PW?*^ ■w^vvuM^iL . 1 1 u lu rd«i^an^f^!«|W(PiM QH431 C377 ^ortI| Carolina ^^taie College This book was presented QlH431 NORTH CAROLINA STATE ;. M ,'ERSITY LIBRARIES S009607 3 Q J0SEPH.BU2IC|{A BOOKBIWOER'S BAIT ! MO^e, MD. GHEEHSBORO.N.C. WA3HIN0TeN.0.C. This BOOK may be kept out TWO WEEKS ONLY, and is subject to a fine of FIVE CENTS a day thereafter. It is due on the day indicated below: «^rt«l 22Dec35j FE8 -419 yi 6Jul42E " FEB25' 81 efeb '50^ 3lJan'562 ^PR 1 - 1! ifli 22May58V 7i«»'6|P jSPR 1 89 m ^Ww IBPw^ ^ " • * • MflV ?^) 2008 STUDIES OF HEREDITY m RABBITS, RATS, AND MICE BY W. E/ CASTLE RESEARCH ASSOCIATE OF THE CARNEGIE INSTITUTION OF WASHINGTON Published by the Carnegie Institution of Washington Washington, 1919 CARNEGIE INSTITUTION OF WASHINGTON Publication No. 288 [From the Laboratory of Genetics of the Bussey Institution] Copied of this Book were first issued JUL 24 \m PRESS OF J. B. LIPPINCOTT COMPANY PHILADELPHIA CONTENTS. Part I. Further Experiments upox the Modifiability of the Hooded Character in Rats. PAGE Selected races crossed with wild 2 Part II. The Inheritance of White-spotting in Rabbits, with Special Reference to Gametic Contamination. Dutch 5 English 19 Relation of Dutch to English 25 Part III. Observations on the Occurrence of Linkage in Rats and Mice. Tables 37 Bibliography 56 5r?S12 STUDIES OF HEREDITY IN RABBITS, RATS, AND MICE. By W. E. Castle. I. FURTHER EXPERIMENTS UPON THE MODIFIABILITY OF THE HOODED CHARACTER OF RATS. In publications Nos. 195 and 241 of the Carnegie Institution of Washington reports have been made on the results of a series of experi- ments designed to show to what extent a mendelizing character, the hooded pattern of piebald rats, may be altered by selection or by crossing. At the last report (Castle and Wright, 1916) the plus- selection series had been carried without out-crossing through 16 successive generations, in the course of which the mean grade of the offspring had advanced from +2.05 to +4.13, in terms of the arbi- trary grading scale depicted in both previous publications. Since then the plus-selection series has been carried through four additional generations of selection (17 to 20) and the mean of the selected race has been raised to 4.61. In some respects this part of the series is less satisfactory than that previously reported on, because smaller numbers of animals were available from which to select and the selec- tion has therefore been less rigorous. The race has unmistakably fallen off in vigor and fecundity in later generations. It is uncertain whether this should be ascribed to inbreeding alone, uncorrected by selection for vigor (as in Miss King's experiments), or to increase in the prevalence of disease, or to both causes. Certain it is, however, that notwithstanding increasing care in regard to feeding and sanita- tion a very large proportion of our breeding-pens in the case of the selected races produce no young at all. Tables 1 to 4 show in detail the grade distribution of the young produced by plus-selected parents of generations 17 to 20. The numbers of young produced in each of these generations are respec- tively 351, 420, 280, and 92. The mean grade of the young advances from 4.13 (in generation 16) to 4.48 (in generation 17), remains prac- tically stationary in generations 18 and 19 at 4.46 and 4.49 respectively, and then advances again (in generation 20) to 4.61, The minus-selection series — which at the last report had been car- ried through 17 generations, with an advance in mean grade of the young from grade minus 1.00 to grade minus 2.70 — has now been 1 Library N. C, State CoUen^e 2 HEREDITY IN RABBITS, RATS, AND MICE. carried through 4 additional generations of selection. (See tables 5 to 8.) But this race has shown even poorer vitality than the plus- selection race, so that in order to keep it alive practically no young could be rejected as parents and consequently no further progress has been made. The numbers of young recorded for the four addi- tional generations have been 330, 130, 79, and 35 respectively, and their mean grades -2.84, -2.89, -2.78, and -2.74. The race is now practically stationary in grade, but seems likely soon to become extinct despite our strongest efforts to keep it alive. Notwithstanding the fact that the race is verging on extinction after long-continued close breeding, the variability of the hooded character is still as great as ever. The standard deviation ranges from 0.25 to 0.45 within about the same limits as in the previous 17 generations from minus- selected parents. In the plus-selection series the standard deviation was also fully as high in the last generations as it had been in the previous 10 generations. Only the initial 7 generations had shown an appreciably higher variability. SELECTED RACES CROSSED WITH WILD. We may now inquire what happens to the races modified by selec- tion in opposite directions, when they are crossed with an unselected, non-hooded race, the wild race. This question was considered in some detail, so far as the plus-selected race is concerned, in a previous publication (1916), where it was shown that a cross of the plus-selected race reduced the grade of the hooded character, undoing in a measure the work of selection. Selected animals which mated with their like should have produced young of about mean grade 3.75, actually pro- duced hooded grandchildren, extracted in F2 from a wild cross, of mean grade 3.17, a falling off in grade of over 0.50. A second cross with wild showed no further falling off, but instead a movement in the reverse direction to 3.34 (a movement probably not significant, in the light of further experiments). A third cross with wild has been made on a small scale; 19 hooded grandchildren extracted in F2 from this third cross have a mean grade of 3.04. (See table 9.) It seems probable, therefore, in the case of the plus-selected hooded char- acter, that the maximum efTect exerted by the residual heredity of the wild race is to reduce the hooded character in grade by about three- fourths of a grade. Selection in 10 previous generations had elevated the grade of the hooded character by about If grades. A cross with wild eliminated less than half of this change. The remaining change must be ascribed to changes effected in the course of selection. The minus-selected hooded race has also been crossed with this same wild race. Originally of grade -1, it had been altered by 15 generations of selection to the extent of about 1|- grades, to mean grade -2.54. Females of generations 15 to 16 (table 10) and of MODIFIABILITY OF HOODED CHARACTER 3 grade —2.75 were crossed with wild males of the same race used in the crosses of the plus-selected race. The extracted hooded Fo young were highly variable, ranging in grade from —2.25 to +3.00, mean — 0.38, a remarkable change in the plus direction of over 2 grades. A second cross with the wild race (table 11) brought about a further movement of the mean in a plus direction but by a somewhat smaller amount, 1.39 grades, the mean of the twice-extracted hooded young being -fl.Ol. A third cross with the wild race (table 12) brought still further contamination of the hooded character, which now ceased to vary below grade +1.00, and had a mean of +2.55 in the case of over 100 thrice-extracted hooded young, this being a change in the plus direction of 1^ grades. It will be observed that the hooded grand- children of 0^2068, table 12, the most plus in character of the hooded grandparents, are very similar in grade to the hooded young resulting from three crosses of the plus-selected hooded character with the same wild race (table 9). In other words, the same wild race, when its residual heredity is made fully effective by repeated crosses, brings both the plus-selected and the minus-selected hooded lines to a pheno- type of common grade. This shows, contrary to my earlier opinion, that what has really happened in the case of the selected races was more largely due to residual heredity than to any change in the gene for the hooded character itself. ]My critics have been wrong when they insisted that selection could not change racial characters that mendelize and change them permanently, and when on this ground they denied to gradual change through selection an important part in the evolution of characters and thus of races. But my critics have been right w^hen they insisted that evidence is wanting that change in single genes occurs other than spontaneously, uninfluenced by sys- tematic selection. II THE INHERITANCE OF WHITE-SPOTTING IN RABBITS. WITH SPECIAL REFERENCE TO GAMETIC CONTAMINATION.! One of the commonest color variations of mammals is white- spotting — the occurrence of whollj^ unpigmented areas in the skin and the hair arising from it. Small unpigmented areas are frequently found in the coats of wild mammals, as, for example, in the fur of wild mice, rats, the common North American rabbit {Lepus sylvaticus), and cavies {e.g., Cavia cutleri). The white-spotting found in these wild forms is usually not extensive. It consists of a white ''star" in the forehead or a spot on the chest, or at the end of the tail, or on a foot. Such locations of the white-spotting suggest a deficiency of pigment in the skin, either where it closes together in the median line during development of the embryo or at the extreme limits of its peripheral extension during development. At places where the skin regenerates after injury, even in self-colored animals, a white spot frequently develops. This is especially noticeable on the backs and shoulders of horses where the harness has ''galled" them. That such shght congenital deficiencies of pigment as occur in wild mammals are hereditary has been shown by Little in the case of the house- mouse, and by Phillips and myself (unpublished observations) in the case of the field-mouse, Peromyscus. We observed in a colony of Peromijscus reared from animals taken in Massachusetts the occur- rence of individuals having tails partly or wholly white. This con- dition was found to be a Mendelian recessive character in crosses. After one or two selections of white-tailed individuals, we noted exten- sion of the white area on to the belly. In some wild mammals the white-spotting is more extensive, taking the form of a definite pattern, as in skunks, the harp-seal, and the Malay tapir. The color pattern of skunks, w^hile characteristic, is known to vary slightly, the value of a pelt increasing with the amount of black which it contains, a fact which the incipient industry of skunk-farming in the United States notes with interest. Selective breeding is being directed toward the establishment of all-black strains and no doubt it will ultimately be successful. White-spotting is so common in the domestic animals as to need no comment. White-spotting in more or less definite patterns char- acterizes the majority of our breeds of cattle, horses, dogs, and swine. Often the pattern is so definite and so strongly inherited as to con- stitute a sort of trade-mark of breed purity, as in Hereford (white- faced) cattle, Dutch belted cattle, and Dalmatian coach-dogs. Much » Valuable assistance in the conduct of this investigation was given by my former pupil, Prof. H. D. Fish. INHERITANCE OF WHITE-SPOTTING IN RABBITS. 