in this large progeny group is clear. The apparent exceptions to gametic expectation need some discussion. In the B matings (class 29 9) the record shows 23 in the ‘Under 30’ class where none is expected. Actually out of the 30 individuals from these matings only one daughter laid fewer than 30 eggs in the winter period. There were, however, 3 individuals which laid exactly 30 eggs in this period. So, in accordance with the convention adopted at the beginning, the record of 25 is made up as follows: 1+34+34+44 =23. The one bird under 30 witha record of 17 eggs was late hatched and probably represented a somatic fluctua- tion. This bird was bred, but unfortunately got no offspring. Her eggs were nearly all fertile but the embryos died during incu- bation. Of the two birds with a zero winter record it may be said that one (E96) was pathological, and on that account failed to lay. The autopsy on this bird, which died April 13, 1909, showed that it must have been functionally deranged for a long time preceding death. Yet there was clear evidence of functional activation of ovary and oviduct at some time. before death. In this case the 200 RAYMOND PEARL bird without question carried either L, or Ly (or possibly both) and the reproductive system started to function in the normal way and bring to somatic expression these gametie factors. But before this could be done the diseased condition of the organs brought the bird as a whole into such a condition of reduced vital- ity that egg production was impossible. The other bird’s zero record is probably a somatic fluctuation from an ‘Under 30’ hereditary constitution. She began laying very shortly after the end of the winter period. It is of interest to note that the mean winter productions ‘are relatively rather low for the ‘Over 30’ classes in all matings. The contrast between ~ D31 progeny and that of ~ 563 (vide supra) in this respect is striking. This matter will be discussed in detail later. Summary and discussion of matings of class 7 Barred Plymouth Rock males. Having now presented in detail the evidence respecting the matings of class 7 males with various types of females it is desirable to collect and summarize this material. In tables 11 to 16 inclusive are given the assembled results of all matings of certain particular types. It will be understood that these are all pure Barred Rock matings and represent the summa- tion of the data previously given. These tables give the total numbers of different males and females from which data were obtained in each class of matings, as well as the classification of the adult female progeny in respect to fecundity. TABLE 11 Showing the results of all matings of class? 7h X class 1 Q Q flile . file X flnk. . Ful NUMBER OF INDIVID- ‘ UALS INVOLVED IN WINTER EGG PRODUCTION OF ADULT DAUGHTERS MATINGS OF THIS TYPE Nea aes cee {7 fosfos 29 Class Over 30 Under 30 Zero roe z 10 | 75 Observed | 923 1033 7 203 | Expected 101.5 101.5 0 Mean winter egg production of all daughters in designated | | Classe... aes ooh Suenos | Otel meres || lb bomeras 0 eggs INHERITANCE OF FECUNDITY 201 TABLE 12 Showing the results of all matings of class? iS X class 2 9° flule 4 flily xX fIn Le 5 FIyl. NUMBER OF INDIVID- } UALS INVOLVED IN | WINTER EGG PRODUCTION OF ADULT DAUGHTERS MATINGS OF THIS TYPE) loses ge | Class | Over 30 Under 30 | Zero ee Q 9 | 38 | Observed ill 6 | 0 117 | | | Expected ipl7¢ 0 | 0 | Mean winter egg production of | all daughters in designated | Classen eet ea) aoetbG47 eggs | 20/33 eggs TABLE 13 Showing the results of all matings of class7 7o& X Class 3 2 9 fhe. flils X flvk . Falk NUMBER OF INDIVID- UALS INVOLVED IN WINTER EGG PRODUCTION OF ADULT DAUGHTERS _ MATINGS OF THIS TYPE | Total adult 9 | offspring ilos oe) Class | Over 30 | Under 30 | Zero ise | 5 19 Observed 29 23 2 |. 64 Expected a7 | a7 0 | | | Mean winter production of all daughters in designated | | GEISER ERamee aerate ee ie ee 47.93 eggs 15.30 eggs 0 eggs TABLE 14 Showing the results of all matings of Class? io X Class 4 @ file. file X flhok . FIil: NUMBER OF INDIVID- | : UALS INVOLVED IN - WINTER EGG PRODUCTION OF ADULT DAUGHTERS MATINGS OF THIS TYPE) | ~ = 7 : ] ae F | Total adult 9 vie F f g | Class Gree ay if Under 30 | fae a * offspring 2 | 38 | Observed 9 ieee oe wa 10 | Expected 10 | 7) | Mean winter egg production of all daughters in designated | GLASS eae aso riot Sena a 39.56 eggs | 0 eggs THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 13, NO. 2 202 RAYMOND PEARL TABLE 15 Showing the results of all matings of class? 7% X class 6 9 9 fils. fils X flile. Ful. NUMBER OF INDIVID- UALS INVOLVED IN WINTER EGG PRODUCTION OF ADULT DAUGHTERS MATINGS OF THIS TYPE | | Class | Total Adult @ ad | foefe) Over 30 | Under 30 | | Zero offspring 5 | 11 Observed || 12 | 164 | 0 18 | Expected | 0 18 | 0 -- = _ — ——_ — — — | = | | Mean winter egg production of | | all daughters in designated | | Classe Eee oc ct tae ASE OU eggs! 14.44 eggs | 1 The record of the single ‘Over 30’ bird. TABLE 16 Showing the results of all matings of class? io with all classes of 2 2 General Summary = : = NUMBER OF INDIVID- | : UALS INVOLVED IN | WINTER EGG PRODUCTION OF ADULT DAUGHTERS MATINGS OF THIS TYPE) a Under 30 | eG | Total adult 9 fofron 22 | Class | Over 30 offspring 11 | 146 | Observed | 243.0 149.0 | 10 402 | | Expected | 255.5 | 146.5 0 Mean winter egg production of | all daughters in designated | | Glass) Aces ee oa, ee ites so OD AON geen Loaot CLES 0 eggs From these tables the following points would appear to be definitely established: 1. The numbers of different individuals used as parents in these - matings and also the numbers of adult daughters obtained from them are great enough to give an adequate test of the hypothesis under discussion. In other words, we are not dealing here with the results of a few matings, and a small offspring series. One hun- dred and forty-six separate and distinct matings to test out males of one gametic constitution must be regarded as an adequate number. INHERITANCE OF FECUNDITY 203 2. The evidence for a definite and clean-cut segregation of high fecundity and low fecundity in gametogenesis is clear and indubit- able. ‘The expected proportions of high producers and low pro- ducers are closely realized in all the different types of matings. 3. Furthermore, the mean egg productions of the birds in the several gametic classes are widely separated, showing that the segregation is of perfectly distinct physiological entities. Refined biometric tests are not necessary to show that the birds carrying high fecundity hereditarily lay more than those with low fecundity hereditary factors. The birds in the ‘Over 30’ class have aver- age winter productions from three to five times greater than those of birds belonging to the ‘Under 30’ class.- 4. The agreement between observation and expectation for the several types of mating is as close as could be expected con- sidering the nature of the material. The only discrepancy of note is caused by the 10 birds with zero records, where none are expected. In the detailed discussions in connection with e&ch mating it has been shown, however, that nearly all of these 10 cases, when studied individually, have a physiological explana- tion, which makes it impossible to regard them as real exceptions to the gametic expectations. A determination might be made of the ‘goodness of fit’ of theory to observation by Pearson’s (42) method, were it not for the fact that that method cannot be applied to cases like the present.” ” The difficulty lies in the fact that Pearson’s test depends upon a variable ae ee Mr where mr is the theoretical frequency and m’; the observed. Now obviously in any distribution where even one m, is zero, the value of x. must be infinity, what- ever may be the values of the other m;’s or m’;’s. That is, if the theoretically expected frequency on any base element is numerically zero the probability against the whole curve becomes infinite. Thus, for example, suppose a system of fre- quencies like the following, a type which is continually arising in Mendelian work. (CUEYSI hae onda © 2 aa eham ele ie Bice © ar eee eee 3 4 5 Theoretically expected frequency... 595 827 68 0 96 Actually observed frequency....... 594 828 67 1 96 Now, it does not need a mathematical measure of any kind to tell one that in this case the theoretical and actual distributions are in very close agreement. 204 RAYMOND PEARL Further discussion of various points brought out by these tables is deferred to a later section of the paper. Matings of Barred Plymouth Rock males of class 4. Males of class 4 have a gametic constitution fL,L. . fll. That is, they are heterozygous with respect to both fecundity factors. Among the progeny are to be expected high, low and zero winter layers. Four male birds of this genotypic constitu- tion have been used in the breeding experiments. Their records follow. B.P.R. 2 569. Indicated constitution = fL,L, . fll. This male was hatched in 1909, and bred the following year. His breeding history was as follows: Matings: A. With 1 @ indicated to be of class 2 = fliLe . FIle. Q Progeny Winter Production: Over 30 Under 30 Zero Observed acn Sshtas hee ee 2 0) 0 LOG WAGIN ari oeeee ta ore Hem SO. 6 1 1 0 Mean winter egg production of ® 2 indicated class.......... 67.00 eggs Yet, because the theoretical frequency on class 4 is zero, the probability by Pear- son’s test is literally infinite against the observed distribution being regarded as a random sample of a population distributed in accordance with the theoretical frequencies. Pearson (loc. cit., p. 164, footnote) had indeed himself noted what is essentially this same difficulty in using the test on ordinary frequency distribu- tions. The point noted obviously limits greatly the applicability of Pearson’s test, and in a most unfortunate direction. Tests of goodness of fit are much needed in Mendelian work. But it is just here that classes where the theoretical frequency is zero often occur. To determine the probable error of the individual frequency in measuring the goodness of fit of Mendelian observation and theory, as was first practised by Weldon (52) and later by Johannsen (21) and by Mendelian workers generally, does not appear to the writer to be an altogether sound procedure. It fails to take account of the correlations in errors amongst the several frequencies. Yet these are just as important and just as certainly existent in a Mendelian ‘cate- gory’ type of distribution as in the ordinary variation polygon of a continuously variable character. This point I have alluded to elsewhere recently (Pearl, 32). Pearson’s test covers this point, and were it not for the other difficulty noted above would be much more widely useful in Mendelian work than is actually the case. INHERITANCE OF FECUNDITY B. With4 9 2 indicated to be of class 6 = fli, Ly . Fhe. Q Progeny Winter Production: Over 30 Under 30 Zero Obsenviedesstce ss sc. asus ees 2 6 3 (EDDC CLEC Ae I oe Par ee 2.75 5.5 2.75 Mean winter production of 2 @ imindicatediclass':...-......- 75.00 eggs 7.33 eggs 0 eggs C. With 4 2 Q indicated to be of class 1 = fly L2 . Fhile. Q Progeny Winter Production: Over 30 Under 30 Zero @bsenvedtie nme ce cin ass De 63 EDD EGUCO arte re cece 2 4.9 6.5 1.6 Mean winter production of Q@ @ in indicated class. | ......... 44.60. eggs 8.00 eggs 0 eggs D. With3 9 2 indicated to be of class 4 = fll, . FIyls. Q Progeny Winter Production: Over 30 Under 30 Zero Obsenvieds saosin dela sos e 3 3 0 BEGD CCLEU eet MOR oe 3 3 0 Mean winter egg production of @ @ in indicated class........ 45.33 eggs 7.33 eggs 0 eggs All 2 Progeny Winter Production: Over 30 Under 30 Zero Observedeee eee ne ee 123 154 4 THE ANACUAS sc ccw eon otage ae pete 11.65 16 4.35 Mean winter production........ 53.58 eggs 7.60 eggs 0 eggs The agreement between observation and expectation is plainly very close here. The three fecundity classes are represented and in proportions as near to those indicated by hypothesis as could be expected, considering the numbers involved. B.P.R. 2 566. Indicated constitution = fL,L, . fll. This bird was used in the breeding pen in the season of 1910, having been hatched in the spring of the previous year. - His sire was « D556, a class 4 male to be taken up later, and his dam a class 2 female. His breeding history was as follows: 206 . RAYMOND PEARL Matinas: A. With5 2 @ indicated to be of class 1 = file . Flile. Q Progeny Winter Production: Over 30 Under 30 Zero Obsernveds! ies. oe ee ee 4 8 2 DEH NAOT hao big c slowne 4.'3 Giaicobee & 5.2 7 1.8 Mean winter egg production of 9 9 in indicated class........ 35.00 eggs 20.50 eggs 0 eggs B. With6 2 9 indicated to be of class2 = fliL2 . FIle. Q Progeny Winter Production: - Over 30 Under 30 Zero Observed 2 Sins 2-6 ae 9 6 0 BE pected ween aay eee eee 7.5 7.6 0 Mean winter eggs production of 2 9 in indicated class........ 50.44 eggs 11.83 eggs 0 eggs All 9 Progeny Winter Production: Over 30 Under 30 Zero Obsenvedesswcckass see eee 13 14 2 Biecp CCted ss: coe ta eer eee Deep 14.5 1.8 Mean winter production........ 45.69 eggs 16.79 eggs 0 eggs Here again the agreement between observation and expecta- tion is very close, quite as close as could be expected with the numbers involved. The mean production of the 4 birds in'the ‘Over 30’ class in the A matings is low. } B.P.R. ¢ D35. Indicated constitution = fliL. . flilr. This bird was one of the original males with which the present breeding experiments were started in 1908. The only thing known about his ancestry is that he was the son of a hen laying 200 or more eggs in the year. He got only a small adult female progeny, and was used as a breeder only one year. His breeding — record follows. Matings: A. With 4 9 @ indicated to be of class 1 = fliL2 . Fh. Q Progeny Winter Production: Over 30 Under 30 Zero Observedes7. 4 eae 3 5 1 Hapected Na8 sa sees Mat eee Sey 4.6 il Mean winter production of 2 @ imandicated! class. em sas er 58.67 eggs 15.20 eggs 0 eggs INHERITANCE OF FECUNDITY 207 B. With 2 2 2 indicated to be of class 3 = fIilz . Flys. 2 Progeny Winter Production: Over 30 Under 80 Zero @bsenvedsatsessde nas ee sh oe 2 5 0 PRE DECLEU een SP Es, cic 2:6 3.5 0.9 Mean winter production of 2 9 imamndicabediclassseccc-ce--- deen wees 0 3 1 iBamectediutt: Soars seo k eee 0 fy) 1 Mean winter production of 2 Q Imundicatediclasss-.-7sees soa 6.00 eggs 0 eggs All Q Progeny Winter Production: Over 30 Under 30 Zero Obsernvedeay tes a es 4 17 3 Expected.. EOE Cue eos. 18 6 Mean winter Penducnos joousas DLE eas 14.06 eggs 0 eggs Matings of CJI.G. 3» 657. Indicated constitution: class 3 (CaAnGS) of = fll, eplib: Matings: A. With 6 B.P.R. 2 9 indicated to be of class 2 = flil2 . Flals. Q Progeny Winter Production: Over 30 Under 30 Zero Observedia.= tow haere 0 17 0 Expected.. a Nie Aa 0 if 0 ‘Mean winter Reeoaccenn ine 9 9 in indicated class...........: 12.88 eggs INHERITANCE OF FECUNDITY 229 B. With 1 B.P.R. @ indicated to be of class 4 = fIylz . FIle. Q@ Progeny Winter Production: Over 30 Under 30 Zero @ bsenvedeer area cat aes coc 0 1 0 Expected. . nee shee aw 1 0 Observed SL ae pr pdaehion 10 eggs All Q Progeny Winter Production: Over 30 Under 30 Zero Observed ceases ce lies 0 18 0 Expected... ep ieet eee: 8) 0) 18 Mean w er pr snathnattien adetieisce a0 0 12.72 eggs The accordance between observations and expectation here is perfect. Mating of CI.G. 2 529 . Indicated constitution Me TEG:. ) = flak. s flile. sielassat2 Matings: ! A. With 1 2 indicated to be of class 1 = fly L2 . File Q Progeny Winter Production: Over 30 Under 30 - Zero Observedeas ati e steel 1 1 1 Expected. . Beha eeeters AO 2.25 0.75 Observed eigen srothatien., 46 eggs 13 eggs 0 eggs This mating by itself is, of course, without any particular sig- nificance. B. With 6 2 @ indicated to be of class 2 = fli L2 . Plils @ Progeny Winter Production Over 30 Under 30 Zero ODSenVede stern teen ts 5 12 Expected. . : 0 0 Be 17 0 Mean winter ferodme nian! af all ® @ in indicated class 41.60 eggs 11.67 eggs Here, again, as in the case of 7 558 there are seen to be several birds with winter records of over 30 eggs, when none is expected 230 RAYMOND PEARL C. With 3 2 Q indicated to be of class 7 = fll . Fhe. Q Progeny Winter Production: Over 30 Under 30 Zero Observed! st. 45.0 An eke ace 1 3 2 Expected... Dae eRe oe ee 0 3 3 Mean saline prodncrion of all 2 Q in indicated class........ 45 eggs 13.00 eggs 0 eggs All @ Progeny Winter Production: Over 30 Under 30 Zero Observedaniwes Bes) exes Gi 16 3 UOLANIOT AI nak epee Ree ae kre 0 22.25 3.75 Mean winter production........ 42.71 eggs 12.00 eggs 0 eggs The seven birds with records ‘Over 30’ belong gametically to the ‘Under 30’ class, and their records are somatic fluctuations. This is shown both by their history and by their behavior in F’,, all having been bred. Matings of CI.G. 2 578. This male has been shown from his matings with pure Cornish females to belong to class 3 of C.I.G. (= fll. . fil,). His matings with Barred Rock females are as follows: Matings: A. With 2 B.P.R. 2 2 indicated to be of class 2 = fliLy . FIle. Q Progeny Winter Production: Over 30 Under 30 Zero Observed tea ace ves pee 0 3 ) Expected. . eee ed) 8 0 Mean “dha Prono on . datuchtercskie see) eer tee 6.33 eggs B. With 1 2 indicated to be of class 4 = fll. . FIyls. Q Progeny Winter Production: Over 30 Under 30 Zero Observedera cute eae eee 1 4 0 Ee cpected ease ace tetera en, 5 0 Mean winter production of daughters in indicated class.. 