5 interest attaches to the inheritance of these patterns in crosses, a subject which has been studied for some years at the Bussey Institu- tion. The present paper will deal with the subject of white-spotting in domestic rabbits. Patterns of white-spotting mendelize without known exception but with some irregularity as regards dominance. In some cases white-spotting is not expressed in the heterozygote, but if expressed at all in the heterozygote its expression is always stronger yet in tlie homozygote. When the character is nearly or quite suppressed in the heterozj'gote, we may call it recessive; when the character is strongly expressed in the heterozygote, we may call it dominant. But neither term is applicable without qualification in the way that recessive is applicable to complete albinism in rodents. With this qualification of terms, it may be said that there occur among domestic rabbits two forms of white-spotting, probably of independent origin and certainly of quite different genetic behavior, since one is recessive and the other dominant in crosses with the same race of unspotted rabbits. The dominant form of white-spotting is found in the so-called English rabbit and its inheritance has been discussed by Castle and Hadley (1915). The recessive form of white- spotting is found in Dutch rabbits, as observed independently by Hurst and by Castle. Punnett assents to this conclusion with the qualification that the inheritance is possibly not that of a simple (one-factor) sort. ^, ,^^, , ^ ' DUTCH. In September 1910 three standard-bred Dutch rabbits were obtained from a fancier who had bred and exhibited prize-winning animals derived from stock imported from England. They resembled grades 7, 8, and 9 respectively (plate 1). The female proved to be sterile and was ultimately discarded. One of the males (cf3037, grade 7) was mated with two heterozygous English does, which produced self-colored blacks as recessives when mated with English bucks of their own race. For the present we shall consider only the non- English young produced by these matings. Such young would be of the same character as those produced by self-colored ani- mals mated with Dutch, since they would arise from a se// gamete furnished by the mother and this would be fer- tilized by a Dutch gamete furnished by the father. Six young of this character were produced of the accom- panying grades. This same male was mated also with 3 Himalayan albino does of a race entirely free from spotting but which lacked the color factor. Potentially these does were self- colored. This cross produced 18 young of the accom- panying grades. (See table 13) ' Grade signifies a self animal, i.e., one without white spotting. Grade. No. 1 o 3 2 1 3 (See table 13). Grade. No. •0 1 2 3 13 2 6 HEREDITY IN RABBITS, RATS, AND MICE. Two does derived from the first-mentioned cross and one derived from the second were employed in various matings presently to be described. The results observed in the case of all three were so similar that they may conveniently be described together. It will be borne in mind that all are Fi hybrids between Dutch and self. When crossed back with the other original pure Dutch buck ( cf3036, grade 9), these three does produced 20 young of the grades shown in table 14. We get here indications of segregation into two groups, one like the Fi mothers in grade, the other like the Dutch father, but no sharp line of division separates the two. The same three Fi females were also mated successively with an Fi male from each of the two crosses already described, with the results shown in table 15. The results are similar in both cases, but it will be noticed that the lower grade Fi male (5029, derived from the Himalayan cross) produced Fz young of slightly lower grade. The F2 range extends from to grade 5 inclusive, average 1.80. The back- cross range was from 1 to 7 inclusive, average 4.60. Certain of the F2 young and the back-cross young of grade 4 or higher, which presumably would be homozygous for the Dutch char- acter, if it mendelizes, were employed in building up a race of Dutch rabbits for further study. This was done by back-crossing the selected does a second time with the pure buck, cf3036, grade 9, with the results shown in tables 16 and 17. Young were obtained which ranged from grade 1 to grade 17, but which grouped themselves round two modes situated at about grade 6 and grade 15 respectively. We shall presently consider the distribution further. These same does were also mated with a male similar in origin to themselves, \az, cf5167 (table 18), a typical and evenly marked Dutch buck of grade 7, produced by the original back-cross (table 14). He bred in all respects like his father (cf3036, grade 9) when mated with the same does, producing a bimodal group of Dutch young of only slightly lower mean grade than the young which his father sired, as might be expected from the fact that his grade was less than his father's grade. (See table 18.) It was now evident that we had secured a race of Dutch rabbits which produced only Dutch young and which derived their Dutch character exclusively from the two bucks 3036 and 3037, and yet which fluctuated in grade around two different modes. In fact, it was soon discovered that the two original Dutch bucks were them- selves heterozygotes of two different types of Dutch pattern which corresponded with the two modal conditions found among their descendants. Our next task was to isolate these two types in homozygous form. This was easy in the case of the higher grade (whiter) type, which proved to be recessive. A male of this ''white" type, 6175, grade 17, when mated with females of the same sort INHERITANCE OF WHITE-SPOTTING IN RABBITS. 7 produced only one type of Dutch young varying around grade IG. (See table 19.) By studying the results of various matings of our Dutch does it was found possible to classify them in three categories: (1) The ^'white" type of grade 15 to 17, which produced only the "white" type when mated with bucks of the same sort, as already described. (See plate 2, fig. 19.) (2) A "dark" type of grade 1 to 7, which when mated with bucks of the white type produced no "white" offspring, but only those of the dark type. These mothers were evidently homozy- gous "dark,'' the other pure type (see plate 2, fig. 20). (3) The third type of doe is scarcely distinguishable from the pure dark type except by breeding test. It consists of heterozygotes between the two types, a little whiter on the average than the dark type, but not conspicu- ously so. When they are mated with "white" bucks, young of two types are produced in about equal numbers, viz, heterozygous darks and pure whites. The original Dutch bucks from which the entire race was derived were both of this heterozygous dark type. Neither the dark type nor the white type isolated from this race of rabbits conforms closely with the ideal Dutch type of the fancier (our grade 8). The one is usually too dark and the other too white. It seems probable that the fanciers in breeding "prize-winners" have consciously or unconsciously been producing heterozygotes, very much as in the case of the Andalusian fowl. Certain it is that all the rabbits which we have produced from this stock, which would have any chance of winning a prize at an exhibition, have been heterozygotes between these two types. While the experiments with standard-bred fanciers' Dutch rabbits were in progress, and after the white and the dark types of Dutch had been isolated, a third type of Dutch was discovered which kept crop- ping out in a stock of black-and-tan rabbits under observation for another purpose. This stock was derived from a single pure-bred black-and-tan buck which had been crossed with various other stocks of rabbits then in the laboratory. The Dutch pattern had been introduced as a recessive character in a certain j^ellow rabbit of unknown pedigree obtained by purchase. When the descendants of this yellow rabbit were bred with each other, certain of the young produced were Dutch marked. This type of Dutch resembled the fanciers' type of Dutch (grade 8, plate 1) so far as the head markings were concerned, but the belt was very narrow and placed far forward over the shoulders. (See plate 2, fig. 21.) Because of its origin within the black-and-tan stock we have adopted the name "tan" Dutch to distinguish it from the other two types. In describing its variations we have used the same set of grades (shown in plate 1) which were used in classifying the variations of the other two types, but it must be understood that rabbits of the 8 HEREDITY IN RABBITS, RATS, AND MICE. dark and of the tan type which are given the same grade are not exactly ahke in pattern. The grade is at-^signed to express roughly the total amount of white-spotting on the animal. A rabbit of grade 3 in the tan series will usually have a whiter head with a broader white spot on the nose but with a narrower collar than one of the same grade in the dark series. With this explanation it may be stated that tan Dutch rabbits bred with each other have produced 40 Dutch young with the grade distribution shown in table 21. The variation is close about grades 3 and 4, the race being very uniform in character for a white-spotted race. In origin it is derived from a single gamete which introduced the character as a recessive into the original yellow ancestor. Having now secured 3 distinct strains of Dutch rabbits, it was our next task to determme what were their genetic relationships to each other, whether they were allelomorphs or due to wholly independent factors; whether due to single or to multiple genetic factors, and whether these factors were constant or variable. Before these ques- tions can be intelligently discussed the variability of each type by itself must first be known. That of the tan race has just been referred to. It is shown in table 21. It will be observed that the mean grade of the young has a tendency to rise with the grade of the parents. The variation of the uncrossed ''dark" race is shown in table 20. The same homozygous buck (6701, plate 2, fig. 20) was mated with ten different homozygous dark does ranging in grade from 2 to 5. They produced 172 young ranging in grade from 1 to 7, mean 3.30. The variation is sho\\Ti graphically in text-figure 1, d. The higher-grade mothers, it will be observed, produce higher-grade young, although the differences are not striking. The variation of the uncrossed ''white" race is shown in table 19 and text-figure 1, w. Two bucks of grade 17 were mated with 8 does of grade 15, 16, or 17; they produced 59 young with the same range of variation as the mothers and of the mean grade 16.25. Again we observe a tendency for the higher-grade mothers to produce the higher-grade young. The same homozygous white buck (6175) which was used in matings recorded in table 19 was mated also with 5 homozygous does of the dark race, with the results shown in table 22 and text-figure 1, Fi, D X W. These matings produced 28 young of mean grade 7.28. All the young, from their parentage, should be heterozygotes between white and dark Dutch, like the original animals from which these two races were isolated. In reality they agree closely with the foundation stock in grade. Table 23 shows the results of matings of homozygous white bucks (one of which was the same individual, 6175, as sired the young of tables 19 and 22) with dark does which were heterozygous for white INHERITANCE OF WHITE-SPOTTING IN RABBITS. 9 and so were like the foundation stock. The young tlius produced numbered 130 and fall into two distinct groups, each varying about a different mode. The two groups apparently do not overlap. Taking the minimal class, grade 12, as on the line between them, there are 65 individuals below this class and G4 above it. The former should be Grade. I 10 II M2 13 14 15 16 17 Text-figure 1. heterozygotes like the young recorded in table 22. Their mean grade is very similar, 7.04 as against 7.28. The latter are evidently homozy- gous white like the young of table 19. They have a similar but slightly lower mean grade, viz, 15.56 as against 16.25. This result indicates a 1 : 1 segregation of white and dark Dutch and that accordingly white and dark are allelomorphs. If so, an ¥■> generation should show a 3 : 1 segregation. To test this point we may summarize frnm 10 HEREDITY IN RABBITS, RATS, AND MICE. tables 16 to 18 all matings between two heterozygous parents (those which produce both dark and white young). In this way we get the totals shown in table 28, ''white X dark." Two groups of young are here shown, one dark (grade 10 or lower), the other white (grade 12 or higher). The numbers of individuals in these groups are 56 and 25 respectively (a rather poor 3 : 1 ratio); their mean grades are 5.82 and 14.40. (See text-figure 1, Fa, DxW.) The dark group evidently includes homozygotes (mode on 5) and heterozygotes (mode on 8) which overlap in the intervening region. Accordingly all facts thus far noted indicate that white and dark are allelomorphic forms of Dutch marking. One other test of this hypothesis is possible. Fi may be back- crossed with the dark race. The result of such a test is shown in table 29, "Fi (white X dark) X dark," and text-figure 1, Fi X D. The expectation here is the formation of two groups of equal size, homozygous and heterozygous dark, with modes on 3 and 7 respec- tively. In reality the intervening grades are the modal ones. This does not disprove segregation. The flat-topped variation curve ob- served is exactly what we might expect from the combination of two simple variation curves which overlap. Compare F2, Sx W, text-figure2. The several facts developed as regards the relation of "dark" to "white" Dutch pattern are presented graphically in text-figure 1. The variability of the uncrossed races is shown at the top; the races are distinct, monomodal, and do not overlap in range. Immediately below is shown the character of Fi. It also is monomodal and it is intermediate. Next lower is shown the variability of F2. The extracted white race is here seen to have a lower mode than the uncrossed white race and the extracted dark race has a higher chief mode than the uncrossed dark race. In other words, the extracted races reappear in Fo in character mutually modified, although their extreme range is the same. The mode of the heterozygotes lies between the modes of the homozygotes, as in Fi just above. The results of back-crosses with each of the parental races are showTi in the lower part of text-figure 1. In the back-cross with the white race (FiXW), the 1 : 1 segregation is unmistakable; in the back-cross with the dark race (FiXD), segregation is obscured by the closeness to each other of the modes for heterozygous and homozygous dark Dutch, which results in producing a composite flat-topped curve. Everything indicates that dark and white are simple allelomorphs, but are quantitatively fluctuating and mutually modify each other in crosses. If they were not allelomorphs, individuals which contained neither form of Dutch should appear in F2. None such is produced. We may next consider whether white Dutch is a simple allelomorph of self or is composite. The pertinent facts are recorded in tables 27 to 29 and are shown graphically in text-figure 2. The same white INHERITANCE OF WHITE-SPOTTING IN RABBITS. 11 male (G175) that was crossed with dark females was also crossed with self-colored (unspotted) females. (See table 27.) There wore pro- duced 33 young, all showing a small amount of white-spotting, ranging Klo. ,20- 10- ^ ^ 20- 10- #. F,,SxW in grade from 1 to 3, mean 1.51 (see text figure 2, Fi, SxW). F2 is somewhat puzzling in character. (See table 28 and text-figure 2.) Nearly 25 per cent of the 191 young recorded are extracted selfs, but 12 HEREDITY IN RABBITS, RATS, AND MICE. if we take the whitest 48 individuals to be extracted "whites" it is evident that they are considerably modified from the condition of the uncrossed race, since their mode Ues at 12 to 14, not at 16 to 17, and the highest grade (17) of the uncrossed white race does not reappear at all. The back-cross of Fi with white must be investigated before one can interpret the F2 result with confidence. But the back-cross (table 29 and text-figure 2, Fi X W) makes it very clear that segrega- tion occurs on a 1 : 1 basis. The 116 young thus produced fall into two groups which do not overlap and each of which is monomodal. Each contains 58 individuals. But the ex- tracted groups show mutual modification. The mode of the lower group is not at grade 1, as in Fi, but at grades 4 to 6, while the mode of the upper group is not at 16 to 17, as in the uncrossed white race, but at 15. The back-cross of Fi with self (table 29 and text-figure 2, Fi X S) gives a variation nearly covering the combined range of self and Fi, as expected, but the two expected groups (if they are distinct) lie so close together that it is impossible to separate them. The several facts developed in crosses of white with self indicate that white is a simple allelomorph of self. If so, and if dark is also an allelomorph of white, as indicated by text- figure 1, then self and dark should be allelo- morphs of each other. Such is probably the case, but crosses of self with dark are incon- clusive because the two conditions are so close to each other on the grading scale that it is difficult to demonstrate segregation. See tables 27 and 29 and text-figure 3. This figure shows (at the top) the variation of uncrossed dark in relation to self. Below is shown the variation of Fi, self being usually dominant. A back-cross with dark (Fi X D) produces less than the expected propor- tion of self (nearly 50 per cent as indicated by Fi) and produces extracted darks of lower mean grade than the uncrossed darks. This is evidence of mutual modification of self and dark in the heterozygote, so that they emerge in the gametes modified. So far, then, the evidence indi- cates that self, dark, and white are all allelomorphs, but that they fluctuate quantitatively and mutually modify each other when asso- ciated in heteroz3'gotes. We may now consider the relation of these three to tan Dutch. If tan is an allelomorph of either of the other forms of Dutch, it should be an allelomorph of all three conditions, in fact sl fourth allelomorph. Grade 1 2 3 4 5 6 7 Text-figure 3. INHERITANCE OF WHITE-SPOTTING IN RABBITS. 13 No. 20- 10 ■ 10- 0^ [^^^ F,,TxW 20 10 ith white have given the results shown in tables 26, 28, and 29, and also graphically in text-figure 4. Fi is intermediate. F2 is likewise interme- diate, but varies to or into the range of the uncrossed races with indications of segre- gation of modified tan and modified white individuals. The ab- sence of selfs in Fo shows white Dutch and tan Dutch to be allelomorphs. The back-cross with white (FiXW) produces in- dividuals varying (as expected) all the way from the Fi to the un- crossed white race, but with an apparent q. tendency to form ^^^^« modes on 10 and 15. These, it seems, represent the mutually modified modal conditions of the Fi and the white race. The extracted whites have tlieir mode low- ered to 15, instead of on 16 or 17, as in the uncrossed white race; and the extracted Fi's have their mode advanced from grade 7 (in tlie orig- inal Fi's) to grade 10. The races dark and tan are similar to each other in grade (see text- figure 5), but differ in the location of their unpigmented areas, as already explained. Dark has a wider collar and a darker head; tan has a narrower collar and a whiter head. Fi is nearly or completely self-colored, since dark tends to make the head pigmented and /rv^ 0- 90- SO- TO' 60- 50- 40- 30- 20- 10- ?^^^S\ -f^ \' iL. Grade I 2 3 4 5 6 7 8 9 10 II I 2 3 4 5 6 7 6 9 10 11 12 13 14 15 16 17 Text-figure 5. with white; (c) tan united with white; and (d) self (neither dark nor tan) united with white. The observed modal values of these four kinds of combinations are respectively, in terms of grades: (a) 10 16 HEREDITY IN RABBITS, RATS, AND MICE. (text-figure 1, W); (6) 7 (text-figure 1, Fi); (c) 7 (text-figure 4, Fi); and (d) 1 or 2 (text-figure 2, Fi). That is, this cross should, under the hypothesis considered, produce a trimodal figure, with a mode for 25 per cent of the individuals at either end of the grading scale and with a still larger mode (for 50 per cent of all individuals) at an inter- mediate point. But what is actually observed is very different. The figure (text-figure 5, Fi X W) is not trimodal but bimodal, and neither of the modes is where the hypothesis demands that modes should be. This is conclusive evidence against the correctness of the hypothesis in question, but is entirely in harmony with the alternative one, that dark and tan are allelomorphs but segregate in modified form, one on the whole darker, the other on the whole lighter than before they were crossed with each other. This matter of modification on crossing is one deserving further consideration. It is in evidence in all the crosses made. It is clearest where the races crossed differ most in grade, and it seems to consist in a partial obliteration of those differences. Thus the uncrossed white race has its mode at 17, the self race at 0. (See text-figure 2.) In F2 the extracted whites contain no individual as high in grade as 17, and the highest mode lies at 12 to 14, facts which indicate that white has been lowered in grade by its association with self in the Fi zygotes. The back-cross of Fi with white also shows modification but intermediate in amount, as might be expected from the fact that, in the case of each zygote formed, only one of the two conjugating gametes had been subjected previously to modifying influences, viz, that one which was furnished by the Fi parent. The mode in this case lies at 15, instead of at 17, as in the uncrossed race, or at 12 to 14, as in the F2 extracted whites. That the extracted white has less influence in whitening an Fi zygote than the uncrossed white is shown further by a comparison of Fi self X white (text-figure 2) with the back-cross of Fi with self. Fi had its mode at grade 1 and ranged upward to grade 3, but contained no self individuals. If extracted white were identical with uncrossed white in its whitening influence, then in the back-cross with self half the zygotes should be of grade 1 or higher. But in the observed back-cross less than one-third of the zygotes show any white, viz, 29 out of 94, and these are lower in grade than the original Fi's, viz, 1.10 as compared with 1.51. The self character also emerges modified after the cross with white. For the Fi zygotes, consisting of pure self united with pure white (text-figure 2, Fi, S X W) were all close to self in grade, with a mode on grade 1 and ranging upward only to grade 3. But the zygotes formed by extracted self united with pure white produced in the back- cross of Fi with white (text-figure 2, Fi X W) range in grade from 1 to 9 -with a broad low mode at 4 to 7. Evidently they have been much increased in grade in the direction of white. INHERITANCE OF WTIITE-SPOTTING IN RABBITS. 17 Again, in the crosses of dark witli wliite (text-figure 1), we sec tlie mutual modification of the contrasted conditions taking place. Un- crossed dark has its mode on grade 3, and uncrossed white on grade 17. Extracted white as seen in F2 has its mode on grade 15, but in the back- cross with pure w^hite it has its mode on the intermediate grade, 16. As regards the modification of dark, it will be observed that the original Fi individuals are of lower a\erage grade than tlie lower group of individuals produced by the back-cross with white. The mode of the former also is on grade 7, that of the latter is on grade 8. The one consists of pure dark united with pure white, the other of extracted dark united with pure white. The extracted dark has evidently been whitened, exactly as extracted white has been darkened. Uncrossed tan and uncrossed white have their modes on 3 and 17, respectively (text-figure 4). The mode of Fi is on grade 7, but the mode of extracted tan united with white is on grade 10, as seen in the back-cross of F: with white. This shows that extracted tan has been whitened as compared with uncrossed tan. That white has had its grade lowered by the cross with tan is also shown in the back-cross of Fi with white. Its mode lies at 15, not at 17 as in uncrossed white. If, in the several cases considered, crossing tends to mutual modi- fication and assimilation to each other of the contrasted conditions brought together in the cross, why does crossing of dark with tan extend rather than shorten the range of variation, producing in F2 individuals darker than either uncrossed race and others lighter than either uncrossed race? The answer to this question is perhaps to be found in the imperfection of our scale of grades. The scale is a linear one, whereas the variation is not entirely linear; for dark Dutch and tan Dutch have the same modal grade, 3, yet are different in somatic character, as has already been stated. Tan Dutch has a white head and narrow collar, dark Dutch has a dark head and wider collar. In Fi, if pigment simply dominates whiteness, we may expect to get a dark head and a narrow collar simultaneously, i. e., a condition with less white than either parent possessed, which in general is the result observed. Since uncrossed dark Dutch varies down to grade 1 and uncrossed tan down to grade 2, an extension of the dark areas due to crossing naturally carries the pigmentation in Fi down to grade (self) in a certain percentage of cases (1 in 6 observed). See text-figure 5. In F2 the percentage of selfs is still larger, being about 1 in 3. There are also found in F2 whiter individuals than either uncrossed race contained. How these have arisen is indicated by the back-crosses of Fi with dark and with tan respectively. The back-cross with dark (text -figure 5, Fi X D) produces a monomodal group closely resembling the Fi group, but with slightly higher range. This shows that the potent factor in lowering the Fi range was the dark gamete, since it is the only common factor entering into both crosses. On the other 18 HEREDITY IN RABBITS, RATS, AND MICE. hand, the back-cross with tan (Fi X T) shows bimodal variation with the modes at and 5 respectivelj^ This indicates that the gametes formed by Fi are really of two types, extracted dark and extracted tan. The former uniting with pure tan produces a group like Fi but apparently of even slightly lower grade, since the modal condition is now 0, not 1. The extracted tan gametes uniting with pure tan pro- duce a group like pure tan, but of apparently higher grade, since the mode is now on 5, not 3 as in pure tan. Hence extracted tan is poten- tially of higher grade than pure tan, a conclusion supported by F2 from dark crossed with tan, for here we observe that extracted tan meeting extracted tan produces zygotes of grade 5 to 8 or even higher, whereas pure tan does not exceed grade 5. Yet, to return to the imper- fection of our grading scale, these higher grades consist merely in combining a wider collar with the same form of head markings as are found in the tan series of grade 4 or 5. Hence it appears that tan is regularly modified through its contact with dark in an Fi zygote in the way of acquiring a wider collar, whereas dark is modified by the same agency in the waj^ of acquiring a narrower collar. Yet there is no indication that head marking and collar marking are due to dis- tinct single genetic factors, but merely that they are qualitatively different. This difference tends to disappear through mutual influ- ence in the heterozygous condition, but the difference disappears more rapidly in collar markings than in head markings; hence the extended range of grades in F2. Because their collars become more alike the extracted darks rank lower in grade and the extracted tans rank higher; for it will be recalled that the uncrossed darks, though having wider collars than the tans, were graded lower on account of their dark heads. The collar changes, then, are actually blending in this cross, as in all the others studied, but give the appearance of segrega- tion with differences emphasized, merely because of the inadequacy of our linear grading scale to record simultaneously changes in head and in collar markings when these occur with unequal rapidity. It has already been shown that we have conclusive evidence that Fi, from the cross of tan with dark (text-figure 5), produces two types of gametes, not four types, this evidence being (a) the bimodal variation seen in the back-cross of Fi with tan and (6) the bimodal variation seen in the cross of Fi with white. These results indicate that tan and dark are to be regarded as allelomorphic but mutually modifying conditions, as had already been found to be true for tan and white, for dark and w^hite, and for self and w^hite. We have, then, a condition of multiple allelomorphs in white-spotting patterns of Dutch-marked rabbits, which includes the forms self, dark Dutch, tan Dutch, white Dutch, and possibly many other types or conditions of white-spotting which with suflSciently accurate observation might be distinguished from each other. INHERITANCE OF WHITE-SPOTTING IN RABBITS. 