42 eggs 11.00 eggs O eggs INHERITANCE OF FECUNDITY Zou All Q Progeny Winter Production: Over 30 Under 30 Zero Olsen diame ecu cenn cists ce 1 7 0 GE DCCLEGMS RRO Mn IAT ra er OO 8 0 Mean winter production........ 42 eggs 9.00 eggs There are no data from which to make sure whether the one bird with an ‘Over 30’ record represented fluctuation from the ‘Under 30’ class. It probably did, but this cannot be positively asserted. Summary of all F, matings Putting together the results of the matings of all Cornish Indian Game males with Barred Rock females, we have for the actual observations: Q Progeny—Raw Data Winter Production: Over 30 Under 30 Zero @bsenvedhin, pet sack he 12 58 6 LEED CCLEOM Ee ae tO 66.25 9.75 Mean winter production........ 45.67 eggs 12.46 eggs 0 eggs In view of the fact that the 11 of the 12 birds with ‘Over 30’ records represent somatic fluctuations from the ‘Under 30’ class it is desirable to present another summary table in which the progeny are distributed in accordance with their gametic consti- tution. Q@ Progeny on Gametic Basis Winter Production: Over 30 (L,L2) Under 30 (Iyl2) Zero (lls) Obsenveds sees oes et 1 69 6 ECCLES EL ae! 66.25 QMS The contrast between these distributions and those of the reci- procal cross discussed before is very striking. Taken together these reciprocal crosses support strongly the general hypothesis of fecundity inheritance here being tested. Da? RAYMOND PEARL Matings of the second cross-bred (F»2) generation The F’, birds discussed in the preceding sections were mated in all possible ways inter se and with the parent forms. The results of these matings will be discussed in the present section. At the outstart it should be noted that in spite of the fact that as many F’, birds were hatched and reared as the available facili- ties would permit, nevertheless, the number of adult daughters available for fecundity study is small in case of some of the mat- ings. ‘There are several reasons for this. Besides the obvious one such as mortality, depredations of thieves, hawks, crows, rats and the like, there is another important but not so obvious one. This is the failure or great difficulty experienced in getting certain of the F’, eross breds to grow into normal, full-sized, healthy adult birds. After rather wide experience in handling ‘eross-bred chicks, I am convinced that certain gametic combina- tions which are to be expected on Mendelian theory, and can be produced in the expected numbers in the breeding pen, are never- theless physiologically abnormal or unsound. Such birds do’ not make a normal growth, but in spite of the best care and atten- tion grow up into stunted weaklings, which always show, both in their structure and their physiological economy, the effect of this retarded, abnormal development. JI am further convinced that this result is primarily due to the hereditary constitution of the individuals in question. Certain combinations of hered.tary factors do not produce physiological sound and vigorous zygotes. Of course, there is nothing novel in such a result. It is of a piece, for example, with the parts respecting the relation between hereditary constitution and physiological vigor in maize, which have been so clearly set forth and analyzed by Shull (45, 46, 47) and East (9). Other examples of the same phenomenon might be cited. The whole phenomenon is precisely what would be ex- pected from Jobannsen’s general conception of inheritance and ontogeny (22). . This relationship between hereditary constitution and physio- logical constitution or normality takes on particular significance when one is dealing with fecundity. As has been pointed out INHERITANCE OF FECUNDITY Zao earlier in this paper one cannot expect to get a normal somatic expression of the hereditary constitution in respect to fecundity unless the bird is a physiologically normal, well-developed individ- ual. Stunted, under-developed, or physiologically unsound birds will lay but very few if any eggs, regardless of what fecundity factors it may carry. A marked difference is here apparent be- tween structural and physiological characters so far as the study of inheritance is concerned. A definite structure either is or is not present in the zygote, however weak physiologically the indi- vidual may be. But if the general capability of an organism with respect to the transformation of matter and energy is markedly reduced, then all physiological characters will be affected, and fail to reach complete normal expression. In the study of cross-bred poultry I have found pure extracted whites from crosses involving originally two heavily pigmented parent races to be conspicuously good examples of the phenome- non under discussion. It is only very exceptionally, in my experience, that such white birds are physiologically normal. Indeed because of this fact it is only with the greatest difficulty, and after many failures, that I have been able to get such extracted whites to breed, and thus form a pure white race. If the hens lay eggs, which some do not do, they are usually either infertile, or else all the embryos die at an early stage. These facts have some bearing on the popular belief of animal breeders that whites in general are delicate in constitution and hard to rear. This belief is so well known that it is not necessary to cite in detail references regarding it in the literature. As a consequence of the above considerations, I have felt justified in leaving out of account, or rather in considering. apart from the others, a few of the F’, individuals, in all some 7 out of - over 200 F’, birds all told. In each case these birds were physio- logically abnormal, and obviously so to the most casual observer. The fact that they did not lay was no criterion whatsoever of their hereditary constitution. In order that there might be no possibility of unfairly influencing ratios by leaving these birds out, the whole families (usually of two or three individuals only) to which they belonged have been rejected. As a matter of fact THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 13, NO. 2 234 RAYMOND PEARL whenever one individual in a family is physiologically abnormal in this way all the other members will usually show the same con- dition in greater or less degrees. In the F, generation following the reciprocal crossing of Barred Rocks and Cornish Indian Games there are a number of possible matings. The nature of these matings and the results as to color and pattern have been discussed in another place (41). That paper may be referred to in case one is not clear as to the nature of the matings. The different matings will be discussed in the following. order. 1. fF, -¢ (out of BPR. a eeCu.G:e 2). x iia? of barred and non-barred. 2. fF, ¢ (out of C.L.G-@ X B.P-R: ?@) x Fixe es banmedvang non-barred. Balla. cv e(Outeor BsPAR seme wo sk.G. 0). iB Rane JF oatout of BPR oe CAG. 9) i CauG ie ic (out of (CG oie B: PR: 2) 2GiBaR Re oO pH gout01 @aA- Gano «ib. Riko) SG@meG? re B.P.R. 2% X F, Barred ¢ ¢ B.P.R. a¢ X F; non-barred ¢ 9 9. C.I.G. 2 X F,:2? 9 Barred and non-barred It will be recalled that the barred F, females come from the mating B.P.R. 7 x C.I.G. 9; and that the non-barred (black) F, females comes from the reciprocal mating C.I.G. 7 B.P.R. ¢. Matings of F, 2 576 with F, females. The pedigree of F, 7 576 was as follows: iBze Re of 559 (fl: Le : fl L2) x CrliG: Q 456 (fll, < Fixls) . | 40 410 +40 140 F, 3 576 The hereditary constitutions of both 7 559 and ¢ 456 were known, both from their pedigrees and their progeny in other . matings. From this pedigree it is evident that the gametic formula for ¢ 576 must be either fl,L. . fll, or fll . fli. A study of his progeny in all matings shows clearly that it is actually the former. In other words he produced gametes of four kinds, viz., flils, fl, Le, fIiile, flr. INHERITANCE OF FECUNDITY 235 This bird was mated with barred F 9 @ which had been pro- duced by the following mating. B.P.R. o@ 559 (fil. . fll.) x C.1.G. 99 of type flu . Fh { A. fl, LZ. . FL,l,-Winter record over 30 and B. fll. . Fil, -—Winter record under 30 Male 576 was mated only with 9? ¢ of the A class,“ namely those having a gametic constitution fl,L. . FL,l., and producing two sorts of F-bearing gametes, Fljl, and FLyl. This same # 576 was mated with non-barred (black) Ff’; females which had been produced in the following way. ‘B.P.R. 992 some of which were of type fl,L, . Phils, and some fll, . FL,l. (both x ; , | producing F'-bearing gam- | CEGae 008", | or 529 (fll: . fli) J etes of two kinds FLil, L and Fil.) A. fil, . FL,l, Winter record under 30 B. fll, . FLjl, Winter record under 30 C. flyl, . Fl,l., Winter record under 30 D. fll, . Fl..l., Winter record zero. The three non-barred ¢ @ with which ~ 576 was bred were of the B-C type, producing two kinds of F-bearing gametes, P'L,l, and Fl,l,. They were thus identical, so far as concerns F’-bearing gametes with the barred F, birds with which 7 576 was mated. All the progeny may then be treated together since all did, as a matter of fact, lead to the same result, having regard to the errors of sampling in such statistically small lots. 14 It would of course be desirable to have data from the other mating, o 576 X B 2°. If one could have foreseen what the mechanism of the inheritance of fecundity was going to turn out to be, such matings would have been made. Actu- ally these cross bred birds were being studied primarily with reference to color characters, and the matings were made relative to that line of investigation. _ Naturally highly fecund females would be chosen as breeders whenever possible, in order to get more chickens for the color studies. Actually, however, all of the possible gametic combinations in respect to fecundity were tested in F, either from one mating or another. 236 RAYMOND PEARL TABLE 27 Showing the results of mating F, & 576 with F, 99° Type of mating: fll, . flule * flily.FIyl, and (or) flil, . Fils INDIVIDUALS USED IN THESE MATINGS WINTER RECORD OF DAUGHTERS a 9 Over 30 | Under 30 Zero ease 576 F68 4 2 0 6 576 | F89 0 3 0 3 576 F41 1 3 0 4 576 F79 1 3 1 5 576 | F416 0 0 7s 2 576 | F421 0 2 2 4 576 | F33 5 3 2 10 576 F415 0 4 1 5 Motallobserviedi: 4.4455" 11 | 20 8 39 Motal Bxupected.s...2:.. 23 14.6 | 19.5 4.9 Mean winter production of | 9 9 in indicated class....| 42.81 eggs | 12.05 eggs 0 eggs 1 These two individuals ought really to be excluded on the ground of physio- logical abnormality of the sort discussed at the beginning of this section. | Neither of them made a normal growth. No poultrymen would have regarded these birds as reliable material for the study of egg production. Leaving these two birds out the totals stand as follows: Winter Production: Over 30 Under 30 Zero Observede pest eee eee mala 20 6 J Ha DT ONEG baer Bes tera eee a lore. iksyoe) 18.5 4.6 The same kind of evidence for the segregation of different degrees of fecundity which has been seen in all the previous mat- ings appears again in these F, birds. Matings of F, 3 577 with F, females. This FP, 2 577 was pro- duced in the following way: C.1L.G. 7 558(flil. . fl) x B.P.R. 9 234 (fil, . Fil) | F, 2 507 Such a mating as this would be expected to produce males of four (really three different) kinds as follows: A. fly, . fIiite C. flile . flute Big plone icles D. flle . fliLe INHERITANCE OF FECUNDITY Dam The results indicate that « 577 was of the last (D) type, pro- ducing two kinds of gametes, fll, and fl,L.. He was mated with 4 barred Ff, ¢ 9 and 4 black F,¢ 9. All of these females, as in the preceding case, produced F-bearing gametés of two kinds in equal numbers; Fil, and FL, l.. Of these matings three pro- duced small families in which all of the individuals were so far from being normal physiologically that they cannot fairly be included in the tabulation. The details regarding them are as follows. From one barred F; @ was produced two adult daugh- ters, both of which were undersized and stunted in development, and failed to lay. One of these daughters died early in the year. From one of the matings with black F, 2 ¢ only one adult daughter was obtained, which again failed to develop normally and was only put into the adult house because of its interest from the standpoint of color inheritance. Another of the matings with a black F, ¢ produced four adult daughters. Two of these were extracted whites and very small, poor specimens. The whole family was saved because of these birds. Neither of them laid. Of the other sisters one died early in the laying year, never having laid. It, like the other members of the family, was from the start a weakling. Finally the fourth sister made a winter record of 8 eggs. It presented the same evidence of abnormality as the other sisters, and its egg record could by no means be taken as a just indication of its gametic constitution in respect to fecundity. No one of the seven birds under discussion would ever by any chance whatever have been put in the laying house as normal individuals for the study of fecundity. The only reason they ever were put in was simply, as already explained, because the primary object of the F’, birds as a whole was the study of color and pattern inher- itance. Even though a bird is an undeveloped weakling physio- logically one may make a record of its plumage color and pattern, and see whether these change with advancing age. However, since these birds really were in the adult house, and in order to forestall the possibility of a suggestion that any records were suppressed in this study of fecundity, it has seemed advisable to take the space for the above detailed discussion of the matter. 238 RAYMOND PEARL The records for the other matings of Ff; ~ 577 are given in table 28. Matings of F, 2 576 with Barred Plymouth Rock females. This F, # was mated with three pure Barred Rock 9? ¢ of class 2 (table 6). The results of these matings are shown in table 29. TABLE 28 Showing the results of mating Fi¢ 577 with F, 99 Type of mating: flle . file X flile. . FInl, and flil, . Fil. Gametes: fll, and fl,Ls. F-bearing gametes: FLyly and Flyls INDIVIDUALS USED IN THESE MATINGS | WINTER RECORD OF DAUGHTERS rol : 2 | Over 30 Under 30 Zero | ates: 577 F401 | 0 3 0 3 577 F18 0 2 1 3 577 F44 2 4 1 7 577 F99 2 2 0 4 577 F418 1 5 5 11 Total observed............. | 5 16 7 28 Uiotalnempected see f: a 14 if Mean winter production of | | 9 9 in indicated class... | 36.33 eggs | 11.25 eggs) 0 eggs TABLE 29 Showing the results of mating Fi S676 with Barred Rock 2 9 Type of mating: fliLz . flile X flaLy . FLile INDIVIDUALS USED IN THESE MATINGS | WINTER RECORD OF DAUGHTERS Mean winter production of 2 9 in indicated class....| 46.78 eggs | 22.83 eggs rot | fe) Over 30 | Under 30 Zero ree 576 1107 1 1 0 2 576 F115 1 2 0 3 576 F1ll1 73 Bo 0 11 Motaliobsenvedsasss. eer gi | 64 0 16 Niotalken pected ae ee 8 | 8 0 | INHERITANCE OF FECUNDITY 239 These results are suggestive in connection with the problem of the absolute fecundity value of the same genes from different sources, a matter which will be fully discussed later. The figures give clear evidence of Mendelian segregation of high and low fecun- dity. | _ Matings of Fy 2 576 with Cornish Indian Game females. There were but two matings of 7 576 with pure Cornish Indian Game ¢ ¢ which produced adult female offsprings. Other matings were made but got no adult progeny. The successful matings were with C.1I.G. 9¢ of constitution fil, .Fll.. ‘The results were as follows: Gu SIG flils .flals < flights @ Progeny Winter Production: Over 30 Under 30 . Zero Observed sna ets 0 3 1 0 2 0 Motaliobserveds.. 25. s.06- 0 5 1 MO ges CVA sac oobe caecac. leo 8 1.5 Mean winter production of 2 9 im indicated class...........- 11.80 eggs 0 eggs The numbers here are too small to give definite results, but there is nothing incompatible in the observations, having regard to the smallness of the numbers, with what would be expected. Matings.of F, 2 577 with Barred Plymouth Rock females. ‘There were three matings of this sort, but the families were all small. The females used were of class 2 (table 6). The results were as follows: SUT flils flile X flab, SEG Q Progeny Winter Production: Over 30 Under 380 = Zero (COMPETI ere 8 ce irae om aoe nae 0 1 0 (COP RRO S S85 ope ain ae te oR neem br 15 0 GBH 235) Race rer: 1 2 0 Wotalkopsenvedase sheen - 24 ‘ 4 0 Motal expected). .y.c2. aes 2, OO 3.5 0 Mean winter production of 9 @ imundicatedkelassy. + ..252 ee 35.00 eggs 11.25 eggs 240 RAYMOND PEARL Again the numbers involved are too small to be of any particular significance when taken by themselves. They are in conformity, however, with all the other data and therefore have cumulative value. Matings of F, 2 577 with Cornish Indian Game females. Two matings only of this sort got adult female progeny. The families are small. The pure Cornish females used were of constitution flyl, . Flhly. The results follow. Gi 517 file flula: X flalee Pls os Progeny Winter Production: s Over 30 Under 30 Zero (COMBE eee ee en ee 0 1 1 (eH) oad aba parason ¢ 0 2 1 Total observed....