19 The foregoing observations show unniistakal)ly that tlie several members of this allelomorphic series tend, as a result of crosses, to become more Hke each other. This has been described as mutual modification, but it should be expressly stated that in the light of our experiments with rats ''modification" need not be regarded a.s change in the nature of a single gene, but merely as equalization of the residual heredity additional to the single genes which produce mono- hybrid ratios. ENGLISH. In November 1909 there were received at the Bussey Institution four "English" rabbits, 1 male and 3 females, bred by R. W. Wills, of Hornerstown, New Jersey. In terms of the grading scale shown in plate 3, the male was of grade 2^; the females were of grades 2, 2f , and 3, respectively. In matings of the male with each of the 3 females, there were pro- duced both English and self-colored young, as shown in table 30; of the former, 21 ; of the latter, 8. The self young were later found to produce no Enghsh young when bred inter se. Hence it seems clear that English is a dominant jNIendelian character, that self is recessive in relation to it, and that the 4 English parents were all heterozygous dominants. The question now arose whether homozygous English rabbits could be produced and why English rabbits were not regularly bred in homozygous form. We did not have long to wait for an answer to these questions. Table 30 shows that the English young of our 4 original English rabbits fall into two groups quite different in appear- ance. Of the 15 3"oung which were graded, 5 were of grade 1 or 1^, while 10 were similar to the parents in grade, varying from grade 2 to 3. The group of low-grade English was found to consist of homo- zygous individuals which produced only English young in crosses with each other or with selfs. The higher-grade group, twice as num- erous in individuals, was found to consist of heterozygotes. These are preferred by the fancier because of their much more striking color- pattern. The homozygote is in appearance only an impure white animal, but the heterozygote is beautifully mottled. It is therefore clear why the fancier breeds heterozygotes. (See plate 3.) Our original English buck, 2545, was also mated with self-colored does of several different sorts, \az, gray, cream, yellow, sooty yellow (tortoise), black, and black-and-tan. In regard to color inheritance, these matings gave us such results as are already familiar through earlier publications by Punnett, Hurst, and ourselves. We may therefore confine our attention to the behavior of the English pattern in crosses. Table 30, b, shows the results obtained; 20 English and IS self young were recorded from these matings. No grade was recorded for 17 of the English young. The others varied in grade 20 HEREDITY IN RABBITS, RATS, AND MICE. from U to 4, the mean being 2.80. Evidently the heterozygous English" produced by these matings with unrelated does were much more variable than those produced by matings of the original English individuals with each other. It now occurred to us to see to what extent this variabiUty could be carried farther in a plus direction by selection. Accordingly we chose the grade 4 individual produced by 9 1492 as the starting-point of the experiment. This individual had been recorded as d'2711. He constitutes generation 1 of the selection experiment now to be described, all animals produced in those experiments deriving their English from him. He was mated with a black-and-tan doe, with three black does, and six Himalayan (albino) does, all free from racial white spotting (EngHsh or Dutch). These matings produced 56 young (table 31, b) equally divided between heterozygous English and self. The English young were of higher grade than the English young produced by 0^2545. (Compare tables 30 and 31.) They ranged from grade 3 to grade 5, mean 3.89, as compared with a mean of 2.80 for the young of cf 2545, produced in similar matings. Male 2711 was later mated to 5 of his heterozygous English daugh- ters produced in the matings already described, and also to one of the resulting grand-daughters. The character of their young is shown in table 31, a. The mothers form generations 1^ and 2^ of the selected English race. They vary in grade from 3|- to 5. They produced three classes of young: low-grade English (homozygotes), high-grade English (heterozygotes), and selfs. Their respective numbers were 12, 11, and 8. The mean of the low-grade English group was 1.52, that of the high-grade group was 3.93, which agrees closely with the grade of the same group produced by matings with non-English (self) does (table 31, b). Grouping the mothers by grade, the relation shown herewith is observed between grade of mother and grade of young. This indicates that selection of higher-grade parents would probably result in produc- ing higher grade young. A son of 0^2711, viz, cr5086, generation 1^, grade 4|-, was chosen to succeed him in the selection experiment. This buck was mated with 10 different heterozygous English does, 5 of which had also been mated with his father. (See table 32.) There resulted 20 homozygous English young, 44 heterozygotes, and 35 selfs. The mean grade of the homozygotes was 1.38, that of the heterozygotes was 3.96, averages not very different from those which characterized the young of d^2711 (table 31). Consequently no advance can be claimed as a result of the selection of cf5086. He was mated also with 3 black-and-tan does (table 32, b), producing thus 10 English j^oung of mean grade 3.15, a lower average than that given by cf 2711 in matings with self Mother. Mean of young. 3.50 4.75 5.00 3. S3 3.80 4.25 INHERITANCE OF WHITE-SPOTTING IN RABHITS. 21 does, but it must be borne in niind that the mothers were nf)t all identical in the two cases. Black-and-tan does seemed in general to give lower-grade oftsjiring than does of the self black and Himalayan races, with which cf 2711 had been mated. We next used as sire in the selection experiment cfoSTo, generation 2^, grade 4|, a son of cf 5086 by his half sister. (8ee table 33.) He was mated with 9 different heterozygous English does, all but one of which had also been mated with his father. They produced English young of somewhat lower mean grade than they had borne bv the father, 6^5086. (See tables 32 and 33.) 17 homozygous English young were of mean grade 1 .20; 29 heterozygous English young were of mean grade 3.79; there were also 25 self young. Again the higher- grade mothers produced the higher-grade young. Hence there wa.s evidently material favorable for selection among the mothers, if not among the fathers. This male was now discarded and replaced by an own brother of slightly higher grade, viz, cf 5555, generation 2^, grade 4f . (See table 34.) This male (cf5555) was bred more extensively than any of his predecessors and produced higher-grade offspring. He was mated to the same does as his father and grandfather and also to a number of new ones which now^ became available. By all classes of does he pro- duced higher-grade 3'oung than had an}- of his predecessors. He also produced higher-grade young by his mates of higher grade than by his lower-grade mates. (Table 34.) In his case, then, a second advance had been made by selection in the male line and the necessary variation was evidently' present to make possible similar advances by selection in the female line. JNIale 5555 produced 41 homozygous English young, 125 heterozj-gous English, 2 ungraded English, and 65 selfs, or all together 168 English and 65 selfs. The next sire tested was a son of cf5555; viz, 6^6370, grade 5, generation 3, as regards selected ancestry. (See table 35.) He was more advanced in grade and generations than any male thus far tested and produced higher-grade young by females of the same grade. Many of the older females had now been discarded, but enough remained to form a standard of comi:)arison between the genetic properties of this male and those of his predecessors. The mean grade of the heterozygous English 3'oung of this male was 4.66; the grade of his homozygous English young was 1.79. The corre- sponding figures for his father were 4.40 and 1.36 respectively. Three other sons of cf 5555 were also tested by matings with sub- stantially the same group of does, although tests in the case of the less promising ones were terminated sooner. Male 6420 (table 36) was of slightly lower grade than his father and was found to be probably inferior to him and so was soon discarded. His heterozygous English young were of mean grade 4.33. 22 HEREDITY IN RABBITS, RATS, AND MICE. j\Iale 6071 (table 37), although of higher grade, gave no better results. His own brother, cf6072 (table 38), born in the same litter and graded the same, did much better. He was bred very extensively and gave a record very similar to that of his father, 6^6370, who was of the same grade but had half a generation less of selected ancestry. IVIale 6072 had 75 homozygous Enghsh young, of mean grade 1.97 (father's record 1.79); he also had 159 heterozygous English young of mean grade 4.63 (father's record, 4.66). Increase in the grade of the young with increase in the mother's grade is very clearly shown among the young of this sire. (See table 38.) The next male tested was 6964 (table 39) a son of cf 6071. He was discarded after a set of matings which showed him probably not better than his uncle, 6072, who w^as still in service. He had 21 homo- zygous j^oung of mean grade 1.49 (his uncle's record was 1.97) and 29 heterozygous English young of mean grade 4.68 (his uncle's record being 4.63). Next was tested cr7699 (table 40), son of 6^6072, who shared with his half brother, cf 6370, the position of best sire so far. All were of the same grade, 5. This male was mated with all available does and produced 354 recorded young. He has a better record than any sire so far tested. By heterozygous does he has sired 75 homozy- gous young of mean grade 2.31 and 149 heterozygous young of mean grade 4.80. Another male of the same grade and generation as the foregoing, indeed his half-brother, being also a son of 6072, was tested, but appeared not to be superior to 7699 and so was soon discarded. This animal, 9532 (table 41), sired 16 homozygous English young of mean grade 2.53 and also 27 heterozygous English young of mean grade 4.73. Three sons of the superior male, 7699, have since been tested, viz, 9806, 1212, and 534 (tables 42-44). The first one shows no probable superiority over his father, but the last two are more promising, each having produced a total of over 60 heterozygous English young with a mean close to grade 5. In the case of their father the corresponding group of young were of grade 4.80. The homozygous young produced by their father were of mean grade 2.31; those produced by the sons were of mean grade 2.87 and 2.95 respectively. Accordingly, as regards both heterozygous and homozygous young, the sons have a distinctly better record. This, no doubt, was due in part to the fact that their mates were of higher grade or from more highly selected stock, but it was not wholly due to this cause, for their half-brother (9806, table 42) did not show the superiority which they showed, even when mated with females of high grade and advanced generations. Hence we must conclude that these two males, 1212 (table 43) and 534 (table 44), were genetically superior to their father. Table 45 shows the grade distribution of the young produced by a homozygous English male, 1173 (plate 3, fig. 6), when mated with does INHERITANCE OF WIIITE-SPOTTINC IX RABBITS. 23 of the three categories used in testing heterozygous Enghsh males. It will be noted that he produced only English young, however mated, conclusive evidence of his homozygous dominant character. His hetero- zygous English young were of mean grade 4.77 and 4.84 by homozygous English and self does respectively. This male was a son of male 9532, table 41, with whose genetic character his own was very similar, judg- ing by the grade of their heterozygous English young. His line was not continued. Table 46 enables one to survey at a glance the summarized results of this entire selection experiment. The course of the experiment is followed only in the male line, because only in the case of the males is the number of young large enough to show beyond question the genetic properties of the individual. From the beginning of the experi- ment fluctuation was observed in the grade of the 3'oung produced, and this fluctuation was in part at least genotypic, since the higher-grade mothers have given higher-grade young in matin gs with the same male. That the fluctuation was also in part lihenotypic is shown by a comparison of the records made by different males of the same grade when mated with the same group of females (tables 31-45). The entire selection race derives its English character from cf2711. This animal was a heterozygote deriving the English character in a single gamete from cf2545, who was also heterozygous. Hence the English character had evidently changed in transmission from father to son, a sufficient refutation of the idea of unit-character constancy. AVhether the change resulted from a directly changed unit-factor (gene) or from the introduction of one or more modifying factors is a matter for further consideration. The advances made in the male line seem to occur as five successive steps corresponding roughly with generations of selected ancestry (table 46). The first advance comes with the selection of the (single gamete) male 2711, founder of the race; the next in the selection of his grandson, 5555 ; the third in two sons of 5555, viz, 6370 and 6072 ; the fourth occurs in the selection of two sons of 6072, viz, 7699 and 9532; the fifth is seen in 1212 and 534, sons of 7699. The direct line of advance is through 2711, 5555, 0072, and 7699. The amount of advance at each step, as indicated by the average grade of the young of these males, is shown in table 47. The rate of advance has evidently decreased as the experiment progressed. That modification of the English pattern resulted imme'ever, which was challenged by those who maintained that the gametes v^^ere "pure." Subsequent investigation has shown beyond question not only that unit-characters are frequently greatly modified by crosses, but also that they can be modified by selection alone unattended by crossing. Those who formerlj^ maintained the doctrine of gametic purity now shifted their ground, and while admitting that unit-characters might change, insisted that single factors or "genes" could not change. This is the doctrine of pure genes which Morgan has made so familiar. This doctrine it is difficult either to prove or disprove. Pragmatically speaking, it is of small consequence, since it is admitted (1) that single Library N. C, State College INHERITANCE OF WHITE-SPOTTING IN RABBITS. 25 factors do sometimes change, leading to the formation of nmhiple allelomorphs; (2) that the action of single factors is not limitetl to any particular part of the organism, but may affect parts apparently unrelated; (3) that the total number of factors concerned in the genesis of even the simplest organisms must be very great ; and (4) that in what should theoretically be "pure lines" (asexually reproducing organisms, Jennings) genetic changes are constantly occurring. Trag- matically, then, genetic variability by minute gradations is a reality, precisely as Darwin assumed it to be, and this fact allows races to be altered steadily and permanently by selection, either natural or artificial, as Darwin also assumed was the case. The hypothesis that stable organic forms come into being onl}' suddenly, by aljrupt changes from preexisting forms and not by gradual modification — this hypothesis, the "mutation theory" as commonly understood, is not substantiated. RELATION OF DUTCH TO ENGLISH. It remains to consider the genetic relations to each other of Dutch and English spotting. Dutch, as we have seen, behaves as a reces- sive in crosses with self-pigmented races; English, on the other hand, behaves as a dominant. Dutch marking appears to result from a simple deficiency of pigmentation, as if in development the pigment supply failed at an extremity or at an embryological point of finishing- off. When Dutch marking is reduced to its lowest point of expression by selection or crossing, the only white visible is found at the tip of the nose, or on the toes of a fore-foot, or as a spot in the middle of the forehead. English spotting, on the contrary, appears to result from some positive inhibiting force, some agency which uses up the pigment-forming materials here and there in the epidermis and con- verts them into an end-product not colored but white. That English individuals possess all the agencies necessary for full pigment formation is shown by the fact that English parents may produce fully pigmented (self) young as recessives, which then produce only self young if mated with each other. We have seen that there occur different forms of Dutch spotting, which apparentlj^ behave as allelomorphs, but which tend to become less distinct, one from another, when they are associated in the same zygote. It would seem probable that they represent quantitati\ely different stages of reduction in the amount of some substance carried in the germ-cell. But undoubtedly this substance, whatever it is. is located in the chromatin, since the defect is transmitted e(iually through egg and sperm. There are also cjualitative differences among the different forms of Dutch, as for example between 'Mark" and "tan" Dutch, in one of which the white collar is more in evidence, while in the other it is the head markings that are more in evidence. Probably, then, the different forms of Dutch are variants of a single 26 HEREDITY IN RABBITS, RATS, AND MICE. Dutch young. English young. Grade. No. Grade. No. 1 2 3 3 2 3 1 If 2 2i 2\ 2f 3 2 15 1 2 1 5 Total 1 9 20 ' ' locus," in the terminology of the chromosome theory. But the physi- ological and genetic behavior of English are so difTerent that it would seem improbable that they are variants of the same gene. Nevertheless, when English is crossed wdth Dutch, the two appear to be either allelomorphs or closely linked, as the following results show : Heterozygous English rabbits of grade 5 were crossed with ''white" Dutch of grades 15 to 17. Two matings were also made of the homozygous EngHsh buck, 1173 (table 33, and plate 3, fig. 6), whose English character was of equivalent potency with that of heterozygotes of grade 5. These matings produced 26 Enghsh and 9 non-English (Dutch) young, which were graded (with no great exactness) as follows : There was probably no real discontinuity in the grouping of the English young, but owing to the rough manner of grading them the numbers heap up on the even grades, fractional grades being neglected. The Dutch young are similar in grade to the young pro- duced by crossing "white" Dutch with self . (Compare table 27.) But the English young are much whiter than heterozygotes between English and self. The latter are 4.75 to 5.00 in grade when produced by the same English sires. (Compare table 45.) But the Enghsh-Dutch heterozygotes in no case are of higher grade than grade 3 and in the great majority of cases are no darker than grade 2. In other words, thej^ are of about the same grade as English homozygotes. This means that a white Dutch gamete has about the same whitening effect on English as another gamete of Eng- lish would have. The pattern of the English-Dutch heterozygote is indistinguishable either quantitatively or qualitatively from that of a homozygous English animal. The Dutch is not at all in evidence except as a whitening influence on the dominant English. Even in a single dose it completely counteracts the darkening influence introduced into the English race in the process of the cross with self and the subse- quent 5 generations of selection. Since it has been shown that both white Dutch and English segre- gate from self in monohybrid fashion, much interest attaches to the inquiry whether they segregate from each other in the gametes of the Fi individuals. Two methods have been employed to test this matter, one being to mate Fi individuals inter se, the other to back- cross them with white Dutch. If English and Dutch are allelomorphs (invariably pass into different gametes) nothing but English or Dutch young should be produced by either mating. If, however, English and Dutch are not allelomorphs, then a certain number of gametes should be formed by Fi individuals which are neither English nor Dutch, but which are self, and a like number should be formed which INHERITANCE OF WHITE-SPOTTING IN RABBITS. 27 carry both English and Dutch, the genetic properties of wliich woukl be to produce a very white (low-grade) English. These two new kinds of gametes would respectively be as numerous as the simjile iMiglish and simple Dutch gametes, if English and Dutch are inherited inde- pendently (in different chromosomes, for example). If English and Dutch are linked in inheritance (are borne in homologous chromosomes but not at the same locus), these two classes of gametes would represent the "cross-overs." The existence of the gamete which bears both English and Dutch would be difficult to demonstrate, since one char- acter is dominant, the other recessive; but the existence of gametes bearing neither English nor Dutch would be easy to detect, either in the straight F2 generation or in the back-cross generation. Such gametes uniting with each other would produce self individuals, or uniting with a Dutch gamete would produce a Dutch of grade 3 or lower. Now, among 47 F2 young neither of these classes of individ- uals has appeared. The 11 non-English Fo young are all Dutch of grade 7 or higher. An even better test for the existence of gametes transmitting neither English nor Dutch is the back-cross with Dutch; for in this case any such gamete would produce one and the same kind of zygote, viz, Dutch of grade 3 or lower. No such zygote has appeared in a total of 88 Dutch and 105 English young obtained in back-cross matings. This indicates that Enghsh and Dutch are either allelo- morphs or closely linked. The grade distribution of these back-cross young is showTi m the table herewith. It should be of interest to compare this distribution with that obtained in the orig- inal English-Dutch cross (page 26) as indi- cating whether the contrasted characters have been contaminated in the Fi zygote. The English back-cross young are of lower grade (whiter) than the Fi Enghsh. The respective means are 1.36 and 2.25. This whitening of the English is the result of contamination from white Dutch in the Fi zj^gote. No English individual produced in the back-cross is darker than the English produced in Fi. In both cases the darkest Enghsh are of grade 3. This indicates ab- ' sence of gametes transmitting neither English nor Dutch, for any such gametes uniting with an English gamete should prothice lOnglish darker than gxade 3. The Fi Dutch (page 26) were self-white Dutch heterozygotes. (Compare table 27.) None was darker than grade 3. In this back-cross, any self gamete (neither Dutch nor English) which might arise by cross-over should produce young ecjually dark, grade 3 or lower. But the lowest-grade Dutch recorded are of grade 0; Dutch young. English young. Grade. No. Grade. No. 6 2 1 52 7 1 2 25 8 1 3 2 9 4 26 10 6 1 11 4 12 13 13 1 14 15 15 14 16 22 17 3 ? 