-...:.-- 0 3 2 MotaikenDeCied anes tae eee. 2.5 1.25 Mean winter production of Q 9 ral ibaKo WORMHEYe CES Sse op oedasoce 7.33 eggs 0 eggs Matings of Barred Plymouth Rock males with Barred F, females. While several matings were made here they all fell into one or the other of two gametic types. The Barred Rock males used in these matings have, of course, already had their gametic constitu- tions determined through their matings with pure Barred Rock females. A. Type of mating: fll. . fll. X fll. . FLIjl. The results of this type of mating are shown in table 30. There are several points which need to be noted about this table. While in general it is apparent that the observed result falls out in fair accord with expectation, the three zero birds are outstanding exceptions. No zero birds should occur in any of these matings. RESPONSE * EGE) cc. per liter percent ++ | percent — per cent « | | 6.49 | te 90 10 yl OF Gir teeSOihat | gets 6.49 89 11 0 | 80 | 18 ytd | 89 10 1 80 | 18 6.90 91 9 0 80 | 18 6.90 Ca mee eh MSO tee or dls 7.00 es Od 9 Os by leeueo 18 7.40 | 90 8 Pe i tt 15 Average positive response 89.4 per cent. Greatest deviation from average 4.9 per cent. *1 indicates a positive reaction. — indicates a negative reaction. « indicates an indifferent reaction. regarding moulting, and were entirely normal in every way. Throughout their life they had been kept in still water having an average of about 6 ec. of oxygen per liter. From table 1 it will be seen that with animals in approximately the same physiological state, the experimental error of the method used is almost 5 per cent. This error is too large for purely quan- titative results, but it will not interfere greatly with the compari- sons that are to be made in this work. In no case however is any importance attached to experimental results that do not show a difference of at least 20 per cent, so that the possible error of 5 per cent cannot affect the conclusions drawn. 278 W. C. ALLEE The results of another check on the constancy of the rheotactie response to a circular current are given in figure 1, which shows eraphs of twenty-five successive responses of one male Asellus. The animal used was a stream isopod 11 mm. long, that had been Fig. 1 Twenty-five successive rheotactic reactions of one stream Asellus kept in water with the oxygen at air saturation. In this case the trials were made in a glass dish 10 cm. in diameter. The current used was strong enough to sweep the isopod to the center each time and the presence of the spiral current is plainly shown by the path taken in reaching the circumference of the dish.. RHEOTAXIS IN ISOPODA 279 In the figure the starting point is indicated by a large dot; the end point, by a cross. The arrow on the outside of the circle indi- cates the direction of the current. In all cases, excepting no. 23, the animal responded positively to the current, that is, it gave a 96 per cent positive response. Evidently its normal reaction was a positive one and the one fail- ure to go against the current would then mean a chance turning that was not corrected. That is, in this case there is an experi- mental error of 4 per cent, which checks well with that shown in table 1. These results Hes show that the isopods will give their normal response for at least twenty-five successive trials, so that it is entirely safe to take the first ten responses as S URGISENS the normal behavior of the animals tested. During these experiments the oxygen content of the water has been determined by the Winkler method. During the first part of the work, the method was followed as outlined in the report of the committee on standard methods of water analysis tothe laboratory section of the American Health Association (’05, pp. 74-77). The free carbon dioxide was determined by direct titra- tion with N/22 sodium carbonate using phenolphthalein as an indicator. This method is described in the same Yrepert (pp. 72-73). After the appearance of Birge and Juday’s work (’11 pp. 138-24), their methods were followed wherever the technique appeared better. Birge and Juday compared this method of determining the oxygen content of the water with that of boiling and found (1. ¢., pp. 11-12) that the amount of variation in results from the two methods was not more than was the case in dupli- cate determinations by the same method. V.E. Shelford and the writer verified these results and the methods used have proved eminently satisfactory for rapid biological work. 280 W. C.. ALLEE II. RHEOTACTIC RESPONSE UNDER NATURAL CONDITIONS 1. STREAM ASELLI a. Normal adults Table 1 and figure 1 show responses of stream isopods at their highest rate of positiveness. Usually under the conditions used the per cent of: positive response was somewhat lower. A list of the results of these tests is given in table 2. It will be seen that the results of different trials vary somewhat, the greatest varia- TABLE 2 4 Normal rheotactic response of adult stream Aselli Oo | RESPONSE | Silage OF | cadet ce. per liter per cent + | per cent — percentx | | je | 72 20 | 8 COs 1 5.46 | 78 12 10 50 17 5.5 | 70 12 | 18 50 18 5.69 78 | 12 10 50 | 18 5.92 70 24 | 6 80 18 D.97 70 18 | 12 60 i? 6.08 ; 80 20 | 0 30 18 6.03 | 88 6 | 6 60 18 6.15 76 14 | 10 80 18 6.26 | 80 14 | 6 50 20 6.36 75 25 0 40 17 6.49 89 e 8 3 60 18 6.55 93 7 0 80 12 6.57 | 72 18 10 50 19 6.77 | 86 13 1 70 15 7.00 88 6 6 50 23 7.02 72 8 20 50 13 7.61 82 17 1 80 10 7.54 73 27 0 | 30 18 8.14 | 79 | 21 | 0 80 18 8.14 | 85 | 11 | 4 80 17 8.14 | 83 | 12 5 80 17 8.59 86 | 4 | 10 50 16 9.22 | 84 | 4 | 12 50 6 9.31 | 78 | 8 | 14 50 6 9.34 | 74 10 | 16 50 8 10.25 78 18 | 4 50 11 6.41 79 14 | 7 1570 RHEOTAXIS IN ISOPODA 281 tion being 18 per cent from the mean. These variations may be due to length of time since moulting or age. The group giving the highest positive response had been carefully selected to repre- sent animals in the best possible condition. In all the other eases the animals were simply picked at random from the general stock. The variation in oxygen content of the water within the limits given in table 2 does not appear to have any marked signi- ficance although in general the lowest positive responses are found with the lowest oxygen content. Apparently there is sufficient oxygen here for the usual activities to be carried on at the usual rate and not enough to stimulate them greatly. The length of time the animals have been in the laboratory does not affect their response providing they are kept in conditions resem- bling as nearly as possible, those in their usual habitat. The response of these stream isopods to the current is vigorous and usually definite, that is, they are either definitely positive or negative. When the current is reversed they tend to reverse within the first ten seconds after the new current is set up, and often they are all reversed before the stirring is stopped. They also move vigorously, sometimes completing two circumferences of the pan in the minute reaction time allowed. This would mean a rate of about 80 em. per minute. They sometimes pivot on their posterior end, turning their head in a complete circle before starting a definite reaction and thi& testing reaction is usually followed by a positive response. In all these cases the amount of free carbon dioxide in the water was very low. That is, it never exceeded 3 cc. per liter of water and usually ran much lower, the average being about 2 cc. per liter. Thus the variations of the free carbon dioxide are too small to be of any significance, and the fact that it was present in such small amounts will have to be considered in determining the cause of the high percentage of positive reactions. b. Breeding season The breeding season of these isopods begins before the ice is out of the water in the spring. It reaches its culmination by the last of April and gradually diminishes. Occasionally breeding THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 13 NO. 2 282 W. C. ALLEE occurs throughout the summer but is very rare during August and September. As the weather becomes colder, a new period sets in, but this is of much shorter duration than the spring periods. The shortening cannot be wholly due to the increasing coldness because animals brought into the laboratory and kept under normal con- ditions do not long continue breeding. In these animals in the laboratory, however, a new period of breeding begins about the first of December. Staiting with a few individuals it slowly increases in importance until by the middle of January, it is the dominant activity of the animals. Curiously enough this is much more pronounced in animals kept at temperatures about 5°C. than in those at 20°. In these laboratory isopods the season stops about the time it is reaching its height in the field. One sign of the approach of the breeding season is the increased tendency to collect in bunches. Bunching is apt to occur at any time during the year if conditions become unfavorable, as when there is a sudden drop in temperature, but the bunching tendency of the breeding season is even stronger. Often these close irregu- lar groupings occur, containing six or eight individuals. This is especially apt to happen when the animals are stirred in a current so that they are thrown against each other. The copula- tion occurs much as Holmes has described for amphipods (Holmes 03, p. 288). The females may become quite helpless as the brood pouch develops and unless they are clinging to some support, they are often brought to the surface and float around ventral side up, entirely unable to right themselves or to regain the bottom unless they chance upon some solid object. The effect of the breeding season on the rheotactic response of stream Aselli is shown in table 3. The first part of the table gives the cycle of reactions due to the breeding season as shown in the field experiments from April till October. The second part traces the progress of these influences upon laboratory stock dur- ing the winter months. One of the most noticeable changes in the rheotactic response is the marked decrease in the percentage of positive responses. Another almost as striking is the extreme variability in animals selected at random from the breeding stock. This variability is not so apparent when the same animals are é TABLE 3 Part 1. Field trials. The effect of the breeding season on the rheotactic response of adult stream ‘Aselli Oz RESPONSE see PRE MOLL REMARKS cc. per liter) per cent + | percent — | percent « | 9.31 38 14 48 50 13 4-6 By) Lose Pols 7.80 | 9 74 17 80 9 492 |494¢ 7.08 48 45 7 60 9 4-22 | all 9 7.68 33 60 a 40 10 4-22 | all 9 ADD 40 A 18 50 10 4-22 | all 9 WesCo 36 60 4 70 14 4-29 | all with b.p Uf tbo} 12 44 44 50 14 4-29 | all & small Wee8 50 32 S| 50 14 4-29 | all o& large Tees Sone | ss) | 10 50 14 5-16 | mixed 7.28 48 52 0 50 14 5-16 | mixed 1.02 14 80 6 50 18 5-21 | mixed; no bp. 1.02 | 11 63 26 | 50 18 5-21 | 5 prs. cop: 5.30 | 42 | 52 6 50 18 o-21 |491¢ AO! | 48 52 0 40 18 5-21 | mixed 4.65 0 70 30 | 20 18 5-21 | 2 prs. cop. 5.63 60 40 ONES ieeelO 23 6-8 | Q 6.68 68 24 8 50 23 6-8 mixed §.01 | 53 42 5 45 16 6-17 |} 822¢ 6.26 14 82 4 50 23 7-8 mixed 7.00 | 88 6 6 50 23 8-1 normal 5.46 78 12 10 50 17 10-4 normal 8.54 | 50 46 4 50 13 10-5 selected with | b.p. i *b.p. stands for females with brood pouches. Part 2. Breeding season im laboratory stock 6.05 |) ) 67 40 She 30 ig | 124 | copulating | 1912 7.84| 650 47 Se are GO) 4 1-11 | 1 pr. cop. 9.54 | 44 20 coun || 50 6 1-19 | mixed GbL Poe 12 34 - 50 8 1-19 | mixed 9.54 53 33 14 ,.| 50 10 1-19 | mixed 9.54 60 20 20 20 8 1-23 | mixed 9.54 70 10 20 | 20 8 1-23 | 2 prs. cop. 10.47 28 51 Di x80 4 1-28 alllarge 7 8.73 | 21 23 56 80 4 1-30 | mixed 7.83 | 30 29 41 80 4 1-13 | mixed 8.88 27 30 43 70 6 1-31 | mixed 9.11 29 16 Doe uly.) 100 6 22 | mixed 8.82 27 23 0) | 70 4 2-4 | mixed 8.48 29 Das. | 48 70 6 2-8 mixed 7.97 al Sie Me aa a 70 5 2-13 | mixed 8.48 BY 33 a 60 5 2-17 | mixed 284 W. C. ALLEE tested from time to time. Thus the last trials recorded in part 2 of the table were all made on the same group of animals, with the exception of two cases the oxygen content of the water was high and in all cases the amount of carbon dioxide present as free car- bon dioxide was very small, so that these changes in reaction are not due to a modified gas content of the water. It will be seen in later experiments, that the degree of positive- ness depends upon the metabolic rate of the animals. That is, in animals having a high rate of metabolism there is a high ° positive response. From this point of view the decrease in the positive responses would be accounted for by assuming that the animals are in a state of lowered metabolism during the breeding season. This view is supported by those plants and lower animals that reproduce asexually during conditions favorable for growth and respond to poorer external conditions by sexual reproduction. The cause of the variation in response is not impossible of solu- tion although at present it cannot be treated entirely from the experimental side. From the results given it is evident that during the breeding season not all isopods are in the same physio- logical condition at any one time. The results from any one day as listed in the table show wide variations, yet these were taken from almost identical external conditions, so the variable quantity in this case must be an internal one. This view is further sup- ported by comparing those results where the animals were in the copulating position. This term is used in the table to show when the females were being carried by the males. There are three cases given when all the individuals tried were in this position. Of these one gave a positive response of 11 per cent, another of 70 per cent, and the third gave no positive response at all. The first were taken from a very low oxygen content and as will be seen later this tends to decrease the positive response. Yet with this added complication they gave a much higher response than did two pairs taken the same day from another place in the same stream. It may be that in the case where no positive re- sponse was given the animals were near the actual copulation time, and the animals giving the 70 per cent response may have been far from this period. RHEOTAXIS IN ISOPODA 285 The large brood pouch on the ventral side of the females offers a serious mechanical obstacle to making progress against a strong current, but since the same reaction tendencies occur in the males this mechanical hindrance cannot be the only faetor in the re- sponse. The condition of the germinal glands during the breeding season and the exact connection existing between their activity and the rheotactic response will be presented in detail in another paper. In comparison with the normal behavior, the large increase in indifference to the current is remarkable, since this is just the con- dition that tends to prevail in pond animals. The action inthe current is also decidedly different. The animals are much more easily swept from their footing; they do not reverse so rapidly as the current changes, often failing to reverse at all and the speed of reaction is greatly lowered. In some cases there is no response at all; the animals are then in the same state as that caused by the strongest depressing agents. This breeding behavior brings up some niteeeeane points in the ecology of the isopods. Although they are taken in streams they are rarely found in rapid parts of permanent ones, being limited for the most part to the pools and protected places. Out- side of the breeding season they are fitted by their positive reac- tion to the current and their strong clinging ability to maintain themselves in much stronger currents than those in which they are found. The reason for their absence in these places must be due to the influence of the breeding activities upon their behavior. This is especially significant since their period of least ability to maintain themselves corresponds to the time of the strongest cur- rent inthe stream. Hence their breeding behavior limits them to those streams where they can find ample lodging places during this time of weakened responses. c. Juvenile mores When the isopods are first liberated they are about 2 mm. in length and usually give no response to the current but cling passively to the bottom. Consistent rheotactic responses were made by the time the animals were about 3 mm. long; that is when 286 W. C. ALLEE they were about a month old. As table 4 shows, the general positive response of the young isopods is much lower than that of the adults but that the positiveness increases with age. Again there-is more variation in the positive response than seems consistent, for, as the table shows, the isopods give at times a low response regardless of size and of the external conditions here controlled. It will be noted that the increase in positive reaction was not due primarily to the oxygen content of the water because this was high when the positiveness was low. The carbon dioxide was also low as it has been in all cases discussed. Evidently the growth in size was the most important factor in the increased positive response. TABLE 4 Rheotaxis in juvenile stream Aselli SIZE Oz ; RESPONSE eee Date tae ee mm. cc. per liter per cent + | per cent — | per cent « 3.5 6.26 joa \eeecGs | 0 40 | 17 10 7.28 Suh a -¥8S 1 50 | 21 4.0 BUGS) il) CREME (84 Of 25 | 17 4.0 5.63 JOm mae 3G) ||. 