2 Total 88 105 28 HEREDITY IN RABBITS, RATS, AND MICE. hence there is no evidence that cross-over gametes are produced as often as once in 192 times. It, therefore appears that Enghsh and Dutch are either very closely coupled or are variants of the same locus. As regards the effect of the English cross on the grade of the Dutch character, this is indicated in the grade of the back-cross Dutch young, which range in grade from 6 to 17, average 13.88. Similarly produced back-cross Dutch obtained from Fi (white X self) X white (table 17) range from 13 to 17 and are of mean grade 15.15. This indicates that the English cross has darkened white Dutch even more than self did in a similar cross (table 27). The superior darkening effects of English over the self used in table 27 may be attributed to the more highly selected character (for darkness) of the English used in the cross and to the less highly selected character (for whiteness) of the Dutch used in the same cross, but the difference is not great. At any rate, from the grade distribution of the young produced in the back-cross of Fi (English X white Dutch) X white Dutch, it is clear (1) that English and Dutch behave like allelomorphs, or closely linked factors, since in nearly 200 cases studied no cross-over is observed, i.e., no gamete transmitting neither English nor Dutch; (2) that the segregated English and Dutch borne by the gametes of Fi individuals are mutually modified, the English (previously selected for darkness) being made lighter, and the Dutch (previously selected for whiteness) being made darker. In these mutual modifications we are dealing probably for the most part with residual heredity, but it is possible that quantitative variation of the English and the Dutch genes is in part responsible, yet such an interpretation is not favored by the results obtained from the critical experiments with hooded rats, which point strongly to changed residual heredity as the correct explanation of changed phenotypes, when only a single Aiendelizing character can be observed. III. OBSERVATIONS ON THE OCCURRENCE OF LINKAGE IN RATS AND MICE. In publication 241 of the Carnegie Institution evidence was pre- sented showing that the red-eyed yellow and pink-eyed yellow varia- tions of the common rat {Mus norvegicus) are due to genes which are linked with each other. Upon crossing with each other the two yellow variations, which visibly differ in eye-color only, young are obtained which differ from both parent races in coat-color as well as in eye- color. These young are black-coated or gray-coated and have black eyes. This result shows clearly that the two variations, which are both recessive in genetic behavior, are due to independent genes. In the F2 generation the two yellow varieties were recovered, each wnth its distinctive eye-color, and certain individuals, which visibly were pink-eyed yellows, were found from breeding tests to carry the genes for red-ej'ed as well as for pink-ej'ed yellow. These double recessives obviously had arisen by the process known as "crossing- over," in which genes, although introduced in a cross by different parents, yet later emerge together in the same gamete formed by an FjL individual. It is supposed that genes which behave in this way lie in homologous chromosomes and that when crossing-over occurs a gene (A) leaves the chromosome in which it originally lay and crosses over into the homologous chromosome in which the other gene (B) lay. Thus both A and B come to lie in the same chromosome and at gametogenesis pass into the same gamete. From an examina- tion of the proportion of the double recessive yellow individuals found among the F2 yellows, it w^as concluded that cross-over gametes (those which carry genes for both yellow variations or for neither) represent about 18.5 per cent of all the gametes formed b}' Fi individ- uals. If no linkage occurred, such gametes would form 50 per cent of the total. To test more fully the strength of the linkage between these two genes and to find out whether this linkage has the same strength in both sexes, further experiments have been undertaken. A race of homozygous double recessives (genetically both pink-eyed and red- eyed) was built up from the F2 cross-over individuals and with this race Fi individuals were crossed. If we designate by r the gene for red-eyed yellow and by p the gene for pink-eyed yellow, an Fi individual might be expected to form gametes of the four sorts l^R, Pr, pR and pr. Of these 4 combinations, Pr and pR would corrcsjiond with those furnished by the parent races, rcd-eyetl yellow and pink-eyed yellow respectively; but the other two, PR and pr, would be new and 29 30 HEREDITY IN RABBITS, RATS, AND MICE. could therefore arise only by crossing-over. A test mating of an Fi individual with a double recessive would give zygotes as follows, it being understood that the double recessive individual produces only one type of gamete, viz, pr. Non-cross-over gametes, Pr and pR, would give zygotes Pprr and ppRr, visibly red-eyed yellows and pink-eyed yellows respectively. Cross-over gametes, PR and pr, would give zygotes PpRr and pprr, visibly dark-eyed (black or gray) and pink-eyed yellow respectively. The pink-eyed yellows could not be distinguished readily from yellows arising from non-cross-over gametes, but the dark-eyed young could be distinguished immediately at birth from all other classes. Since theoretically they would con- stitute half the total cross-overs, it is evident that the simplest way of estimating with accuracy the proportion of cross-over gametes is to double the number of dark-eyed young observed in new-born litters. This number divided by the total number of young would give the percentage of cross-over gametes. Following this procedure, we have reared from matings of Fi indi- viduals with double recessives a total of 1,714 young, of which 174 were dark-eyed. Doubling the number 174, we have 348 as the prob- able number of cross-over gametes among the 1,714 Fi gametes which entered into the production of these young. This is a percentage of 20.3, a little higher than the calculation 18.5 of publication 241, based on a study of a much smaller F2 population. The difference between this figure, 20.3 and 50, the percentage expected where no linkage occurs, would be a measure of the strength of the repulsion shown between the genes for red-eyed yellow and for pink-eyed yellow respectively, when they enter a cross in different gametes — that is, each through a different parent, the condition realized in this cross. But, on the chromosome theory, an attraction or "coupling" equal in strength to this repulsion should occur between the same two genes when they enter a cross together. Entering together, they should tend to hold together, because they would lie in the same mem- ber of a pair of chromosomes and so could pass out separately only in consequence of a cross-over. This point, repeatedly verified in the case of other organisms, was tested for rats by producing Fi individuals through a cross of double recessive yellow (pprr) with a pure non- yellow individual (PPRR). In reality Fi zygotes of this same sort were being produced in considerable numbers in the matings already described to test the strength of repulsion. Such were the 174 dark- eyed young already mentioned. Each resulted from the union of a pr gamete with a PR gamete, the relationship which would give the expected coupling. Accordingly many of these dark-eyed young were used instead of Fi parents in the experiments to test the strength of "coupling" between p and r. In these experiments, as in those to test the strength of repulsion, Fi individuals were mated with double OCCURRENCE OF LINKAGE IN RATS AND MICE. 31 recessives. In both cases the Fi individual was of the formula PpRr, the double recessive was of the formula pprr. The only difference in the two cases was that in one case the Fi arose from a union of Pr with pR, and in the other from a union of pr with PR. But the importance of this circumstance is seen in the different results obtained in the two cases. In one case (where p and r enter the cross separately) 10 per cent of the young were dark, in the other case (where p and r enter the cross together) more than 40 per cent of the young were dark. The exact figures for the repulsion series have already been given, 174 dark young in a total of 1.714, or 10.1 per cent dark young. For the coupling series the figures are 1,255 dark young in a total of 3,032, or 41.3 per cent dark young. To compare the strength of repulsion with the strength of coupling we may estimate the percentage of cross-over gametes produced in each case. Either sort of Fi individual would produce 4 kinds of gametes, PR, Pr, pR, and pr. But in the repulsion series PR and pr would arise from crossing-over, whereas in the coupling series Pr and pR would arise from crossing-over. In either case dark-eyed individuals would arise only from the same type of Fi gamete, viz, PR, but in the repul- sion series this would be a cross-over gamete, whereas in the coupling series it would be a non-cross-over gamete. While in the repulsion series the number of dark-eyed young would measure half the total number of cross-overs, in the coupling series it would measure half the total number of non-cross-overs. Applying these criteria, we have found in the repulsion series, as already stated, that the number of dark-eyed young being 174, the probable number of cross-over gametes is twice this, or 348, in a total of 1,714, which is 20.3 per cent. Turning now to the coupling series, we find that the total number of dark-eyed young is 1,255. Doubling this we have 2,510 as the probable number of non-cross-over gametes. Deducting this number from 3,032, the total number of young, we have 522 as the probable number of cross-over gametes, which is 17.2 per cent. This we may compare with the 20.3 estimated for the repulsion series, and the earlier esti- mate of 18.5 based on the census of an Fo population (publication 241). These differences are not large enough to lead us to think that there is any consistent difference between the strength of repulsion and the strength of coupling between the same two genes. The chromosome theory would not lead us to expect the existence of any such difference. This case therefore fully accords with that theory. If we combine the results obtained from both the repulsion and the coupling series we have as the average linkage strength (either repulsion or coupling, as the case may be) 18.3 per cent. This is based on a total of 4,740 young produced by the back-cross of Fi with the double recessive. The figures are large enough to have significance and agree remarkably well (almost too well) with the estimate based on the F; population, 32 HEREDITY IN RABBITS, RATS, AND MICE. viz, 18.5 per cent. It is safe to conclude that the linkage strength of red-eyed yellow with pink-eyed yellow is close to 18 per cent. We may pass now to the question whether the linkage strength is the same in spermatogenesis as in oogenesis, whether it is the same among the gametes formed by Fi males as in those formed by Fi females. A priori we might well expect it to be different in the two cases, since in Drosophila crossing-over has been found to occur only in females, whereas in the silkworm it has been found to occur only in males. In publication 241 the fact was demonstrated that crossing- over does occur in both sexes of the rat, but we were at that time unable to state what its relative frequency was in the two sexes. Our back-cross series of matings give data for such a determination. (See table 48.) It will be observed that the estimated percentage of cross- over gametes is somewhat higher for females than for males in both the repulsion and the coupling series and that the difference is greatest where the numbers are largest, viz, in the coupling series. This would suggest that crossing over occurs more readily in oogenesis than in spermatogenesis, but I doubt very much whether such is the case when all other conditions are the same. Summaries made for the concluding period of our experimental work, when conditions had been more carefully controlled and the procedure of taking the records had been best standardized, show no appreciable differences in the case of the two sexes. For this period, in the coupling series, Fi females gave 187 dark and 277 yellow young, or 19.3 per cent cross-over gametes. Simultaneously, Fi males of similar parentage gave 133 dark and 197 yellow young, or 19.4 per cent cross-over gametes. In the repulsion series only Fi males were at this time being used to any great extent. They produced 29 dark young and 269 yellow young, which by the method of calculation already explained indi- cates 19.4 per cent cross-over gametes, a remarkably close agreement with the results given by both sexes in the coupling series at this same period. Whether external conditions have any influence on the percentage of cross-overs we are unable to state, but this seems doubtful in the case of a warm-blooded animal such as the rat. That individual or age differences may occur among Fi animals affecting the percentage of cross-overs is a possibility we have considered carefully, but with only negative conclusions. The indicated percentage of cross-overs varies in the case of particular Fi males from to 44 per cent, but this variation appears to be the result of random sampling rather than of consistent differences in genetic behavior. Several males showing extremely high or extremely low percentage of cross-overs were trans- ferred to new breeding-pens and mated with other double recessive females. Their indicated percentages of cross-over gametes before and after the transfer showed no consistency with each other, and OCCURRENCE OF LINKAGE IX HATS AND MICE. 33 SO we are forced to conclude that individual differences as regards the production of many or few cross-overs have not l)cen shown to exist. In publication 241 evidence was presented indicating that all)inism in rats is probably due to a gene which is linked with the genes for red- eyed yellow and pink-eyed yellow. This idea is now fully established and we are able to give provisional estimates of the linkage strengths involved, although the investigation of this matter is still incomplete. When red-eyed and pink-eyed rats are crossed with each other, or either sort is crossed with an all^ino, the Fi young produced are dark- eyed and dark-coated (either black or gray, according as the agouti factor is absent or present). But the Fi young are not quite as dark in color as wild rats. This shows that all three variations are reces- sive and complementary, but that the allelomorph of each is a little less effective in producing pigment when in heterozygous form than when in homozj-gous form (as in wild rats, or in Irish or in hooded rats). Fi individuals from the cross of albino with pink-eyed yellow, when bred with each other, produce an Fo generation of three apparent types, viz, (1) dark, in eye and coat color; (2) pink-eyed yellow; (3) albino. If no linkage occurred we should