9 24.04) 50 20 420 6.55 yommen 48 5 | 102") 30 17 5.0 wes waaay. SO 6 50 21 5.0 7.28 30 68 Da) 50 | 21 5.0 5.01 LOMA SO! |. “Onn 50 12 5.0 5234 e cdi, 2S aD | oo a Ona ee 5.0 6.38 1 N55 20 25 20 17 5.0 6.94 60 Bore Oat 50 | 15 5.5 5.63 19 75 Ohya) 80 23 5.5 7.06 50 Ae) Oa 60 22 BY 7.06 55 25-19) ae 80 29 6.0 5.63 16 Btn!” (oN eaeanao 23 6.0 5.63 15 Toe} 0 60 17 6.0 8.19 32 14 54 | 50 8 6.0 5.23 Hie Nh BG 3 40 21 6.0 6.57 90° 77 3: ieaeeeO 22 6.5 7.00 88 Gre | 6 50 23 6.5 1.14 92 0 | 8 50 29 6.5 1.14 60 230G) 424 3 50 20 8.0 5.23 56 42 | Dn 50 | 21 8.0 6.36 ‘in 25 Ong 40 19 8.0 4.27 90 5 5 20 | 21 RHEOTAXIS IN ISOPODA 287 The importance of this factor is emphasized by the two trials where with a low oxygen content the animals tested gave a high positive response. This is the one case in the progress of this work where the laboratory tests have failed to run parallel with the field results, for as will be seen later, keeping young stream Aselli in water having a low oxygen content kept the animals from developing a high positive response, in all the cases tried. These two high results were obtained in the field on July 4, 1911. The stream at this time was reduced to a series of small pools with no running water, and in the case of the higher response, the animals were in a very high temperature, 29°C., which may to some extent account for the difference between the two trials. Reference to this will be made in another part of this paper in connection with experimental data which may tend to clear up the case. d. Response to straight current Banta (10, pp. 467-468) described a trough which he devised for testing the response of isopods to a straight current. His appara- tus consisted of a simple straight trough in which the current was equalized by passing through a number of wire screens before it reached the experimental part of the trough. He introduced the isopods to be tested into still water and then turned on the cur- rent. After some crawling back and forth the animals collected at the upper end and stayed there from fifteen minutes to two days, afterward reversing their reaction. In order to test the efficiency of the pan response as an index of the rheotactic activity of the isopods, Banta’s experiments were repeated. A different type of trough was devised and is shown in figure 2. This trough has a rounded well 10 cm. in diameter. Fig. 2. Straight current apparatus 288 W. C. ALLEE The water is introduced through a circular tube 3 em. in diameter, which has holes to permit the exit of the water only on the side away from the trough. The trough is 3 cm. across, 2 em. deep, and 50 cm. long. At its lower end it opens into another well exactly like the one at the upper end. This in turn opens into a drainway directly below the end of the main trough, the drain being 1 cm. above the general floor of the apparatus. The whole trough is made of wood and is painted a dead black with water- proof paint. Thanks to the careful workmanship of Mr. Floyd of the Ryerson Physical Laboratory, the trough is very accurate in its dimensions. The animals were confined by wire gauze and their movements measured by means of a centimeter scale at the top of one side of the trough. When in use the apparatus was kept almost level. The animals to be experimented upon, were placed in the cur- rent at the center of the trough. They usually started off in the direction in which they were first headed regardless of the response which they would ultimately make. In some cases posi- tive animals would continue with the current to the lower end and then turn and make their way back along the edge of the trough, toward the upper end. Usually they did not collect in contact with the upper screen as Banta found to be the ease in his experi- ments. Rather the response brought a majority to the upper part of the trough where they settled in the angles between the bottom and the sides, with some of course clinging to the screen. In taking readings the exact position of each individual was recorded and an average taken of the position of the whole group. From five to eight was found to be the most convenient number to be tested at one time. A summary of the results obtained in this manner, with normal stream Aselli, is found in table 5. Often the isopods would move first up stream and then down stream, without giving a definite reversal. In these cases the reversal time was taken to be the time after which there was no decided movement against the current. When the experiments ran over night no readings were taken after midnight. Except in the third and last experiments recorded in the table there was RHEOTAXIS IN ISOPODA 289 TABLE 5 Stream Aselli in straight current TEMPERATURE eee | ce. per liter | | per cent + | cc. 9.24 | after 2:01 84 3:06 1680 6 7 9.21 Q:47 | 3:59 | 1680 17 7 9.52 | after 5:00 | 54* 15:52 | 1000 5 8 5:30 | TO 29:23 1020 5 6 970300. 740 | 7h 29:37 | 1050 8 10 5.13 | 23:48 | 64 59:48 | 1050 7 10 6.33 | TELE: | 58 oc 5i aa 1020 12 15 Balge| Ooe 44:47 1030. 14 14 4.33 | Bsoo. 60 | NSA | 1100 16 16 6:40 | iva 24:07 | 1050 16 16 7.00 6:13 78 1D:355 4) 1250 18 18 6.85 | 8:20 | 78 (istoy 1100 17 16 6.77 | after 7:00 | 86 26:20 |- 1050 | 13 15 * Pan response taken after trough reversal; all others were taken before. The first column gives the oxygen content of the water in cc. per liter. The second gives the time in hours and minutes before there was a decided reversal in the response to the current. The third column shows the percentage of positive responses of the same animals to the circular pan current. The fourth column shows the duration of the experiment in hours and minutes. The strength of the current is given in cubic centimeters of flow per minute, the trough being placed as nearly level as possible. In the three cases where no exact amount of oxygen is Shown, the amount present was well above air saturation at the given tempera- ture. no appreciable change of position during this period when no readings were taken. A study of table 5 will show that there is a relatively long lapse of time before reversal. This length of time is correlated with the high per cent of positive responses. This correlation is by no means Close, nor does it follow the fluctuations of the pan response but the significance of the relation between the two is well shown by comparing the results obtained here with those of isopods kept in water having a low oxygen content. 290 W. C. ALLEE 2. POND ASELLI a. Adults As has already been mentioned, there are two species of isopods found in the ponds near Chicago. Mancasellus danielsii (Rich- ardson, ’05, pp. 417-419) has been previously reported from La- porte, Indiana, only. It isa much flattened form and is generally found in the grasses in the shallower water of the ponds. Its reactions so far as tested agree with those of A. communis and unless otherwise designated all the discussion of pond behavior will be based on the latter species. The pond Aselli are decidedly smaller than those from the stream. In the isopods that have been measured the difference averaged about 3 mm., that is they were about 75 to 80 per cent of the length of the stream forms. However as has already been stated the pond isopods contain all the variations of the pro- podus of the first pair of legs that are to be found in the stream forms. The pond isopods react to light, heat, touch, and gravity in much the same way as the stream animals, although the speed of the reaction and the sensitiveness to the stimuli are probably different. The rheotactic response of the isopods from the two habitats is markedly different. In place of the positive reaction to the current dominating as in the stream mores, these isopods give a high proportion of indefinite responses. Their orientation is less definite and they do not appear to be so capable of ho ding an orientation onee it is attained. Their response is less vigorous than that of the stream isopods and they are much more easily swept off their feet by the current. A typical pond isopod response to current is shown in figure 3. This trial and record was made exactly like that for the stream isopod shown in figure 1. The result of twenty trials is given in which the animal went positive 25 per cent, negative 30 per cent, and indifferent to the current 45 per cent of the total number of trials. The same response is shown in table 6. It will be noted that the table includes the response made during the breeding season and that this response is not markedly different from that of the RHEOTAXIS IN ISOPODA 291 rest of the year although it is somewhat less positive. That is, the depression found in the stream Aselli during the breeding season is not so marked in the pond isopods. The external condition with which the low positive response seems to be correlated is the low oxygen content of the water. Fig. 3 Twenty successive rheotactic reactions of one pond Asellus Thus the highest oxygen content of the pond is a little below the lowest found in the stream during a period of normal response, while the normal amount of oxygen found in the pond water is much below that of the stream. The amount of free carbon diox- ide present in the pond water is higher than in the stream, but 292 W. C. ALLEE TABLE 6 Rheotactic response of adult pond Aselli | NUMBER CO. | TEMPER- | Ox | RESPONSE | OF TRIALS | 2 | Agen REMARKS ce. per liter per cent + | per cent —| per cent = | lec. per liter pee en 16 a 216)

am) “DoG-oG Anjouadwuay, °g 10g f= 4 &@ OG ef) VG 89 0 69g LG eel 09 Gl cl. | & | ¢ 68 91 Sia Oy AO eee ee Gacy I) a | aaa O | (aeuanG cI OL cil Mis | tek | 46 (eta, (8) ; | | I L 08 g 1 th See Gr | 692, L le OS | SP al 8s | é T€°6 0 eater qua0.10d aaa daqyy dad ‘99 | waqy tad *99 | shop \eueotad yua0 wad yu99.0| 109 dad *99 | 4a74 sad *99 O “awa a ASNOASat | #00 70 eS ie) een een ASNOASa | 700 70 IOULNOL) INGWINdd xX I 140d . 312 ypasy woos fo wospas Burpaasq ay) bBursnp srapjoays uodn aprxorp uogsva ybry fo joaffa ay], 91 HTAV.L RHEOTAXIS IN ISOPODA 313 asphyxia, partially due to a lack of oxygen and partially toa direct depressing action upon the central nervous system. When mixed with sufficient oxygen the specific effects of the gas may be observed without asphyxia. “Under these conditions transient stimulation occurs, followed by subsequent depression of the central nervous system and heart. In well diluted vapor, only the exaltation occurs as the anesthesia does not follow. This would - mean that the results shown in the first parts of tables 15 and 16 are due to the action of carbon dioxide as a stimulant, and this may even retard the depressing effect of a decreased oxygen sup- ply (table 11, part 4). But when over 200 ce. of carbon dioxide per liter are present, it acts as a strong depressant. Regarding the effect of large quantities, Cushny (le., p. 588) says that in mammals a large amount of carbon dioxide probably acts as a _ poison to protoplasm, for it lessens the amount of oxygen absorbed, so that in the final analysis it would seem that the depressing effect of carbon dioxide is directly due to increased oxidations and thus it acts in the same way as when the supply of oxygen is decreased. It will be remembered that the free carbon dioxide present in the streams is about 2 ec. per liter, while that of the ponds may run as high as 40 cc. But since in the experiments 45 ce. of car- bon dioxide acts as a stimulant, and since small amounts of the gas are known to be mammalian stimulants, it is improbable that the low positive response of pond isopods is correlated with the increased carbon dioxide content of the water. This then leaves the decreased oxygen supply as the main environmental factor with which the lowered positiveness seems correlated. c. Chloretone Chloretone (acetone chloroform) belongs to the class of sub- stances commonly known as anaesthetics. Authorities generally agree that these anaesthetics inhibit to some extent certain of the fundamental metabolic reactions (Child, ’10, p. 173). Different strengths of chloretone were used in these experiments but 0.005 per cent was found to work best for experiments that were to run some time. Preliminary tests showed that the animals collect THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 13, NO. 2 314 W. C. ALLEE in bunches when first placed in a solution containing chloretone. Later they became acclimated and moved more normally but would bunch again if more chloretone was added. As has been previously stated, this reaction is also induced by a sudden in- crease or decrease in temperature. The exposure to chloretone for the rheotactic experiments was made in sealed glass jars, the solution being changed every twenty- four hours. The results of one experiment only are shown, (table 17) but these are entirely comparable to the reactions given by a number of other tests. In this experiment the con- trol animals remained strongly positive while the reaction of those under the influence of chloretone was cut from 78 to 25 per cent positive. Towards the end of the experiment the animals were evidently becoming acclimated and hence gave a normal response. d. Potassium cyanide Potassium cyanide is known to decrease the amount of oxida- tions by decreasing the ability of the tissues to take up oxygen. (Geppert, 799, p. 208). Then weak solutions of this chemical should show the same results upon isopod reactions as keeping them in a low oxygen supply. Experiments prove this assump- tion to be true. The results of two series of such experiments are shown in table 18. These experiments were conducted in every respect like those with chloretone. The amount of potas- sium cyanide used gave only a faint odor to the water. In the first part of the table, isopods giving a positive pan response of 82 per cent, were placed in N/100,000 KCN solution. At the end of five days they gave a 30 per cent positive response. For the next three days they were kept in N /125,000 solution and during that time the response was practically the same. At the end of eight days half’ of the original number were dead. The remainder were put in tap water and showed a rapid recovery of their normal response. The other trial shows the same results except that when the experiment ended, all but two of the isopods were dead. The lowering of the positive response in the control is due to the fact that the control in these experiments was run with very little food. 315 ISOPODA © RHEOTAXIS IN 0g p 8e Ce eco edna Cr og 8h 9g OT 62°¢ aie OF ge z 69 V8.2 se Ole | 709 z9 0 88 COMO) S |= = Sw aneey 09 9 9 SSiee me GsOr eallanG, 0g 9% OL 89 08'¢ ee og 0 8 CL 602 0 | Om eOce 0 0% 08 €0°9 19}@M de, : PI 0g 0 lz SL PG 2 19}0M dv, | OL OF L OF 89 18°L oye Avy, og | 9 FI 08 9%°9 8 OF 08 Ler alkenes 18°9 are sy ea 81 04 16°S g 0g 08 OF 08 Ges pias OSiae=\ a LT és. 1a? 0 | 2» quan dad | — quad sad | + quao sad | 1aqy wad ‘90 shop x» uaa wad | — uaa tad | + quad sad | sayy wad ‘90 Seman | GASNOdSau . (0) Sees Meant | BENOgE EE ‘O Nigpal TOULNO’D ry | LINGAWNIN ad Xo ainjosadua) WOOY 66-41 aunqouad a,j ‘Upasy puod ijnpp ur srvxpjzoays wodn aprupfia unisspjod fo jaaffa ay 8I WIAVL 09 € 8 68 6F' 9 16 Og rae 9 oY 00°2 | LG 0g FP 9 0g 00° 2 09 g 0% Gls aa Glee «= 0¢ OL GG Ig Le 68°F 08 | 9 £Z Od 66S aI OF le LF 8Z 9 &%'S | tI OF &F G es &% og | 9 91 8d 69° 0 | » 7ua0 Lad — quad wad | + quad sad | a1 Jad *90 shop » 7uad Jad | — qU99 wad | + #uao dad aqu] sad *99 mara nenae | A SISCONEASIC LS | ize) Bete eh a ASNOdSau | 70 TIOULNO’) INGANINAd Xo “DoSI-od] aLnposadwua J, “pasn 9u0}0.10]49 fo woynjos quaa dad 0086/1 “UJasY Uwnadjs YNpH U2 siapjoay UOdN auojaLojyo fo Jaa{Ja au. 4T ATEaV iL 316 ’ W. C. ALLEE e. Low temperature The general effect of lowering the temperature is to cause a decrease in the positiveness of the animals. This decrease is well shown in the experiments listed in table 19. No attempt was made to find how small a decrease or how short an exposure would cause a reaction, the only care being to find if low tempera- ture would affect the response. The results show that a decrease in temperature does affect the rheotactic response in a marked manner, mainly in that it renders the isopods extremely inactive. Although these ex- periments do not show any acclimatization, yet by comparing with results listed in table 2 it will be evident that the isopods do come to give their normal positive response at as low as 4°C. above zero. Again this is a reaction to a change of conditions. General experimental work has shown that a change in 3° of temperature does not usually affect the reaction to current, although in one case a change of 4° did have a marked effect. In TABLE 19 The effect of low temperature upon rheotaxis in adult stream Aselli I ee, a Tae wonnen or | Mas parame cc. per liter per cenl + | per cent — | per cent « 18 5.69 | 78 16 | 6 50 | 7 days 5 8.41 16 84 0 50 | field test 18 4.99 94 4 2 50 | 2 days 4 6.08 42 16h | 42"! 