expect these three classes to occur in the ratio 9 : 3 : 4; but, as was pointed out in publication 241, linkage would tend to equalize the numbers of pink- eyed and albino j^oung, and such a tendency has been recorded. Further, if no linkage occurs, but if pink-eyed yellow and albinism segregate quite independently of each other, then half the albino gametes formed by Fi individuals should transmit pink-eyed yellow and half should not transmit it; conversely, half the gametes which transmit pink-eyed j-ellow should also transmit albinism and half should not. If less than half the gametes which transmit one character transmit the other, the two show repulsion. To test the matter, 45 F2 albinos have been mated with homozygous pink-eyed individuals. Of the 45 so mated, 17 have produced both pink-eyed and dark-eyed young, one has produced only pink-eyed young, and 27 have produced only dark-eyed young. The 17 are evidently heterozygous for pink-eye as well as homozygous for alliinism, their formula being ccpP. The one which produced only pink-eyed young is probably of the formula ccpp. The 27 which produced only dark-eyed young are of the formula ccPP. We may now consider what was the nature of the gametes which produced these 45 individuals. A gamete which furnished both albinism and pink may be called a cross-over gamete; one which furnished albinism only must be regarded as a non-cross-over gamete. The 27 albinos which did not transmit pink-eye evidently arose each from the union of two non-cross-over gametes. This accounts for 2 X 27 or 54 non-cross-over gametes. The 17 individuals which were heterozygous for pink evidently received each a single non-cross- 34 HEREDITY IN RABBITS, RATS, AND MICE. over gamete. This makes a total of 71 such gametes. Cross-over gametes were represented singly in each of the 17 individuals which were heterozygous for pink and doubly in the one which was homo- zygous for pink. This makes a total of 19 cross-over as against 71 non-cross-over gametes, which is 21.1 per cent cross-overs. This is an indicated linkage strength a little less close than that between pink- eyed and red-eyed yellow, in which case the percentage of cross-over gametes was estimated at 18.3. For with no-linkage gi\'ing 50 per cent cross-overs, it is evident that the linkage strength increases as the percentage of cross-overs decreases until (when cross-overs cease) linkage becomes complete. If we measure the strength of linkage by the difference between the observed percentage of cross-overs and 50 per cent (the percentage of cross-overs when no linkage occurs), then linkage between red-ej^ed yellow and pink-eyed yellow is 31.7 and that between pink-eyed yellow and albinism is 28.9, as provisionally determined. The linkage between red-ej^ed yellow and albinism is much stronger than the linkage in either of the cases just discussed. Tests have been made for the presence of the red-eyed yellow gene in 160 F2 albinos and for the presence of albinism in 57 F2 red-eyed yellows derived from the cross of albino with red-eyed yellow. Only a single cross-over has been detected, and even that is not beyond question. One of the Fo albinos, a male, when mated with a pure red-eyed yellow female, sired a litter of young, all of which were dark-eyed except one. This one proved to be a yellow but died, as did the father, before additional breeding tests could be applied. If this yellow individual was really sired by the albino male (and not accidentally introduced from some other cage, a remote possibility), then that male evidently carried yellow as well as albinism and in his genesis a cross-over gamete must have functioned. Each of the other F2 albinos and the Fo yellows tested manifestly arose from the union of gametes neither of which transmitted both yellow and albinism, since as mated they produced only dark-eyed young (4 or more each). On these assump- tions the experiments thus far show that only one gamete out of 434 formed by Fi parents can have been a cross-over gamete, which apparently gives less than one per cent of cross -overs. The experiments are being continued with the hope of finding addi- tional cross-overs and of thus securing a double recessive race, which will make possible a more accurate determination of the linkage strength. The information already presented shows that on the chromosome theory the genes for albinism and for red-eyed yellow are extremely close to each other in the same chromosome and that the gene for pink-eyed yellow, while lying in this same chromosome, is at some distance from the genes for albinism and red-eyed yellow. OCCURRENCE OF LINKAGE IN RATS AND MICE. 35 In mice it has been shown by Haldane 6/ a/, that the genes for albin- ism and for pink-eye are probably linked with each other. This fact, interesting in itself, is made doubly so by the consideration that characters apparently identical in nature with these two are also linked with each other in rats. Since mouse and rat are species grouped by systematists in a single genus, it should be of interest to compare their genetic constitution as fully as possible. With this idea in mind we had already undertaken to study the linkage relations of all)iiiism and pink-eye in mice before the appearance of the paper by Ilaldane et aU This investigation was undertaken by Mr. L. C. Dunn while acting as my assistant. Upon his entering military service, I took over the experiments. It is a pleasure to acknowledge Mr. Dunn's important part in the work. We began, as in the rat experiments, by crossing pink-eyed with albino individuals. Dark-eyed Fi young were produced exactly as in rats. These bred with each other produced an F2 generation of dark-eyed young, pink-eyed young, and albinos, in a 9 : 3 : 4 ratio manifestly modified by linkage. Pink-eyed F2 individuals were tested for the presence of albinism and albino F2 individuals were tested for the presence of the pink-eye gene as a first step toward the production of a race of double recessives needed to ascertain the proportion of cross-over gametes formed by Fi individuals. The simplest way of making the tests was found to be the mating of F2 albinos with F2 pink-eyed individuals. This afforded simultaneously a test of both parents. For if the pink-eyed parent carried albinism, 50 per cent of the young would be albinos, otherwise none would be albinos. But jf the albino carried the pink-eye gene, 50 per cent of the young pro- duced would be pink-eyed. If both these contingencies were realized in the mating, 25 per cent of the young would be pink-eyed and 25 per cent albinos. All other young, as in a cross of pure pink-eyed with pure albinos, would be dark-ej^ed. If no linkage occurred between pink-ej'e and albinism, it would be expected that half the F^ pink-eyed individuals would carry albinism, and also that half the albinos would carry pink-eye. Any smaller proportions than these of pink-eyed carrying albinism or of albinos carrying pink-eye, among the F2 individuals, would indicate linkage. Linkage is very clearly shown by the tests made. Among 03 F? pink-eyed which were tested, 18 produced each one or more albino young in litters otherwise dark-eyed, while 45 produced no albinos but only dark-eyed young. In the genesis of the 18 parents men- tioned, it is evident that 18 cross-over gametes had united with 18 ^ In fact, I have not yet had access to the paper by Haldane et al. but know it only as cited by others. Our copy of the journal in which it appeared is probably at the bottom of the ocean and we have been unable as yet to replace it. 36 HEREDITY IN RABBITS, RATS, AND MICE. non-cross-overs. But in the production of the 45, only non-cross-over gametes had functioned. The total gametes involved accordingly are 18 cross-overs and 18 + (2 X 45) = 108 non-cross-over gametes; total 126. As 18 is 14.28 per cent of 126, the indicated percentage of cross-overs is 14.28 per cent. Among 75 F2 albinos which were tested, 20 produced pink-eyed young (as well as dark-eyed ones), while the remaining 55 produced only dark-eyed young. Reasoning as before, there were evidently in\'olved in this case 20 cross-over gametes and 20 + (2 X 55) = 130 non-cross-overs, total 150. But 20 is 13^ per cent of 150; hence the indicated percentage of cross-overs is 13^. Combining the tests of F2 pink-eyed and of F2 albinos, we have in tests involving 276 Fi gametes an indicated percentage of 13.76 cross-overs. The pink-eyed parents, which in the course of these tests had been found to carry albinism, were now mated with each other, and the albino young which they produced when so mated w^ere used in building up a race of double recessives, for all albinos so produced must of necessity be homozygous for pink-eye. These double reces- sives were next mated with Fi dark-eyed animals obtained by the original cross of pink-eyed with albino, or with dark-eyed individuals of similar genetic constitution which had resulted from the test matings. The interpretation of the results obtained from these back- cross matings is the same as that given by similar matings in the case of rats. The Fi parent would form gametes of the four sorts CP, cP, Cp, and cp, of which cP and Cp would represent the original com- binations found in pure albinos and pure pink-eyed respectively, and so would be non-cross-overs, but CP and cp would arise only by crossing- over. Of the four types of gamete, CP alone would produce a dark- eyed zygote, if mated with a double recessive, cp. But this is one of the two cross-over types. Hence the number of dark-eyed young produced in mating Fi animals with double recessives should indicate half the total percentage of cross-overs. By matings of the sort just described, 3,142 young have been produced, of which 222 were dark- eyed. Doubling this number, we have 444 as the probable number of cross-over gametes in 3,142 gametes produced by the Fi parents, an indicated percentage of 14.13. This agrees very well indeed with the 13.76 per cent indicated by the test-matin gs of F2 pink-eyed and albinos. It seems safe to assume, therefore, that the average cross- over percentage is close to 14 per cent. For the corresponding char- acters in rats, the indicated percentage of cross-overs is considerably higher, viz, 21.1, but it should be borne in mind that the estimate is based on a much smaller series of observations in the case of rats and that further observations may alter it materially. TABLES. 37 TABLES. Table 1. — Classification of generation 17, plus-sclcclion series. Grade of parents. Grade of ofTspring. Totals. Means. 31 4 4i 4^ H 5 51 5i 5J 2 4t 4^ 4j H 4i 5 51 5i 5| 2 1 1 5 1 9 3 1 2 2 1 4 7 12 8 5 2 3 1 3 5 11 7 26 10 10 13 8 2 3 4 7 11 17 13 12 6 1 1 1 4 6 13 5 10 2 2 1 4 2 8 1 10 5 1 1 1 1 4 6 3 4 3 1 2 5 3 5 9 12 31 38 88 43 65 39 16 5 3.95 4.22 4.44 4.47 4.49 4.46 4 48 4.56 4.61 4.30 4.55 4.81 23 46 96 75 44 32 22 11 o 351 4.48 Table 2. — Classification of generation IS, plus-selection scries. Grade of parents. Grade of offspring. Totals. Means. 31 3i 4 4i 4^ 41 5 5i 5^ 4i 4^ 4^ 4i" 4f 4|- 5 51 5i 5i 3 4 5 2 5 7 o 1 5 9 3 10 3 8 11 4 2 3 9 8 22 13 14 25 11 7 2 1 9 10 27 7 5 16 14 5 3 3 7 12 6 2 12 2 3 2 2 2 5 4 8 1 9 7 4 1 1 5 6 3 2 5 2 1 1 3 1 1 2 19 32 35 86 48 35 84 43 27 11 4.10 3.34 4.57 4.45 4.64 4.21 4.43 4.50 4.61 4.82 4.S0 3 26 55 114 97 49 42 25 9 420 4.46 Table 3. — Classification of generation 19, plus-selection series. Grade of parents. Grade of offspring. Totals Means. H 3f 4 4i 4i- 4| 5 5i 5i 4i 4i 4i 4^- 4f 41 5 5i 3 8 4 1 3 2 15 3 6 1 4 25 4 4 15 6 4 3 1 23 4 5 20 12 7 6 9 3 2 8 10 3 3 7 2 1 8 8 3 2 4 2 3 8 1 1 1 2 95 13 17 65 47 22 14 4.21 4.34 4.56 4.50 4.49 4.76 4.47 4.53 4.66 3 16 27 65 78 38 31 18 4 280 4.49 38 HEREDITY IN RABBITS, RATS, AND MICE. Table 4. — ClassificaHori of generation 20, plus-selection series. Grade of parents. Grade of offspring. Totals. Means. 3f 4 4i 4^ 4f 5 5| 5i 5| 4^ 4| ^ 4| 4f 4J 5 51 5^ 1 2 1 2 1 3 5 9 1 1 1 1 3 5 3 4 3 3 7 2 3 1 3 1 3 1 2 1 3 2 2 1 1 3 2 2 1 2 9 15 19 7 14 13 9 3 3 4.22 4.50 4.42 4.57 4.77 5.02 4.61 4.89 4.75 4.66 1 6 21 28 13 11 7 3 2 92 4.61 Table 5. — Classification of generation 18, minus-selection series. Grade of parents. Grade of offspring (minus). Totals. Means. 2i 2h 2| 3 3i H 3f 4 -2| -2i -3 -3i -3i -3f -H -31 -3t 3 3 2 26 5 18 5 1 4 29 28 37 19 15 12 3 9 2 9 9 13 8 8 1 4 1 1 4 2 5 8 6 2 5 1 4 1 1 3 6 1 1 1 2 1 1 61 49 68 47 34 28 9 29 5 -2.63 -2.78 -2.72 -2.94 -2.97 -3.00 -3.14 —3.03 -3.45 -3.09 8 59 152 55 34 16 4 2 330 -2.84 Table 6. — Classification of generation 19, minus-selection series. Grade of parents. Grade of offspring (minus). Totals. Means. 2i 2| 2f 3 3i 3J 4 -2f -2| -3 -3i -H -31 -31 3 1 5 4 3 2 2 11 16 11 7 2 1 2 3 7 15 2 7 2 1 3 7 2 2 1 1 1 2 1 2 1 5 21 36 37 11 13 3 3 1 -2.80 -2.79 -2.85 -2.95 -2.89 -3.08 -3.10 3 15 50 38 17 6 1 130 -2.89 TABLES. 39 Table 7. — Classification of generation 20, rninus-seUction aeritt. Grade of parents. Grade of offspring (minus). Totals. Means. 2 1 ol -4 2i 2J 3 3i 3J -2i -2| -n -3 -3i 1 1 8 6 1 9 2 21 2 5 3 1 6 4 1 6 1 21 5 40 7 6 -2.67 -2.80 -2.81 -2.86 -2.S7 -2.S1 1 2 1.5 39 14 1 79 -2.78 Table 8. — Classification of generation 21, minus-selection series. Grade of parents. Grade of offspring (minus). Totals. Means. 