50 12 hours 10 | Thaw 8&2 i || 1 80 | 10 days Beg Sa pe IPs) vis bie | Il 15 7.40 90' | 8 2 80 over 10 days 0 a AE | 95 80 34 hours 16 6.45 Os 14 16 80 43 hours at 12 6.55 64 23 ie 13 80 | over 10 days 0 6 | 89 70 2 hours 12 6.50 66 20 14 70 5 hours RHEOTAXIS IN ISOPODA Sle a temperature gradient, isopods collect at a temperature near that to which they have been previously exposed so that this optimum shifts with external conditions. The effect of a decrease in temperature upon the bunching reaction has already been mentioned. Since the isopods become acclimated in both these cases, it was to be expected that they would show a similar reaction regarding the effect of tempera- ture upon their rheotactic response. f. Starvation The effect of starvation is shown by the results tabulated in table 20. These experiments were carried on in filtered lake water. The only difference between the experiment and the control was that the latter contained a few leaves for food. The results show that as starvation progressed the positive responses of the isopods were diminished. The decrease shown in the con- trol is due to the approach of the breeding season. As has been explained in other cases this may have helped cause the decrease in the experiment, but by comparison it will be apparent that the breeding season was not the major cause of the change in response. 2. POND ASELLI a. With increased oxygen (1) Normal adults. It is obvious that if the amount of oxy- gen present in the habitat is the determining factor in the rheo- tactic response of isopods, increasing the amount of oxygen pres- ent should increase the number of positive reactions. The amount of oxygen present was experimentally increased in three different ways: (1) The animals were placed in running tap water. This method is objectionable because it introduces a current, and the mechanical effect of this might be the stimulating agent; (2) the aerating device already described was used; and (3) the isopods were placed in water containing a large amount of green water moss. Similar results were obtained from all three methods and these results are listed in table 21. ALLEE Go 318 09 | tS! GG | 99 66 °¢ OF AT | O§& &9 OL 8h°9 OL | F SP | &¢ 8b PF O8 OL | 8& | of if 98° Cs L cee, ||, 209 Gg"9 08 Ce ae eh Se ey 0 Gag Loh OS | Ol ial | 94 GT 9 09 LL 6 06 6 oT 9 0S | 0 IG 64 WI 8 08 SY al It & | eg) O8 | 0 OL | 06 OF 9 OS G 8 06 0 | OFZ I : og Sais peeCiheree une” zo'g og p se 8g Tee all ed 0g 0Z 8 el 0 Z0°L | » Wan sad — quaa sad | + yuao wad 137y sad *99 » quao wad | — quad wad + 7ua9 wad shop | 4374) “ad “99 Romain mgN Odea | 0) aaa EO deat Amare eras | ‘O IOULNOD INAWIN TAX "Dob l-oGT 9Lnouad wa J, I “UISW WYILS Np Ur sixpjooys Uodn wornasn)s fo Joaffa oY, 06 WTAVL RHEOTAXIS IN ISOPODA 319 TABLE 21 The effect of high oxygen upon rheotaxis in adult pond Aselli Or RESPONSE ae | Greats ae TIME Gane ec. per liter per cent + | per cent — | per cent « days iy A 34 24 42 90 | 17 10 5.12 | 49 | 31 29 30 17 11 &.12 SiG | 16 52 50 14.5 19 5.12 | 46 10 44 50 14.5 19 JEG eo hanes samen nme 30 30 13 44 9°31 | 26 28 52 50 ait 54 8.96 | 56 4 40) 50 1.5 75 8.96 44 6 50 40 7.5 | 73 9.22 | 52 2 46 50 8.5 89 5.86 CO ee 50m b 0 10 ao 188 7.42 56 QO | 44 50 | 21.5 15 4.84 | 70 On 30 20 | 17 65 * The amount of oxygen present was at least 5.12 cc. per liter. With the exception of the last two items, this table deals with one stock kept in a large aquarium in running water. The tests were made by selecting individuals at random and so give a fair representation of the reaction of the group. The increase in positive responses is not great but it is marked enough and con- stant enough to show that increasing the amount of oxygen pres- ent will affect the rheotactic reaction. One very significant fact is that there is no evidence of acclimatization in this or any sub- sequent test. The last two items show the same response with isopods kept in still water which gained its increased oxygen supply from the photosynthesis of Amblistigium moss. If increasing the oxygen present in the water 5 cc. per liter caused an increase in the positive response, then increasing the concentration still more should cause a yet greater increase of the positive reaction. In order to test this, oxygen was bubbled through the aerating device previously described. The oxygen gave the following analysis; oxygen, 99 per cent, carbon dioxide trace, nitrogen 0.95 per cent. The results are listed in table 22. The increase in the oxygen caused a decided increase in the positive reactions as well as in the general activity. Where isopods in the control crawled slowly these ran rapidly, often | 4| | ; | | ; | en Oe 2 ie Ope Pace 69d 09 = i90r = Late HST ee Cy eal at 09 le Sy | Ib st 66° 96 - | | 0 | 0g | Ved ah 288s 88 691 TTL Og ST |} She 9S 687 Le OS PL 8 8k | 96°96 0g rg 9T 0g 61g 0 0g 98 vial 0g =| «Brg a I 4 a 9G 0G 0G } Sal G9 97 8 S , FI 0g GBF | SOT 8 er 16 - beet 0g 98 ia 0g er re | 0 OS) = Sie. 07, 8% 6s 981 2» 7Ua2 wad Nata quao wad | + yuao vad 13qy tad *99 shop » 7Ua9 lad | — quaa wad | + 2uUa0 vad 4ajy tad *99 Bae aASNOdsau tO aqqasodxad aAWIL ene GASNoOdsau : 70 TIOHULNOYD oF e ‘ A: OINGAWINAdX I ‘Dol@-od] ainpouaduay, “ijasy puod zjnpp ur srapjoays uodn ‘yuajwoo uabhxo paynunqwos v burany say fo oaffa au. L 6 ATAVL 320 RHEOTAXIS IN ISOPODA é 321 starting suddenly when no apparent stimulus was acting. Even when the rate of positive response was already high, as shown in the second part of the table, increasing the amount of oxygen present caused a large increase in the positive reaction. ‘The third part of the table 22 shows the effect of allowing the oxygen in the water to escape gradually, and under these conditions the animals returned to a normal pond response while the amount of oxygen present was still much higher than that in their normal habitat. However it was very low in comparison with the amount of oxygen to which they had been previously exposed. (2) Juvenile mores. Table 23 gives the effect of keeping juven- ile pond Aselli in high oxygen for long periods of time. Other tests run for shorter times, give similar results but with lessincrease - in positiveness. The right hand side of the table gives averages from the response during normal development and is placed here for comparison. ‘Two things are brought out by the table. First, the pond isopods kept in a large amount of oxygen, develop not as normal pond mores but as stream Aselli, and second, that the time taken to acquire a given size is less when a larger amount of oxy- gen is present. The final results are especially interesting. At the age of 123 days the pond Aselli in high oxygen were 6 mm. long and gave a 72 per cent response. ‘It took those in low oxygen 193 days to attain the same size and then they gave only a 38 per cent positive response. There is no evidence of a return to the response normal for pond isopods. (3) Response in straight current. Again the tests with the continuous straight current support the results with the dis- continuous circular one. The results of these trials are sum- marized in table 24. Isopods from all three methods of furnish- ing increased oxygen were used with similar results, and one is forced to the conclusion that increasing the amount of oxygen present in the water makes juvenile pond Aselli give a positive rheotactic response comparable with that given normally by stream isopods. ALLEE C. 322 ‘ssouoATyIsod jo yuao sad ayy Jo stu} ut osuodser uvd ay} PUL So}NUTUL PUB SINOY UI WATS ST OUIT} OY} ‘OL0Joq SV , SABP GST 109BM G UNI it sAVp PZ 1oyVM SUNY | OT iz {| Gist | skep gg soyem Suny | g | ¢ | gee | 39s | OTS, Baie | oe | skep gy 1oyea Buny | OT eszz | 99 | sAep GJ 1oyvM SUNY | OT 60: § O& sABp G) IoJVM SUNY | OT I SLT Gg | sAvp @) 1oyVM SUNY | OT Pie | Ose 8é sABp Op 1oyVM Suny | eT OL: eI SY sABp GQ SSOUL ST OL 8c: LT OL SABp (OT IO} ddIAVP Suyes1sy Or v 8G: 9G gg unt | | | | sno -enby ul L ONG aSNOdsau SUUVNAY so= t= | -Ndaxa | NvVd =) | 10 HLDN@IT| AALISOd | GUN LVaadWN ak BS:9 98°¢ (lazfD) O18 gc OL 6y:9 Zo '6 98:7 ZS OL 60:9 Ze 6 (4ajf0) 60:8 96°8 (4a7fP) 8T:T 96°8 (aif) OL:@ 286 06:6 G18 66-9 —8°F 86-8 I¢'8 ayy wad *99 recent ze) aymuru dad *90 0011-000 7 waning ] wana jyBro.ys u 1 uablico ybry wo.f wpasy pu Od Go WIAVL eon a] 288 9 14'S S21 S21 lee Gr: x\lieeee o ylinecS lees eso er 8 | Gc 99 lee rele |. oe Mi lk roth 679 GLE 86 VA 648 GZ 99 GO es en O |. 69 [| Taal, © AGUNG Niel sat eee yt a, 09 09 ZG ST Zparcn| SO/ ve meg Mee ale cid shop | + qua? wad | “UL | 4aqu) tad *99 shop shop | | x» qua wad | — quao sad | + quao sad | WU | 4aqu) ad *99 cov | govusay | ame | soaovuay | (QROM | | gov | emvusso | sxoasa eee eit | = = | | : ~ Sa ees ame NADAXO MOT NI MOOLS WOU ASNOdSAU ADVUAAV INAWIYEdI XY UO’UDIMDA WOOL) *9Y.Ee— aun pada ‘wasp puod apruaanl fo asuodsa. ‘ ‘ 00G 6 5 & ATAVL l ayonjoay.t ay} uod n uabhixo Buy fo joaffa ay, RHEOTAXIS IN ISOPODA 323 b. Caffein Physiologically, caffein acts as a permanent stimulant. That is, there is no depressing after effect (Cushny, 710, p. 248). For this reason, experiments were run with pond isopods to find if such a stimulant would cause the same increase in the rheotactic reaction as that already produced by increasing the oxygen supply. For the experiments a solution of caffein in distilled water was made, saturated at room temperature, and this was added in small amounts to the ordinary tap water used in the experiments. The amount of caffein is shown in the table (25) in terms of cubic centimeters per liter of water. The results of a number of trials are summarized in table 25. The first part compares the effects of different strengths of caffein solution upon the rheotactic response. For the time used, 25 cc. of the saturated caffein solution per liter of water proved to be most stimulating although the mortality was high. Part 2 gives a series of trials with a group of pond Aselli exposed to 10 ce. of caffein solution per liter of water. The positive response was increased over 30 per cent and continued high for over eighty hours. The subsequent decrease in the positive response is due to the fact that the isopods apparently became acclimated to the caffein. In the third part is given the record of another set of trials with 10 ce. of caffein solution per liter of water. Under these condi- tions, the isopods showed a slightly higher positive response and gave the same slump in positiveness at about the same time. These experiments were run at the same time as those given in part 2 and the same control was used. Part 4 shows the effect of increasing the caffein after the animals had gone back to their normal reactions, through becoming acclimated to the strength of caffe used. The increase given must be due to the increased amount of caffein, because no other factors were acting. After the acclimatization to this amount of ¢affein occurred, a further increase served to kill the animals rather than to increase their positive reaction. Again as in the case of the saturated oxygen, the isopods are decidedly more active in all their movements so og oP | 2 66 Nero Se oor |) (og oe qyobeee. | Ave or cy | | 6 og - oS GZ or or £87 | | | §8 0g 8I 8% 4g ol Chr 0g o¢ | w% | 9% OF rt | 8F og ZS 0g 87 ol rE 0g a a) a? 08 06° LT og 8% 2G 0g ol 68°F | 0 09 LE 9% ug or 68°F ° II & ] ] | fe | 8gT 0g 91 Of ve 08 4 | | 46 i Glee N 8s IT 62°¢ | | ¥6 Ooo nice or | 86 28 62° Ss | | Oe On a Se | eur 9% 50'S : | | re) eas ie EA a eel 0 | Fore = | 1 quad wad | — quao vad | + quao wad 4aqy) dad *90 | sunoy | » 7Ua9 Jad | — Quad wad | “ quao wad Jaqy Lad *99 aj] Jad *99 pede | ASNOdsau 70 | GWIL | Ala detarey GSNodsau NIGdad vO cre) TIOULNOD | ; INAWINAdIX I ‘Dolé-odT anpouadmay “wyjasy puod ynpp ur sixvjoays uodn uraf{va fo poaffa ay], S$ HTAVL 324 325 ISOPODA IN RHEOTAXIS ‘prop [1B S10YIO y €% «OT OL O1 08 98 GZ F | De elie 08 91 99 81 96 z0'¢ | gI 00) > 20 GZ Soe | 96 68°F | L mes] Esiote tPF 96 $1 pue oy a | shop | AI | | Col 02 | oz Po 96 91 62°F | | 46 OG: Si\ece 02 7 $1 61 F QAOG®B JOI1JUOD 9G &8 OG aeci 0€ 84 $I 89°F i | SF OG-<5 "= Fi ze “9 9] SP | | Tit Og “3 \-2-0z 9g rt $I 68°F 0 OO) si aeeZe 9 ug 0 68°F II 326 W. C. ALLEE that their general activity as well as their rheotactic response is affected. c. Increase of temperature It was shown in a previous section that decreasing the tem- perature caused a lowering of the positive response; theoretically then an increase of temperature should show reversed results. The first experiment to test the effects of increased temperature was carried on by keeping the animals near a hot electric plate. The second, by keeping the animals in an automatically regulated heating tank. The results listed in the third part of table 26 were obtained by keeping the isopods in a melting ice pack ° exposed to a warm room temperature. The first part of the table shows that the effects produced by an increase of 6° operating for twelve hours, will persist for at least that much longer although the isopods are returned to the original temperature. The second set of trials indicates that the effect of continued high temperature is only temporary and the last three tests show that an increase of only 4° may cause a most decided increase of positiveness. The last experiments were upon stream Aselli during the breeding season. Other experiments show that a sudden increase of 10° in tem- perature may cause either an increase or a decrease in the positive reaction, but if an increased response is still given after a 10° raise in temperature, more heat always results in a diminution of the positive response. Whether a 10° increase in temperature will causean increase or decrease in the positivity, seems correlated with the general amount of activity of the isopods. 3. EFFECT OF OXYGEN ON SIZE The relation between rate of growth and the ultimate size of the isopods, with the amount of oxygen present in the water has been mentioned already, yet this subject is of enough importance to merit a short general treatment. Colton (08, pp. 410-447) demonstrated for snails, that the size may be correlated with the general oxygen supply, but the exact amount of oxygen present RHEOTAXIS IN ISOPODA BAT TABLE 26 The effect of raising the temperature upon rheotactic response I O2 TEMPERATURE | RESPONSE | eee | ERT, oe per liter | aaa + ae cent — | per De a | ot ee ee 4.89 19 26 2A 350 50 25 Go 2a 0 a2 | 50 3.75 19 heeeoG puny 24) lemon 23 | 50 Bal 19 Soi) 20 Wy ahd 93 60 II 6.15 (Sei ee4on® |) 736 22 | | 50 | ar | wee | Ge, | 632 22 50 3.93 | 26 68 | 24 8 Af 50 26 40 42 18 70 50 4.55 ove 42 182% jh) GRA0 121 50 III —-- -—-- = a. : | s LOG. » | 4 | 28 Stil © #25 0 | 80 10°47 = | 8 46 40 14 2 | so 8.73 Ae el 23 56 eee |. Wee) was not determined. It has already been stated that the stream isopods are on the average about 3 mm. larger than those found in ponds. One set of twenty pond males averaged 12.75 mm. in length; a set of twenty-eight pond females averaged 9.14 mm. The average of twenty-five stream isopodsof each sex gives the males a length of 15.3 mm. and the females 12.2 mm. During the course of the experiments, the difference in size of isopods reared in low and high oxygen was quite noticeable and those with the higher amount of oxygen, other things being equal, were of a larger size. Although only a few of these were measured, the measurements were entirely characteristic of the general effect and are shown in table 27, together with the average oxygen content and approximate age of the isopods measured. Both sets of isopods were entirely comparable, being taken from the same stocks of animals. Both had plenty of food although perhaps the pond animals were better supplied. The 328 W. C. ALLEE stream isopods were all kept in still water, the higher amount of oxygen being supplied by means of the aerating device. In the pond isopods, however, those from the higher oxygen were kept in running water so that they were in a medium free from their waste products, which other workers have found to cause a diminu- tion of size (Colton, |. c.).. The results indicate that the amount of oxygen present is one of the factors and probably a major one in causing the size difference between pond and stream Aselli. TABLE 27 Effect of oxygen on size Stream isopods Low O | HIGH Oz oe : : : —_ Soon oh ae Ma is Aen ets WM hs ROE sera lGOidays.) | Aged. 2 ee et 5 169 days Average Os, cc. per liter.......... 3.53 | AverageOs» cc. per liter........... 6.02 Average size, 20 isopods.....3.95 mm. Average size, 9 isopods. ...-.5.89 mm. Pond isopods Low O2 | HIGH Oz | ACA: OTR TALS Bese a TGnclawanal! Age 15)... caet ee eee 123 days Average Oz cc. per liter.......... 2748 Average >, ce! periliterss+.. sneer 7.09 Average size, 5 isopods......... 4mm. | Average size, 5 isopods........ 6 mm. 4, REACTIONS TO GRADIENTS OF GASES Earlier experiments had shown that isopods would collect in their optimum light or temperature conditions if subjected to a series of graduated changes in either of these conditions. Since their rheotactic activity is more dependent on the oxygen supply than upon either light or temperature, experiments were tried to determine whether or not the Aselli would respond to gradients of oxygen and carbon dioxide. These experiments were run in galvanized iron boxes especially designed for the purpose. Each box was 50x 30x 7.5 cm. in dimensions. The bottom was covered with a layer of beeswax to give the isopods a better foothold for crawling. All the metal RHEOTAXIS IN ISOPODA 329 parts were painted dead black. ‘Two centimeters from each end a screen of brass wire netting was placed in order to keep the isopods from the current introducers. These introducers were of brass tubing in the form of a capital T. An even distribution of the current across the pan was assured by having the cross bar of the 7 punctured by eight equidistant holes, each 3 mm. in diameter. The water was withdrawn by a brass tube 2 cm. in diameter placed at the middle of the box with its lower side 4 cm. from the waxed bottom. This was also drilled with equidistant holes, which were guarded by wire screening tc prevent the isopods from escaping. Tap water was introduced at one end and the boiled or the high carbon dioxide water, as the case might be, was allowed to flow in at the other. The most striking results were obtained with a flow of 200 to 400 cc. per minute at each end of the pan. This current spread over the 10 cm. width of the pan, gave a flow of 5 to 10 ce. to each square centimeter of cross section, for each minute, and isopods do not react definitely to a flow of this strength. With this device, a gradient of from 1.82 ce. of oxygen at one end of the pan to 8.14 ec. at the other, was obtained. For changes in salts see table 9 and discussion on p. 296. The half bound carbon dioxide was decreased 1 cc. per liter and the nitro- gen content was lowered from 18 to 3 cc. per liter. In all cases a control was run in a box, the exact duplicate in every way of the experimental one. The only difference between the two sets of conditions was that in the control the same kind of water was introduced at both ends, the rate of flow being the same as in the experimental box. From five to ten isopods were used at a time and the experiments ran from one to six hours. Read- ings were taken at five, ten or fifteen minute intervals. In all cases the trials were made in very diffuse light or in total darkness. The readings were taken with a one candle power white light. The temperature was kept near enough that of the water from which the isopods were taken so that there was no temperature interference. THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 13, NO. 2 330 W. C. ALLEE Table 28 gives the summary from eight trials made with isopods from water having an oxygen content of over 5 ce. per liter. The grand totals show a decided positive response to the tap water end, although the control animals are almost equally distributed between the two ends. This mass result was confirmed by transferring the introducers and thus gradually changing the amount of oxygen present at either end, and also by stirring the isopods back to the center. In all cases the final reaction was in favor of the tap water end. Generally this reaction was the result of a series of movements about the box during which the isopods went from one end to the other but in a few cases there wasa definite turning back when the boiled water was encountered. Table 29 shows the response of isopods kept in an oxygen supply of less than 3 cc. per liter. Even under these conditions they did not collect in the boiled water end of the gradient in greater num- bers than in the tap water end, yet the response to the tap water was cut from 77 to 53.3 per cent of the total number of readings taken. This probably means that having been kept in low oxy- TABLE 28 Stream isopods from high oxygen in an oxygen gradient EXPERIMENT ConTROL LOW ConTER HIGH CURBENT: | LOW |CENTER| HIGH iG; cc. per min. O» gradient cc. per liter......... | 2.56, 5.58) 7.68) 1200 | ESP OUSCry tye eavvs ts areata eye cack 15 5 40 20 4 17 PEAGIENGA.2 ese ries ea ee Oe Be 3.07, 5.47) 7.63 600 | EESDONSE: aes ists. Sse 8 ae eal VG Bee 122 eos 9 68 RESPONSE ME eA sy lo ect pee 20 21 67 / no control PEACTEMb cra Apres oN cideacin es eee 1.82 8.14 200 FESPONSE S267 27 sk crc ae ae Nae. 10 =‘{178 | 102) 39)]+ 75 87 22 ~=—(388 | 160 23 | 169 32 18 64 Ly 62 35) |) 36 | 9 4 73 37 | 168} 45 Totals.Ss i Ort biel Fee 197 | 82 (932 | 434] 278] 410 Percentage of total trials....... 16: he 77 ae) 25 36 RHEOTAXIS IN TABLE 29 ISOPODA Stream isopods from low oxygen in an oxygen gradient 331 EXPERIMENT CONTROL LOW |CENTER| HIGH CURRENT: LOW |CENTER| HIGH SOE en ce. ae Bas cc. per liter O» gradient ce. per liter) 3.07) 5.47) 7.63 PESPONSE Ses) ee ees 20 18 40 600 5 3 26 2.44 PaeNOlAM Ree bons ooneneeaall bose 8.14 INSS| OVINE 6 ok gb 56 nove wall! 240 9 {151 400 41 6 | 139 2.12 95 43 286 525 189 38 | 197 2.22 80 26 71 525 65 43 82 2.56 22 9 26 300 25 19 22 3.00 42 38 96 300 39 86 51 2.00 (ANGIE cons eaoen een Bien 11.00 TESPONSC RAE ae eco 12 61 300 67 11 74 0. 3 17 30 28 300 Ue 11 14 2.44 Motels’ ieee ee Sar 1L8o! © 1759 508 | 217 | 605 Percentage of total EPA Se sr eies tras ac (h20 14 57 38 17 45 TABLE 30 ‘ Stream isopods in a carbon dioxide gradient From water having less than 2 cc. per liter current 300 cc. per min. ——__ EXPERIMENT CONTROL HIGH CENTER Low HIGH CENTER LOW CO: gradient ce. per liter 24 3.5 LESPONSC! ete ee ee 25 7 45 30 3 44 gradientys £24 hen}): 45 3 MESPOMSCe aa erie ioe 28 2 40 itt 29 34 PRAGION Gh csi ees cee eae: 59 3 RESPONSEke ayer ne aces 20 3 37 36 8 16 PALOMA ieee Meee ea sits 60 7 RESPONSE? . 2 arses ese: 7 0 36 23 t 15 PARVUM Miocsocegacsoonoo 80 20 LESPONSCe a4 age craes 19 0 59 52 9 17 gradientee herein OO 138: TESPONSCheer te eee ee 5 0 72 38 6 40 Motalsie.cs-s eee lie Os 12 289 196 59 166 Per cent of total num- her otprials: sco 26 3 71 47 14 39 Bou W. C. ALLEE gen conditions, the isopods became more tolerant of low oxygen, although their optimum amount of oxygen was still higher than that furnished by the low end of the gradient. The results with a carbon dioxide gradient are given in table 30. These trials show that stream isopods avoided an amount of carbon dioxide equivalent to the highest quantity found in nature, but that the avoiding is much more pronounced as the amount of carbon dioxide is increased. The result is, that al- though the higher amount of carbon dioxide diffuses against the current to the low end of the gradient, the isopods collect in an amount of carbon dioxide to which they are strongly negative when such a response is possible. These gradient experiments show that the stream isopods tend to collect in the amount of oxygen or carbon dioxide to which they are accustomed. In nature this reaction would tend to keep the stream isopods from collecting in conditions that might affect their general state of metabolism, and in this way affect their power of resistance to the stream current. 5. SUMMARY OF EXPERIMENTAL RESULTS The major experiments may best be classified on the basis of the result, of the materials used, upon the rheotactic response. Under this division they fall naturally into two groups; those that decrease and those that increase the positiveness of the rheo- tactic response. These may be summarized as follows: A. Conditions that decrease the positive rheotactic response: . Low oxygen . Chloretone . Potassium cyanide Low temperature . Sudden extreme increase of temperature Carbon dioxide. . Starvation CO OO Oe NID To these may be added the life history effects as shown in the breeding season and the juvenile reactions. RHEOTAXIS IN ISOPODA 333 B. Conditions that increase the positive rheotactic response: 1. Atmospheric saturation of oxygen 2. Complete saturation of oxygen 3. Caffein 4. Increase of temperature, if not too extreme The rheotactic responses given by the isopods under different conditions are summarized in the following series of curves. These graphs are based upon all the experimental data at hand, and plot only the positive responses of the isopods. They neces- sarily show only two factors, namely, time and the per cent of Aug. Sept. Oct. Nov Dec. Jan Feb. Mar Apr May Jun. Jul Fig. 5 Normal rheotactic response. A, Adult stream; B, Adult pond isopods positive responses. The former is given by the abscissae and the latter by the ordinates. The other accessory data may be found by referring to the different tables. Figure 5 shows the curve for the normal rheotactic response for adult stream (A) and pond (B) isopods. The general sim lar- ity of the two curves is noteworthy. The most striking differ- ence, aside from the lower positive response of the isopods, is the different degree in which the breeding season affects the two mores. 334 W. C. ALLEE The unbroken line (A) in figure 6 gives the effect of a decreased oxygen supply upon adult stream Aselli. The response consists of four periods. First there is the lessening of the positive re- sponse which is followed by partial acclimatization. This is followed by the breeding season after which the isopods are either more vigorous or else are more capable of becoming acclimated than before and so give a stronger positive response. The broken lines (B) show the effect produced by keeping the pond Aselli in tap water, the steeper curve being plotted for younger animals. In neither case is there any sign of acclimatization. Oo Months 1 2 3 4 5 6 7f Fig. 6 Adult isopods in reversed oxygen supply. A, Adult stream; B, Adult pond isopods. The broken line (B) in figure 7 gives the results produced by a low oxygen supply, upon the rheotactic response during the breed- ing season. In general there is first an increase of the positive response, followed by a decided reversal. The unbroken line (A) shows the effect of an increased carbon dioxide supply The effect of the decreased oxygen is probably due to its action first, upon the germinal glands, causing a cessation of their activity, which gives more energy for other activities. This is followed by the direct depressing effect of the low oxygen upon the tissues themselves. The results so far obtained with carbon dioxide may be explained by its general narcotic action. RHEOTAXIS IN ISOPODA Bae) The solid lines in figure 8 give the effect of low (S.) and high (S,) oxygen content upon the rheotactic reaction of juvenilestream Aselli. The rheotactic reactions are too indefinite during the first month to be considered. The second month shows the same state of development in each environment but from that time on the responses are entirely different. It is especially noteworthy that the isopods kept in’a low supply of oxygen throughout their lives, did not become strongly positive even after the breeding season, which came in their fifth month. That is, they did not show the same capacity for regulating, given by adults under the same conditions. ' The broken lines give the reactions of isopod 70 60 50 (0) Days 1234S N67" 8: 9) 10 1112) 43 4S Ber I9K20 Fig. 7 The effect of increased carbon dioxide and decreased oxygen upon the rheotactic reaction during the breeding season. A, carbon dioxide; B, oxygen. of pond parentage (P,) in high, (P2) in low oxygen supply, and it is evident that their rheotactic reactions do not depend on their ancestry but on their environment. Figure 9 compares the results of different depressing agents at the concentrations used. Apparently a rapid decrease in tem- perature (A) is most effective and most transitory. Carbon dioxide (C) from 200 to 300 ec. per liter seems to be second in both these qualities. The isopods become slowly acclimated to chloretone (B) at the strength used, but show no such tendency with the potassium cyanide (D) in the concentration used. Star- - vation (EF) also acts as a continued depressant. 336 W. C. ALLEE Months 1 2 3 4 5 6 rf Fig. 8 Rheotaxis during isopod development in different amounts of oxygen. Si, Stream isopods in high oxygen, So, in low oxygen; P:, pond isopods in high oxygen, P:, in low oxygen. oL Days) 2S 45 6re, (85 SAOM 12 aIsM Am SRI) Mize los2One te fhe so4 Fig. 9 The effect of depressing agents upon rheotaxis. A, low temperature; B, chloretone 1/200 per cent solution; C, carbon dioxide, 200-300 ce. per liter; D, KCN, n/100,000; #, Starvation. RHEOTAXIS IN ISOPODA BEY) Figure 10 gives the curve for the effects produced by stimulating agents. Two distinct results were obtained with water saturated with oxygen (B, D). In one case (B) the isopods were already giving an increased positive response due to the relatively high oxygen content of the water. In both cases, however, the in- creased positiveness was maintained until death resulted. The isopods, however, become acclimated to both a higher tempera- ture (A) and to a supply of caffein (C). When once acclimated to oO | Days ol S44 SLO ee mOMON LINIZNIS 149 15 16) 17S1eMSI20; 22252 3%24 Fig. 10 The effect of stimulating agents upon rheotaxis. A, Increased tem- perature, about ten degrees; B, D, saturated oxygen, 25 cc. per liter; C, caffein, 10-25 ec. saturated solution per liter. the latter, a further increase up to the point of toxicity caused increases in the positive response. The experiments have also shown that the size of the isopods and their rate of growth are correlated with the oxygen supply. Also that when subjected to a gradient of different concentra- tions of oxygen or carbon dioxide, Aselli will collect in the con- centration nearest that to which they are accustomed. 338 W. C. ALLEE IV. DISCUSSION In presenting these results the writer realizes fully the attend- ing imperfections and the need of more work along some lines. But in order to do further work with this material, in this field, in an intelligent manner, these general relationships had to be first blocked out. It is the author’s intention to follow this paper with a study of the effect of the same conditions upon another characteristic isopod reaction, and to further analyze the interest- ing relations shown during the breeding season. The principles upon which this paper is based have been suc- cessfully worked out (under the author’s direction) with amphi- pods by Mr. W. J. Saunders in an unpublished master’s thesis and with amphipods and planarians by graduate students doing course work in experimental animal behavior. The work of these students has shown that the rheotactic reactions of both amphi- pods and planarians are comparable with those of pond and stream isopods and may be controlled by the same external factors. In the summary of the experimental results just given, the basis for classification was made the positiveness of the isopods in their rheotactic responses, but if the heading were rather the effect of these conditions upon the metabolic state of the organism, the summary would not need to be changed. For all the conditions that have been found to cause a decrease in the positive reaction are known to depress the rate of animal metabolism (Child, l.c., p. 173). Low oxygen (Haldane and Smith, ’97, p. 242) potassium cyanide (Geppert, 799, p. 208) and high carbon dioxide (Cushny’ ’10, p. 587) do this by directly decreasing the oxidations. Chlo e- tone belongs to the general group of anaesthetics that are known to have a depressing effect upon certain of the fundamental meta- bolic reactions (Child, l.c., p. 173; Cushny, l.c., p. 195). The decrease in activity due to low temperature is a well known phe- nomenon in both animals and plants, as is also the depressing effect of an increase of temperature above the optimum for life relations, while starvation decreases metabolism by removing the material to be oxidized and so gives the same_results from the other side of the equation. On the other hand, the rate of metab- RHEOTAXIS IN ISOPODA 339 olism increases with the increase of the oxygen dissolved in water (Lingle ’02, p. 83; Martin, ’06, p. 303; Loeb, ’06, p. 95). Then, too, Piéron (08, pp. 1020 and 1061) measured the amount of oxygen present at different times in sea water and found that Actinians expand and retract their tentacles as the oxygen ten- sion of the water increases and decreases. These observations were confirmed by experiments in aquaria, but Bohn (’08, p. 1163) questions Piéron’s interpretation of his results and attributes these rhythms to much more complex factors. An increase in temperature increases metabolism providing, of course, the in- crease is not too rapid or too great. Caffein is known to have a permanent stimulating effect (Cushny, l.c., p. 248). From this evidence, the rheotactic reaction in isopods must depend upon the metabolic state of the animals. All the observed facts concerning their response in nature support this view, except that in young animals the rate of metabolic reaction is higher than in the older ones. Yet the juvenile isopods give either no rheotactic reaction or are very indifferent to the current. But with these young isopods, conditions that favor a high metabolic rate cause positive reactions to appear sooner and to be a great _ deal stronger than in'those animals kept in conditions that depress metabolism. Then, too, the clinging reaction is strong in these young isopods and even under the most favorable conditions, they do not move about rapidly. Since the positive rheotactic re- sponse seems correlated with the degree of motile activity of the isopods, this tendency to cling in one place would account for the apparent discrepancy in the lack of DORLINeHeyS in these juvenile Aselli. : In nature the complex of conditions found in bes furnishes isopods that give a low positive rheotactic response, while stream conditions produce the opposite result. The amount of oxygen in these two habitats varies greatly, as has already been pointed out, and this variation appears sufficient to account in a large measure for the difference in the reactions. Of course the fact that the pond water usually contains more organic waste products than are found in the streams, may help to cause the difference in the reactions. But this does not seem to be the important 340 W. C. ALLEE factor since pond isopods kept in well aerated water containing a large amount of waste products gave the same increase in the positive reaction, as when running tap water was used. Also the cause of difference in response in isopods from the two habi- tats cannot be due to the mechanical stimulation of the current for experiments with both stocks show that this cannot be true. Pond Aselli kept in still water but in high oxygen, increase in posi- tiveness, while stream Aselli under the same conditions maintain their high percentage of positive responses. The isopods which gave the reactions in table 1 had been in still water all their life and yet gave a positive response of over 89 per cent. From these considerations it appears that oxygen is either the most important environmental factor in determining the rheotactic response or it is the best single factor index of the effect of the complete environment upon this reaction. Through the tables one can see evidences of irregularity in the response of the isopods from the same conditions. This irregu- larity was due to two causes. First, the error of the method, which has been shown to be about 5 per cent. This error is due, in part, to the tendency of the isopods to keep on going in the way they may be headed. This would account for an isopod going positive nine times and negative the tenth. The second cause of the irregularity is the fact that an 80 per cent positive response on the part of five isopods may mean that four of them went positive every time while the fifth was entirely negative. At first sight this would appear to be a serious objection to the plan of selecting the animals to be tested at random from the stock under consideration. Conceivably one might pick five isopods that would all give a response opposite to that of the general culture. In practice however this has not occurred and the large number of trials is a sufficient safeguard against such a source of error. However in the later experiments every animal used, both in the control and in the experiment, was tested and more uni- form results were obtained in this way. The question at once arises as to why one member of the stock under the same external conditions as the others should give a different rheotactic response. One of the reasons is the state of RHEOTAXIS IN ISOPODA 341 the isopod, regarding the time distance from the moulting period. The exact bearing of the moulting cycle upon rheotaxis has not yet been worked out, but evidence is accumulating to show that in stream isopods, the moulting period and the time immediately following it are characterized by an indefinite or at any rate a weak rheotactic response. Regarding the permanence of the modifications produced, it has been repeatedly mentioned, that in pond isopods kept in high oxygen, there was no reversal to the normal pond response, although in two cases these experiments ran for over six months. On the other hand, the stream isopods seemed to possess a power of acclimatization, and a return to the normal stream positive- ness in a fairly low oxygen content. However, in the case of the isopods reared entirely in low oxygen, this increase in positiveness was much less marked. Apparently if the stock could have been carried through a few more generations the response would have been entirely that of pond isopods. Hence it would appear that the reactions, with which this paper is dealing, are distinctly dependent upon the environment for their continuance. On the other hand the taxonomic differences, as the number of spines on the propodus of the first thoracic appendage, show the same vari- ation in both habitats. That is, the taxonomic differences are inheritable characteristics of the species, and are not dependent upon external conditions, while the behavior characters here studied are almost independent of heredity. _ The size of the isopods has been shown to be correlated with the amount of oxygen in the water and size is often used in defining taxonomic species. So here we have one structural element that does depend directly upon the environment. This bears out the statement of Shelford (11 a, p. 593) that animal behavior is usually plastic while animal structure is only slightly plastic. Since aside from the size, the taxonomic characters of isopods from the two habitats are the same, the animals cannot be referred . to as a stream ‘form’ as contrasted with a pond ‘form,’ because in its general usage form is applied to a morphological entity. How- ever, the term ‘mores’ as used in general in the Concilium Biblio- graphicum, and as specifically defined by Shelford (11, p. 30) 342 W. C. ALLEE does exactly express the difference between the pond and stream isopods. Thus we are dealing with the pond mores and the stream mores of Asellus communis, which depend on environ- mental rather than on hereditary differences for their distinctive features. In the isopods used, these two mores occur in the same taxo- nomic species, but in the amphipods, Mr. Saunders (1.c.) found a different species in the ponds from that in the streams. This means that with the amphipods studied the natural mores would run parallel with the taxonomic species. But the pond mores could be transferred directly into stream mores without affecting their structural characters. Thus the mores or ‘ecological species’ are independent of taxonomic species, and a single eco- ° logical species may be composed of half a dozen taxonomic species or of only a fraction of one. The general conclusion to be drawn from this series of experi- ments is that in the isopod, Asellus communis, the rheotactic reaction is dependent upon the metabolic state of the animal for its degree of positiveness and that the natural or experimental conditions which affect the metabolic state of the animal, change its rheotactic response. That is to say, the rheotactic reaction is here an expression of the metabolic condition (physiological state) of the isopod and may be controlled by those factors known to control animal metabolism. It is a pleasure to acknowledge my indebtedness to Dr. C. M. Child for his many helpful criticisms but my best thanks are due to Dr. V. E. Shelford, who first suggested this problem and under whose direction the work has been done. BIBLIOGRAPHY Auten, W. C. 1911 Seasonal succession in old forest ponds. Trans. Ill. Acad. Sci., vol. 4, pp. 126-131. Auten, G. D. 1910 Rheotaxis of Planaria dorotocephala. Unpub. Master’s Thesis; Library of Univ. of Chicago. Banta, A. M. 1910 A comparison of the reactions of a species of surface isopod with those of a subterranean species. Jour. Exp. Zoél; vol. 8, pp. 243- 310; 439-488. RHEOTAXIS IN ISOPODA 343 BirGe, E. A. AND JupAy, C. 1911 The internal lakes of Wisconsin. Wis. Geol. and Nat. Hist. Surv.; Bull. 22, 259 pp. Boun, G. 1903 De l’évolution des connaissances chez les animaux marins lit- toraux. Bull. Inst. Gén. Physiol.; no. 6, 67 pp. 1903 a Les Convoluta roscoffensis et la théorie des causes actuelles. Bull. d. Hist. Nat., pp. 352-64. 1905 Attractions et oscillations des animaux marins sous I’influence de la lumiére. Inst. Gén. Psychol., Memoires I. 110 pp. 1908 Les facteurs de la rétraction et de l’épanouissement des Actinies. Comp. Rend d. Séances, Soc. d. Biol., Paris, vol. 44, pp. 1163-6. Cuitp, C. M. 1910 Analysis of form regulation with the aid of anaesthetics. Biol. Bull., vol. 18, pp. 161-178. ‘ 1912 Dynamics of morphogenesis and inheritance in experimental reproduction. Jour. Exp. Zodél., vol. 13, pp. 103-152. Cotton, H. S. 1908 Some effects of environment on growth of Lymnaea colu- mella, Say. Proc. Ac. Nat. Sc., Philadelphia, pp. 410-47. Cusuny, A. R. 1910 Pharmacology and therapeutics, 10th edition. 744 pp. Philadelphia. GamBLE, F. W. anp KEEBLE, F. 1903 The bionomics of convoluta roscoffensis ‘with special reference to its green cells. Qr. Jour. Micr. Sci., vol. 47, pp. 363-431. Geprert, J. 1899 Uber das Wesen der Blausiurevergiftung. Zeitschr. klin. Med., vol. 15, pp. 208-307. Harper, E.H. 1905 Reactions to light and mechanical stimulations in the earth worm, Perichaeta bermudensis (Beddard). Biol. Bull., vol. 10, pp. 17-34. HaLpane, J. AND Situ, L. J. 1897 The absorption of oxygen by the lungs. Jour. of Physiol., vol. 22, pp. 331-59. Houmus, 8. J. 1903 Sex recognition among Amphipods. Biol. Bull., vol 5, pp. 289-292. 1905 The selection of random movements as a factor in phototaxis. Jour. Comp. Neur., vol. 15, pp. 98-112. JenniINGSs, H. 8. 1904 b Reactions to stimuli in certain Rotifera. Carnegie Inst. of Wash., Publ. 16, pp. 73-88. 1904 d_ Physiological states as determining factors in the behavior of lower organisms; ibid., pp. 109-127. 1906 Behavior of the lower organisms. 366 pp. New York. Linctez, D. J. 1902 The importance of sodium chloride in heart activity. Am. Jour. Physiol., vol. 8, pp. 75-99. Lors, J. 1906 Uber die Ursachen der Giftigkeit einer reinen Chlornatriumlésung und ihrer Entgiftung durch K und Ca. Biochem. Zeitschr., vol. 2, pp. 81-111. Martin, E.G. 1906 A study of the absorption and consumption of oxygen in heart tissue. Am. Jour. Physiol., vol. 15, pp. 303-320. Moestus, K. 1873 Die Bewegung der Thiere und ihr psychischer Horizont. 20 pp. Kiel. Mast, S. O. 1903 Reactions to temperature changes in Spirillum, hydra, and fresh-water Planarians. Am. Jour. Physiol., vol. 10, pp. 165-190. 1911 Light and the behavior of organisms. 410 pp. New York. 344 W. C. ALLEE Peart, R. 1903 The movements and reactions of fresh-water Planarians. A study in animal behavior. Qr. Jour. Micr. Sci., vol. 46, pp. 509-714. Préron, H. 1908 De Vinfluence l’oxygéne dissous sur le comportment des in- vertebrés marins. Comp. Rend. de Séances., Soc., d. Biol., Paris, vol. 44, pp. 886-8; 955-7; 1020-2; 1061-3; 1161-2. PrREYER, W. 1886 Uber die Bewegung der Seesterne. Mitth. a. d. zool. Stat. z. Neapel, vol. 7, pp. 27-127; 191-233. Report of committee on standard methods of water analysis to the Lab. Sect. of The Am. Health Ass. 1905. Jour. Inf. Diseases. Supplement, no. i ppa lSI41e RicHARDSON, Harriet 1905 Monograph on the isopods of North America. Bull. U. 8. Nat. Mus., no. 54, 727 pp. Romanes, G. J. 1885 Jellyfish, starfish and sea urchins. 323 pp. New York. Saunpers, W. J. 1911 A study of behavior of amphipods with notes on their life histories. Unpub. Master’s Thesis; Library of Univ. of Chicago. SHEetrorpD, V.E. 1911 Ecological succession. I. Stream fishes and the method of physiographic analysis. Biol. Bull., vol. 21 pp. 9-386. 19lla Ecological succession. II. Pond fishes. Biol. Bull., vol. 21, pp. 127-151. 1911 b Physiological animal geography. Jour. Morph. (Whitman Volume), vol. 22, pp. 551-617. SHELFORD, V. E. AND ALLEE, W. C. 1912 An index of fish environments. Sci- ence, N.S., vol. 36, no. 916. Smitu, AurELIA C. 1902 The influence of temperature, odors, light, and con- tact on the movements of earthworms. Am. Jour. Physiol., vol. 6, pp. 459-486. Spauutpina, E.G. 1904 An establishment of association in hermit crabs, Eupa- gurus longicarpus. Jour. Comp. Neur., vol. 19, pp. 49-61. v. Urxkiitn, J. 1897 Uber Reflexe bei Seeigeln. Zeitschr. f. Biol., vol. 34, pp. 298-318. 1899 Die. Physiologie der Pedicellarien. Zeitschr. f. Biol., vol. 37, pp. 334-403. 1900 Die Physiologie der Seeigelstachels. Zeitschr. f. Biol., vol. 39, pp. 73-112. YerKEs, R. M. 1902 Habit formation in the green crab, Carcinus granulatus. Biol. Bull., vol. 3, pp. 241-244. YERKES, R. M. anp Hiaerns, G. E. 1903 Habit formation in the crawfish, Cam- barus affinis. Harvard Psychol. Studies, no. 1, pp. 565-577. STUDIES ON CHROMOSOMES VIII. OBSERVATIONS ON THE MATURATION-PHENOMENA IN CER- TAIN HEMIPTERA AND OTHER FORMS, WITH CONSIDERATIONS ON SYNAPSIS AND REDUCTION EDMUND B. WILSON From the Department of Zoology, Columbia University NINE PLATES CONTENTS Maar OGG ROTI ec ey echt eae ope 20s 4 cco i eae a ee hse ein 346 I. The maturation-divisions in Oncopeltus and Lygaeus with reference to the sex-chromosomes i Phe diploid chromosome=groups: 3.0)... Gbdscscot oc cee te oe 390 As GUNS invas, SMM OO AKO-CUNMISHOIN. S56 goaccococavcugsacdodncocccsoces 351 SoU Ne PNG eSISe Re ca. Osteria e wees «tA se eR Se aoe Fn eae eT a eee 354 4. The second spermatocyte-division..................2...% sh onnchs yeeree 355 5. Size-relations of the sex-chromosomes in Oncopeltus............... 357 II. The growth-period (PRO UPHINENON EME GUAGE. A. acest cee he oe cies Meee ae ee ee OO 2. Stages a tod. The pre-synaptic stages. Comparison with other I OSXCLS) wigs beth aiees SEE Hid Ia EE Ree Tes Bele att) Oa ol b clo biomes b 362 Gio ISTEIA, Menialsyorsisy, Chin NV ASCIIE Mag aerate eintaldo bam clad 5 gaackeolssonimecsax 376 4. Stages fandg. Post-synaptic spireme (pachytene, diplotene), and HOS CLUTIOTS Ore COMICS Reoadee soo doolsoAbasoonuopocassosopouad 377 De SuagesstOna ebhemprophasess.c: 2s eee ere ee cir ore ert neice 381 6. Comparative considerations regarding the growth-period.......... 387 7. Comment on the sex-chromosomes in Oncopeltus.............--+-+ 390 IIL. Critical considerations on the maturation-phenomena based on a compari- son of the Hemiptera with Tomopteris, Batracoseps and other forms I INNS C WES MOON Eh ON cooeanbecooe son oeedulecoasacyonsbiaddon code 391 2s ithe questionofthereduction-division. «2-4-0 .-6-es-0- ats. e 407 Sen echromosomesa n@nereditye 1 ae oe eee ae eater er 419 345 THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 13, NO. 3 ocTOBER, 1912 346 EDMUND B. WILSON INTRODUCTION In this paper are described observations on certain phases of the maturation-process in Oncopeltus fasciatus (Dall.), Lygaeus bicrucis (Say), and some other Hemiptera, together with the results of a comparison of these species with some other insects, with Tomopteris and with Batracoseps. It was my original object to clear up the relations of the sex-chromosomes in Onco- peltus and to trace as completely as possible their history in the maturation-process; but in doing this it was found necessary to take into consideration many other features of the spermatogene- sis, and I will take this opportunity to present some conclusions based on a broader study of these problems on which I have long been engaged. In respect to the sex-chromosomes, Oncopeltus is of especial interest because it stands on the border line between species in which the X- and Y-chromosomes are visibly unequal in size, and in which a corresponding visible difference appears between the diploid chromosome-groups of the two sexes, and those in which such sexual differences can not be seen.! In my fourth ‘Study’ (09 a) Oncopeltus was classed with Nezara hilaris as an example of the latter class of cases; and much theoretic impor- tance has been ascribed to both these forms as indicating the pos- sibility or probability that the spermatozoa are really sexually dimorphic even when no visible evidence of this is shown by the chromosomes. In my seventh ‘Study’ (11 a) ) I showed, con- trary to my original account, that a dimorphism of the sper- matid-nuclei is in fact visible in Nezara; but in regard to Oncopel- tus judgment was reserved as I was still baffled by apparently contradictory data. I am now in a position to clear up these ! The first account of Oncopeltus was given by Montgomery (’01), who described the sex-chromosomes (‘chromatin-nucleoli’) as of equal size in the male, and found that they remain always separate, without fusing at the time of general synapsis, and divide separately in the first spermatocyte-division. Subsequently (’06) he added that these chromosomes (now called ‘diplosomes’) conjugate to form a bivalent after the first division, and undergo disjunction in the second division. In my fourth ‘study’ (’09 a) I briefly confirmed these accounts, and stated that the female diploid chromosome-groups are not to be distinguished by the eye from the male. id STUDIES ON CHROMOSOMES 347 contradictions and to announce a definite result. Oncopeltus is indeed a case in which the X- and Y-chromosomes are very often sensibly equal in size, and in which the sexual differences of the dip- loid groups are too elusive to be certainly distinguished by the eye. These differences, nevertheless, almost certainly exist. In cer- tain individuals a distinct size-difference between X and Y is clearly evident in a large percentage of the cells at every stage of the spermatogenesis; and even in individuals where they usually appear equal, inequality is unmistakably seen in a small percent- age of the cells. Aside from this, the close similarity—almost identity—between Oncopeltus and Lygaeus bicrucis in all other features of the spermatogenesis makes it extremely probable that the same essential relations of the chromosomes to sex exist in both, though they are only clearly obvious in Lygaeus. Like many other insects of this order, Lygaeus and Oncopeltus are distinctly unfavorable objects for the direct study of synapsis and the reduction-division—indeed the problem of synapsis seems to be practically insoluble in these particular forms. They never- theless present some very interesting features for comparison with other forms. In the first place, the history of the sex-chromo- somes may here be traced with almost unique clearness. They may be identified at a very early pre-synaptic period, and followed thence as individual bodies through every later stage up to the time of their final delivery to the spermatid-nuclei. Every step may be followed in their conjugation and subsequent disjunc- tion without any intervening process of fusion. In case of these particular chromosomes, therefore, I consider synapsis and dis- junction to be indisputable facts. It is far otherwise with the ordinary chromosomes or ‘autosomes.’ It is extremely difficult to gain any clear idea of their behavior in the synaptic period, and I fear quite impossible to trace them individually through the growth-period. On the other hand, their behavior in the pre-synaptic period and in the maturation-prophases exhibits some very interesting features when compared with other forms in which the process of synapsis and its sequel are more accessible to observation. In making such a comparison I have been for- tunate in the opportunity to make use of some remarkably fine 348 EDMUND B. WILSON preparations of other investigators, to whom I am under great obligations. To Professor McClung I owe the loan of a beautiful series of orthopteran preparations, especially of Phrynotettix, Mermiria, Chortophaga and Achurum, which display on a larger scale some of the same phenomena seen in the pre-synaptic stages of the Hemiptera, and leave no doubt of the close parallel between the two groups in this regard. Even more, however, I am in- debted to Dr. and Madame Schreiner, and to Professor Janssens, for some of their admirable original preparations of Tomopteris and Batracoseps, which have enabled me to make a prolonged study of the phenomena of synapsis in these classical objects. In particular, two magnificent slides of Batracoseps by Janssens demonstrate both the complete seriation of the stages and the finest details of the nuclear structures with incomparable clear- ness. Though I have also made many preparations of this form, as well as of Plethodon and other Amphibia, I must admit my failure to equal in all respects the standard set by the slides of Janssens. word, for the ambiguity has arisen simply through a misunderstanding of Moore’s meaning. He applied the term ‘synaptic phase,’ or ‘synapsis,’ to the series of changes following the last diploid division (during the ‘rest of transformation’) STUDIES ON CHROMOSOMES 349 mals (perhaps also in the Orthoptera) the conjugation is a side by side union, or parasynapsis. On the other hand, the evidence of a ‘reduction-division’ in the ordinary use of the term—i.e., the disjunction of the same chromosomes that unite in synapsis— seems to me to be far short of a demonstration. In these forms synapsis is followed by a union so intimate that no adequate evi- dence of duality can for a time be seen in the resulting bivalents. I do not for this reason argue against the conception of the reduc- tion-division. On the contrary, I shall offer new considerations in favor of this conception in a somewhat modified form; but in case of the autosomes it must for the present rest mainly upon indirect evidence. In this respect the autosomes differ notably from the sex-chromosomes, at least in the male sex; and this difference may be of significance for some of the most interesting phenomena of sex-heredity. in the course of which the apparent number of chromosomes is reduced to one-half. “There are thus, after the rest of transformation, only one half as many chromo- somes, i. e., Separate chromatin-masses, as there were before, and the halving of their number, being brought about while the nuclei are still at rest, is to that extent comparable to what is now known to go forward during the maturation of the reproductive elements of plants. I therefore propose the term Synaptic Phase (from ovvarrw, to fuse together) to denote the period at which this most impor- tant change appears in the morphological character of reproductive cells’’ (’95, p. 287. Insubsequent pages the phrase ‘synaptic phase’ is often shortened to ‘synap- sis’in the same sense). This ‘most important change’ is obviously the halving of the number of chromosomes; and nowhere in his paper is the word applied to the contraction-figure, though the latter is stated to be ‘‘characteristic of this particular phase in the spermatogo- and ovogenesis of a great variety of animal forms”’ (p. 305). Though there was, perhaps, some obscurity in his original use of the word, all doubt as to Moore’s meaning is removed in a later paper, published jointly with Farmer (’05), where synapsis is precisely defined as ‘‘that series of events which are coricerned in causing the temporary union in pairs of pre-maiotic chromosomes”’ (p. 490). The fact that so many later writers have misapplied it should not debar us from the continued use of so convenient and appropriate a term—one that seems particularly fitting if it be a fact, as a number of excellent observers have concluded, that synapsis is followed by actual fusion. I can dis- cover no reason why McClung’s term ‘synizesis’ should not be generally employed for the contraction-figure, as it already is by most American writers. 390 EDMUND B. WILSON I. THE MATURATION-DIVISIONS IN ONCOPELTUS AND LYGAEUS WITH REFERENCE TO THE SEX-CHROMOSOMES In Oneopeltus the diploid number of chromosomes is sixteen in both sexes (figs. 1 to 5, photos, 1, 2) in Lygaeus fourteen (fig. 6), and the second spermatocyte-division shows half these numbers that is, eight in the former case, seven in the latter. The first division shows in each case one more than the haploid number, owing to the fact, repeatedly described heretofore in other Hemip- tera, that in the first division the X- and Y-chromosomes divide as separate univalents, while in the second they are united to form a bivalent. In both Oncopeltus and Lygaeus these chro- mosomes conjugate in the final anaphase of the first division, just as the cell is about to divide. 1. The diploid chromosome-groups The spermatogonial and oogonial divisions require but brief description since they present no striking features and the size- differences are but slightly marked. In Lygaeus bicrucis (fig. 6) the fourteen chromosomes are in the main similar to those of L. tureicus as described in my first and third ‘Studies’ (05, ’06) though the Y-chromosome is relatively smaller in the latter spe- cies. The X-chromosome can not be identified by the eye, but must be at least twice the size of the Y-chromosome, as indicated by the maturation-divisions and by the spermatogonial groups themselves. Unfortunately my material of this species does not show a single good equatorial plate in the female; but the relations are here no doubt the same as in L. turcicus. In Oncopeltus, of which I have abundant material of both sexes, the size-differences of the chromosomes are even less marked than in Lygaeus, and it is impossible to identify pairs of different sizes. Careful study fails to reveal any differences between the diploid groups of the two sexes that are sufficiently marked or constant to give any certain result (figs. 1 to 5). In the male one chromosome not infrequently is somewhat smaller than the others (fig. 2), and this may be the Y-chromosome; but very often this is not evident, even in other spermatogonia from the same cyst (figs. 1,3). As will be shown beyond, the X- and Y-chromosomes STUDIES ON CHROMOSOMES é Be) are often somewhat unequal in later stages; but this, too, is incon- stant. Oncopeltus is in fact, therefore, a form in which the sexual differences of the chromosome-groups are too slight or too elusive to be distinguished by the eye. It is, however, perfectly certain that an XY-pair is present, the members of which show all the characteristic peculiarities of behavior that characterize these chromosomes in other forms. 2. The first spermatocyte-division The maturation-divisions are shown with remarkable clearness in Oncopeltus and Lygaeus—indeed, either of them might be taken as a model of those Hemiptera in which a simple X Y-pair is present. The following account applies primarily to Oncopel- tus, Lygaeus being described only by way of comparison. In the first division appear nine separate chromosomes in a grouping of remarkable constancy. Seven of the nine bivalents are grouped in an irregular ring, near the center of which lie the univalent X- and Y-chromosomes, side by side but not in contact (figs. 8,9, 11, photo. 3). The constancy of this grouping appears from the following data. Two hundred clear polar views, taken at random, did not show a single case of more than nine chromo- somes; plus variations of number in this division (such as are occasionally seen in many species) must therefore be very rare— indeed, I have never seen such a case.’ Of the two hundred cases, one hundred and seventy-two showed the grouping just described. In the remaining twenty-eight the deviations were unimportant; the ring may show a gap at one side (figs. 10, 12), one of the biva- lents may lie inside it (fig. 10), or (rarely) one or both the sex- chromosomes may lie in the ring (fig. 13). In only two eases did the sex-chromosomes not lie side by side; in these, they were separated by one bivalent (fig. 13). The size-relations alone sufficiently indicate that thetwo small central chromosomes are the univalent sex-chromosomes (a fact 3 Apparent minus deviations are of course common, but are disregarded because evidently due in most (all?) cases to the fact that one or more chromosomes lie outside the plane of section. ane EDMUND B. WILSON fully established by study of the growth-period and the pro- phases) ; for, were one or both bivalent, the spermatogonial groups should show one or two corresponding pairs, which is not the case. In Lygaeus (fig. 7) the grouping is the same, but only six bivalents are present, and the sex-chromosomes are conspicuously unequal in size. The composition of these chromosomes is better seen in smears than in sections, and better in Protenor (described beyond) than in either of these forms. In sections, side views of the full meta- phases (figs. 14 to 17, photo. 4) usually show allof the chromosomes as simple dumb-bell figures, though indications of a quadripartite form sometimes appear. In smears the bivalents are often seen to be quadripartite, owing to the presence of a longitudinal split in addition to the transverse constriction; but this never appears in the univalents. All the chromosomes alike divide ‘trans- versely’—that is; across the constriction of the dumb-bell. In case of the bivalents, therefore, the early anaphase-chromosomes are double bodies, while the sex-chromosomes are single, but this contrast only appears clearly in smears, owing to the close union of the two halves. In this respect the relations are less clearly seen in these forms than in some others, such as Anax, where the anaphase-chromosomes are clearly double (cf. Lefevre and Mc- Gills ’08), or Aprophora, where the same condition is conspicuously shown (Stevens, ’06). The anaphases are of particular interest because, as has been mentioned, a conjugation between the X- and Y-chromosomes takes place in the later stages. As the division begins and the daughter-chromosomes are separating, a marked contrast in form often appears between the sex-chromosomes and the autosomes (figs. 19 to 21). The latter are more or less extended transversely and often show a slight constriction, thus giving evidence of their double nature, which is accentuated by the very conspicuous double fibres by which they are connected. The latter are so thick, and stain so deeply, as to appear as if spun out from the chromosomes themselves (as has been noted by other observers). On the other hand, the sex-chromosomes do not show such a con- striction, remaining nearly circular in outline, while the connect- STUDIES ON CHROMOSOMES B08 ing fibers are much less conspicuous, and often appear single. Up to the middle anaphases the sex-chromosomes remain always separate (figs. 18, 19). In the later anaphases all the chromo- somes draw more closely together, and often come more or less into contact, though without losing their original grouping; but in case of the autosomes the contact is but casual and tem- porary, while the sex-chromosomes become definitely attached to each other to form a dumb-bell shaped body at the center of the group (figs. 19, 21). By this process the total number of sepa- rate chromatin-elements is reduced from nine to eight (the hap- loid number). In Lygaeus the process takes place in exactly the same way and may be seen with equal clearness, both in polar views and in side-views. In both species polar views of the final anaphases show that the chromosomes, save for their more crowded condition, have retained the same grouping as in the metaphase (figs. 24 to 29), the X Y-bivalent being at the center, surrounded by the other chromosomes in the form of a ring. These facts are seen so clearly and in so great a number of cases as to remove every doubt that in these species the conjugation between X and Y regularly takes place at the poles of the spindle before the first maturation-division has been completed.‘ In Lygaeus the X Y-bivalent thus formed is readily distinguish- able by the inequality of its two components (figs. 28, 29, 46). In Oncopeltus it is often not thus marked but its identity is no less certainly revealed in another way. As the figures show, the autosomes still show but slight indication of a transverse constric- ’ tion, and can hardly be described as dumb-bell shaped until a later period. The X Y-bivalent, on the other hand, is invariably deeply constricted, so as to have a conspicuously dumb-bell shape, and it often still appears like two chromosomes that are merely in contact. This characteristic difference persists throughout the entire interkinesis, and is still perfectly obvious in the ensuing metaphase of the second division. *T described and figured this process in the case of Coenus in my first ‘Study’ (05 b) but did not recognize its constancy. I now incline to think that it will be found to occur in the same way in many other forms. 354 EDMUND B. WILSON 3. The interkinests The interkinesis has hitherto been very briefly treated by my- self and other observers of the insects, because in most species the chromosomes are so closely crowded at this time as to preclude accurate study. Lygaeus (at least in my material) is no excep- tion to this, but Oncopeltus fortunately shows every stage of the interkinesis with remarkable clearness. There is no ‘resting stage’ between the two divisions, no nuclear vacuole is formed, and both the chromosomes and the centrioles retain their individ- uality throughout. At the moment when the equatorial furrow has appeared and the conjugation of X and Y has taken place, the centrioles are already rather far apart, and still lie at some distance from the chromosome-group (fig. 22). All the achromatic elements are now so delicate that it is difficult to make sure of the exact structure; but it is certain that each centriole is surrounded by a small, but very distinct aster, and the two seem to be connected by a delicate central spindle. As the cell divides several other changes take place. The chromosomes, without otherwise changing their grouping become still more crowded together, and thus become massed in a nearly flat plate, while the centrioles move still far- ther apart (fig. 23). Shortly after the division these relations are unchanged save that the centrioles are still farther separated and lie nearly on opposite sides of the chromosome-group. The asters are still present, and between them lies a rather large, irregularly spindle-shaped area. It is difficult to say whether this should be regarded as an actual spindle; but delicate fibrillae often may be seen extending into it from the poles. The chromosome-group lies somewhat excentrically within this area in the form of an irregular flattened plate. In side-view (fig. 30) it is usually impossible to distinguish more than a few of the chromosomes. In face view also, the crowding is often so great that the grouping can not be exactly made out. Here and there, however, it is evident that the original grouping has not been lost, and occasionally plates are to be found in which every chromosome may be clearly seen (figs. 31, 32). Study of such STUDIES ON CHROMOSOMES Be) cases makes it certain that no fusion or process of disintegration has taken place, nor is any evidence of a nuclear vacuole to be seen. The chromosomes still retain the same form as in the pre- ceding anaphases, the X Y-bivalent lying near the center, and still very clearly distinguished by its markedly bipartite form.