2 2\ oi 2f 3 3i 3^ -2| -2f -2} -3 -31 1 2 1 7 4 2 1 2 1 1 1 2 3 1 1 2 3 14 7 3 1 10 -2.61 -2.G1 -3.11 -2.58 1 3 14 7 3 O 5 35 -2.74 Table 9. — Classification of extracted hooded third Fj young produced by a cross of plus-selected un.th wild rats. 2d F-z grand-parent. 2 n ol -2 o3 -4 3 31 3i Totals. Means. a 9 208, +2l 9 9922, +3 2 1 o 1 1 1 1 2 6 1 9 9 .... cf63, + 3i Weighted mean, 3.22. o 1 2 2 1 2 9 19 3.04 .64 Note. Standard deviation oi first F2 was 0.73, that of second Fj was 0.50 (tables 141 and 145, Publication No. 241). Table 10. — Classification of extracted hooded first F2 young from a cross minus-selected with ivild rats. of Hooded grand-parent. n. 2 If n li 1 4 4 13 4 26 f 1 1 2 7 1 10 _i 4 1 1 2 8 1 7 3 19 + i i 2 1 3 f 1 1 1 4 4 8 li 1 1 1 1 2 3 •• t n 1 1 1 3 2 2 1 3 2i 2i 2 23: I >i H 29 I 16 5 53 18 1 121 c 09 2 a 1.26 9 20,331, -2f. gen. 15^ 9 20,482, -2igen. 1.5f 920,359, -2i gen. 16. 9 20,327, -2f, gen. 15| 9 20,480, -2f, gen. lof 2 1 3.. 2 2 2 7 . 6 716 1 6 7 1 1 + .46 -.41 -.65 -.49 — 1 Total 2 -.38 40 HEREDITY IN RABBITS, RATS, AND MICE. Table 11. — Classification of extracted hooded second Ft young from the cross of minus-selected with wild rats. From original hooded. 1st Fa grand-parent. If 1 1 li a 1 1 1 3 4 1 2 1 1 1 4 2 4 _ 7 +i 1 2 2 2 3 1 8 3 4 1 1 9 1 1 2 1 1 5 4 1 5 1 3 2 6 If 1 1 2 4 2 1 2 2 5 2i 2| 2f m o Means. O 9 20,480.. 9 20,327.. 9 20,327.. 9 20,480.. 9 1G9S, -If 91715, -If 91563, -1^ 9 944, +1^ 1 1 1 1 14 17 9 9 1.11 .80 .67 1.64 Weighted mean, —1.11 1 1 2 49 1.01 .92 Table 12.' — Classification of extracted hooded third Fi young from a cross of 7ni?ius-selected with wild rats. From 1st F2. From 2d F2. 1 li . . 1 1^ 1 2 2 If 3 1 1 2 1 2 2 2 1 1 2J 2~ 2f 3 3| H a QJ 9944, -li... 9944, +lh-.- 9 944, +li.... 91563, -li... 91698, -If... . 91563, -li... 9 1924, +2 3 1 2 2 1 9 3 1 1 8 3 5 1 2 8 6 2 1 4 4 3 3 2 1 4 5 41 23 13 7 6 14 2.52 2.51 2.68 2.25 1.79 3.05 9 1925, +^ cf 1926, +lh 9 2008, +^ 1 1 cf 2048, +lf c^2068, +3 Weighted mean, +1.62. .. . 6 1 5 8 6 5 14 19 21 14 104 2.55 .66 Table 13. — Grade distribution of Fi young sired by the standard-bred Dutch ('41T X 97641(2)D 2 2 4 3 1 5 2 5 16 2 2 4 5 3 5 2 9 X $6058(4)D c?6701(5)DX $7209(2)T cf 5757(3)T X 9 5170(6)D " X 95939C6)D Totals 24 1.00 Table 26. — Grade distribution of the F\ young -produced by the cross of white Dutch with tan Dutch. Parents. Grades of young. Totals. Mean. 6 7 8 9 10 11 12 cf5757(3)T X 97003(16)W X 97185(16)W d^6175(17)W X 96424(3)T 1 1 6 3 •• 1 •• •• 3 8 4 3 5 .... X 96539(4)T 2 2 1 •• •• •• Totals 4 11 1 1 ■• ■• 3 20 7.70 Table 27. — Grade distribution of the Fi young from crosses of dark Dutch with self and of ivhite Dutch with self. Parents. Grades of young. Totals. Means. 1 2 3 (Dark X self.) 0^6701(5) X 9 7413(0) 3 9 1 ■• 4 9 X 98012(0) Totals 12 1 •• 13 .08 (White X self.) c^6175(17) X 9 61.33(0) X 9 7123(0) X 9 7124(0) X 98265(0) 4 11 3 5 o 4 2 2 .... Totals IS 13 2 33 1.51 TABLES. 45 Table 28. — Grade distribution of the Ft young from the several crosses made between the three types of Dutch and between white Dutch and self. Cross. Grades of young. 1 2 3 4 5 6 7 8 9 10 U 12 13 14 15 4 13 16 1 2 17 Totals. Means. White X self 41 45 1 22 1 14 2 1 8 9 12 7 11 5 15 12 17 5 6 13 4 5 14 ft ft 8 4 10 4 4 12 2 8 1 10 5 1 4 6 4 2 191 White X dark (taVjles 16- White X tan -18). 7 3 lS'>ft 56+25 5.828nd 14.40 130 275 Dark X tan 89 84 25 8 4 14 9 1 Table 29. — Grade distribution of young produced by other crosses of Fi animals. Cross. Grades of young. Totals. Means. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 13 7 5 17 6 2 1 Fi (white X self) X white Fi (white X self)X self.. Fi (darkX self) X dark. Fi (white X dark) X dark Fi (white X tan) X white Fi (dark X tan) X dark. Fi (dark X tan) X tan. . Fi (dark X tan) X white 65 5 4 19 1 2 26 36 3 12 9 1 3 3 9 12 6 1 5 4 11 3 10 10 11 2 36 13 23 11 33 14 14 14 30 4 5 9 12 1 4 7 3 2 2 1 1 3 5 1 3 1 8 8 4 6 12 1 13 8 14 3 10 24 18 3 58+58 94 59 143 73 25 76 112 5.15and 15.15 Table 30. — Grade distribution of young sired by the original English male, 2545, grade 2\. A. Mother English. Mother. Grade. Self young. English young of grade — Total English. 1 u ih If 2 2\ 2^ •21 3 7 2649 2650 2651 2 3 2f 1 5 2 2 2 1 1 2 1 2 1 1 1 1 4 2 3 13 5 Totals . . . Means . . 8 4 1 4 2 1 1 2 6 21 1.0. =5 2.32 B. Mother S ELF. Mother. Self English young of grade — ! Total young. Ih li 2 2J o3 -4 3 3i 3J 3f 4 / English. 1443 2 .. 3 3 1492 5 1 1 3 5 2053 2 . . 5 5 2502 5 1 2 3 2867 3 3 3 2912 2 1 1 1 . . 5 2916 1 •• •• 1 1 O Totals . . . Mean . . 18 2 ■ • 1 ■■ 1 •• 1 2 1 1 17 26 2.80 1 46 HEREDITY IN RABBITS, RATS, AND MICE. Table 31. — Grade distribution of young sired by English male 2711, grade 4, generation 1, ancestor of all English in the selection series. A. Mother English. Mother. Grade. Gener- ation. Self young. English young of grade — Total English. Means, higher group. 3 1 4 J 1 . H U lf22i 2i 2|c i3i 3^ 1 1 2 3|4 1 1 2 I4i 4^ 4f5 6051 5052 6206 6083 5084 6402 3h 3h 4f 5 5 5 2 2 1 2 1 1 ] 1 . 1 1 1 1 1 1 3 . . 1 . . 1 4 4 6 4 3 3 5 js.ss 3.80 4.25 . . ] . . ] [. . i. . 2 1 2 1 Totals . 8 1] L 2 4 2 1.. 1 3 26 Mpanfl 1.52 3.93 B. Mother Self. Mother. Self young. English young of grade — Total English. 3 3i 3^ 3f 4 4i 4^ 41 5 2765 1 2 1 3 2770 1 2 .. 1 2 5 2840 4 1 1 1 ^ 3 2862 4 , , 1 1 2 2867 3 1 1 , , 1 , 2 5 2878 2 1 _ , . , 3 2947 4 , , . , 2948 4 , , , , , , 2 1 . , 3 2983 4 1 1 , , , , . . 1 3 3019 3 1 •• •• •• •• 1 Totals .... Mean 28 3 5 4 2 5 3 1 5 28 3.89 Table 32. — Grade distribution of young sired by English male 6086, grade 4h generation 1 \. A. Mother English. Mother. Grade. Gener- ation. Self young. English young of grade — Total. English. 1 2 3 1 1 li l\ If 2 2i 2\ 2f 3 3i 3i 3f 4 4t 4i 4f 5 5051 3| U 2 1 1 1 1 1 1 1 1 1 9 5052 3i U 8 2 1 1 2 1 2 2 2 2 2 18 5188 4i 2h 2 , . ^ 1 1 5206 4f 2h 2 1 1 2 4 5053 5 n 2 1 2 1 4 5083 5 n 6 _ 1 1 2 1 . , 6 5084 5 u 3 1 1 3 5101 5 u 3 1 2 1 1 1 1 1 3 1 12 5102 5 u 6 1 1 1 1 1 5 5398 5 If 1 1 8 1 2 Totals . Means . 35 1 3 5 4 4 3 3 4 5 8 4 5 6 1 64 1 ss .■^ Qfi TABLES. 47 Table 32, continued, B. Mother Self. Mother. Self young. English young of grade — ToUl Engliih. li li li 2 2i 2i 2J 3 3J 3i 2 1 3 3} 4 1 1 4i 4i 4J 5 4146 4147 4148 3 2 3 1 1 1 1 1 •• 1 7 1 2 Totals. . . . Mean .... 8 1 1 •• 1 2 1 10 .•? Mi Table 33. — Grade distribution of young sired by English male 6S76, grade 4\, generation £\. Mother. Grade. Gen- era- tion. Self young. English young of grade — Total Eng- 1 lish. Means. 1 1 2 1 1" Lli I 1 ■ 1 1 1 2 lf2 2 . . . 1 2 2 2i 1 2f 1 1 3 1 1 2 3J 1 1 3i 1 1 1 3 33^ 1 1 ] 1 . 1 1 1 ': 2 . 2 . 9 i 14i 4J 1 4f£ 5051 5052 5188 5205 5206 5083 5084 5102 5398 3i 3^ 4i 4i 4i 5 5 5 5 2^ 2^ 2J 11 If 2 2 1 1 1 2 5 9 2 6 8 4 5 4 1 6 8 4 I3.4O js.ss 3.75 4.11 1 J . . 5 . . 1 .. 1 .. 1 I . . 2 . . 2 Totals . 25 46 Means . 1.20 3.7Q Table 34. — Grade dist •ibution of young sired by English male 5555, grade 41, generation 2\ Mother. Grade. Gen- era- tion. Self young. English j'oung of grade — Total Means, - IT. . 1- -• t _ 1 2 " 1 ] \ 1 1 1 2 2 2 2 I . . 2 U 1 1 1 2 3 1 1 2 1 1 1 1 ■ If 1 2 1 1 1 2 1 1 2i 2h 1 2f; j|3i 1 I.. 1 1 2 3i 2 1 2' 1 1 7 3f 1 1 1 1 1 1 1 '_' 7 4 1 5 2 1 4 1 1 1 1 4i 3 2 2 1 2 4J 2 1 ■3 1 3 4f 1 1 1 2 1 1 1 5 i 1 1 5 (Eng- .' lish. Qigner group. 5557 5051 5052 5561 5701 5752 5993 5188 5205 5672 5769 5952 6074 5206 5793 5801 5988 5084 5102 5398 5951 3 31 3^ 3i 4 4 4 4i 4i 4^ 4^ 4^ 4i 4f 4f 4S 4^ 5 5 5 5 2i ih 2i 3i 3 3 2i 2^ 3 3i 3 3i 2^ 3i 3 3 If 3 3 3 2 1 1 9 1 6 2 1 2 6 2 3 1 12 2 7 1 7 I 11 3 1 11 5 2 19 3 [ 6 3 5 5 22 8 15 4 15 12 9 3 4.04 4,47 4.44 1 ] i 1 3 3 2 4 2 2 3 6 4 2 2 2 2 . 1 . 2 . 2 . 1 . Totals 65 I2J9 7 6 4 1 . . ] isjio 2' 168 Aleans 1 "* 1.36 4.40 ' 48 HEREDITY IN RABBITS, RATS, AND MICE. Table 35. — Grade distribution of young sired by English male 6S70, grade 5, generation S. Mother. Grade. Gen- era- tion. Self young. English young of grade — Total Eng- lish. Means, higher group. 1 li 1§ If 2 2i 2| 2| 3 3i 3^ 3f 4 4i 4i ^2 4f 5 5752 6993 5188 5672 5769 6074 6089 7193 6815 6369 6417 6693 6788 4 4 4i 4^ 4J 4^ 4^ 4^ 4| 5 5 5 5 3 3 2\ 3 3i 3i 3^ 4i 3i 3 3 4i 4i 4 2 4 2 3 1 1 1 3 l' 1 2 1 1 1 1 1 3 1 1 1 1 1 1 1 • ■ 1 1 3 1 2 1 2 2 1 2 4 2 1 1 1 2 1 1 1 2 1 2 4 1 1 1 2 2 5 1 1 4 7 7 1 7 2 7 4 17 4 3 7 2 ■4.67 4.63 Totals . 21 1 3 3 2 5 2 2 1 1 8 15 11 18 72 Mean 1.79 4.0(1 Table 36. — Grade distribution of young sired by English male 6420, grade 4 h generation 3. Mother. Grade. Gener- ation. Self young. English young of grade — Total English. 1 n I5 If 2 9i -4 2^ 2| 3 H 3h o3 4 4i ^ 4f 5 5993 4 3 1 1 • • 1 2 4 5188 4i 2i 2 2 1 , , 5 6074 4^ 3i 1 3 2 1 7 5102 5 H 1 1 1 1 1 1 5 6417 5 3 1 _ 1 1 1 3 6841 5 41 1 1 1 2 1 5 Totals . 4 2 2 2 1 1 1 1 2 1 5 1 3 5 2 29 Means i.sn 4..'^.'^ Table 37. — Grade distribution of your 9 sired fey English mi iZe 6071 grad e5 , generation S | Mother. Grade. Gen- Self English young of grade — Total Eng- Means, higher era- 1 1 1 tion. young. 1 n u li 2 9191 2f 3 3i H 3f 4 4i 4| 4| 5 lish. group. 5051 3^ H 1 2 3 5 3.75 5752 4 3 1 1 1 1 1 2 6 4.56 5769 4i 3^ 1 1 2 4 4.83 5891 ^ 3* 2 1 1 1 3 •4.25 6074 ^ 3^ 1 1 1 1 1 1 5 5206 4f 2^ 1 2 1 4 5793 4| 3i 2 1 4 5 5988 4f 3 3 2 1 1 1 2 7 [4.47 6073 4f 3^ 1 1 _ 2 1 4 6189 4f 3 1 2 1 1 4 5084 5 H 2 1 1 1 1 4 \ . ,0 5951 5 3 2 1 2 3 U.12 Totals . 16 4 3 3 7 3 2 3 5 2 5 3 6 8 54 Means. 1 fii A 50 TABLES. 49 Table 38. — Grade distribution of young sired by English male 6072, grade 5, generation S\. A. Mother HETEnozraous English. Mother. Grade. Gen- era- tion. Self young. . English young of grade — Total Eng- lish. Means, higher group. ill I' L . lili IJ 2 2i 2i 2iC 1 ! 2 ! 1 1 (3i 3J 3jU!4i 4i!4j 5 2 1 1 , . 2 2 1 1 3 4 1 2 1 2 2 1 2 5 1 1 1 61 5752 5701 5993 5188 5769 5891 6074 6079 6089 6452 6795 5206 5801 5988 6189 6264 7450 7817 8813 5084 5102 6188 6369 6416 6417 6622 6693 6841 7476 4 4 4 4i 4i 4^ 4^ 4-^ 4^ 4| 4^^ 4f 4f 4f 4| 4f 4f 4f 41 5 5 5 5 5 5 5 5 5 5 3 3J 3 2^ 3i 3^ 3| 4 3J- 4 3| 2h 3 3 3 3f 4^ 3^ 4f ih ih 3 3 3 3 3i 4i 4i 4i o 3 2 4 3 2 1 7 2 3 3 4 6 4 6 5 1 2 9 1 2 2 4 1 2 1 1 1 2 V 2 1 1 1 1 1 1 1 1 i 1 ■ . 1 1 1 2 2 2 1 1 2 1 3 1 2 2 1 1 1 1 1 1 2 • • 2 1 4 1 1 1 1 1 1 1 1 1 1 1 5 1 2 3 1 1 1 4 3 4 1 1 4 1 1 3 1 1 4 4 2 2 3 1 1 1 4 1 1 4 3 3 4 7 4 J 1 4.50 [4.42 4.58 4.62 4.69 2 1 i'.. I 1 I 12 4 8 12 12 1 11 15 13 5 1 3 4 7 16 7 10 22 5 5 12 19 4 2 1 3 3 1 1 1 1 1 2 1 1 1 2 1 2 Totals.. 81 12 8 7 16 15 9 8 5 ' 1 . . 3 13 15 35 56 36 1 234 Means . 1 1 1.97 4.63 B. Mother Homozygous Enolise [. Mother. Grade. Gener- ation. Self young. English young of grade — Total English. 1 U U 1^ 2 0.1 -4 2| o3 -4 3 31 1 3J 1 1 3! 1 1 2 4 2 1 3 4i 4J 4J 2 2 4 5 1 1 5733 7535 7814 8704 2 U ll 3 4 3^ 4i 1 2 1 1 2 2 1 * ' 3 1 1 1 8 1 14 5 Totals . . . 1 2 1 1 4 1 4 2 1 28 Means. . . 2.11 4.17 C. Mother Self. Mother. Self young. English young of — Total English. 3 J n 4 4i 4J 4J 7123 7878 6 3 1 . , 1 1 2 1 1 3 4 Totals .... Mean .... 9 1 1 1 2 2 7 4.32 50 HEREDITY IN RABBITS, RATS, AND MICE. Table 39. — Grade distribution ofyov ing Sired fcy English 7nale 696. 'f, grade 5, generation 4h Mother. Grade. Gen- era- tion. Self young. English young of grade Total Eng- lish. Means, higher group. 3 4 1 li H If 2 ■21 2^i2f 3 3i 3^ 3f 4 4i ^ 4f 5 6074 7193 7475 7356 7817 5102 6417 6693 6841 4f 4| 4f 5 5 5 5 4i 4i 4^ 3i li 3 4i 4i 4 2 1 2 1 1 2 4 1 1 1 1 1 1 1 2 1 1 2 1 1 1 1 1 2 1 1 1 * ' * * 1 4 1 2 1 2 1 3 2 2 1 2 2 • • 1 2 8 4 3 2 8 11 6 6 2 I4.64 ■4.78 < •4.65 Totals. 21 2 4 4 4 1 3 2 1 ■ • • • 2 1 8 10 8 50 Means. . l".'49 4.68 Table 40. — Grade distribution of young sired by English male 7699, grade 5, generation 4h A. Mother Heterozygous English. Mother. Grade. Gen- era- tion. Self young. English young of grade — Total Eng- , lish. Means, higher group. Lli l§ If 2 2i 2§ 2f 3 3i 3^ 3f. t4i 4i 4f 5 5i 9449 6089 5988 6795 7300 7356 7389 7475 7817 9350 5084 6369 6416 6417 6622 6841 7325 7476 7903 8257 8259 9349 9363 9091 9535 31 4i 41 4f 4f 4f 4f 4f 41 4f 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5i 3^ 3i 3 3^ 4i 4i 4i 4i 3^ 5i 3 3 3 3f 3f 3^ 4i 4 4| 4f o\ 5 5i 3i 2 2 1 4 8 1 2 2 4 3 4 5 3 3 3 7 3 7 1 1 2 1 1 1 1 1 1 1 1 2 1 14 3 1 1 1 3 1 10 1 2 1 1 6 1 1 2 1 1 2 2 1 11 2 1 1 3 3 2 13 1 1 2 1 1 2 1 . 1 ] 1 3 i 2 1 L 1 1 1 I 6 1 1 1 1 4 1 1 1 1 1 3 1 1^ 3 4 5 8 6 10 1 3 3 17 6 17 9 10 11 L 10 9 20 3 8 9 . 16 . 15 2 16 I 5 2 4 4.50 4.62 ■4.66 •4.80 >4.S1 1 1 2 1 3 1 2 5 5 3 2 3 1 3 2 4 1 3 43 1 1 1 1 1 1 1 1 3 6 2 2 1 1 3 4 1 5 2 5 6 6 1 2 re 1 . . ] 1 2 1 1 2 8 . 1 2 1 1 I . . 2 1 2 1 1 1 L 5 5 6 Totals . . 64 ] 3 224 Means . . •^ ?.^ 4 >>li a If 2 21 1 2^ 2f 3 31 3i 3| ^ I4i4i .. 1 4J5 2 . 5i^W « £ s:s& Het. Eng 6795 4f 3^ 3 1 5 7475 9362 4| 4f 4i 5 1 . . 2 . . 2 1 '. . . 1 . . .. 1 3 . 1 . 4 5 2.46 4.69 9538 4| 4| 1 . . • . . . 1 . 1 6369 5 3 1 .. I . . . . . . 2 .. 4 6622 8257 5 5 3f 4| .. 2 . . .. 1 . 2 . . . 2 2 . 1 1 .. 5 4 2.30 4.70 8259 5 4f . . 1 . . 1 4 . 6 9535 9592 5i 5^ 3i 4i . . . . . . .. ] ) 2 2 1 1 4 1 2 .. .. 2 1 .. L .. 5 1 . 1 1 16 4 1 4 .. 5 2.85 4.94 Totals. Means. Horn. Eng. Self, 10 1 . . 1 :: 1 43 2.53 1.75 4.73 4.30 2.53 1 4.7: I 1 1 1 6 7814 li ?* . . 1 L 8251 8265 3 . . . . 1 2 . ?. .. 3 2 } 4.80 3 . . Table 42. — Grade distrihudon of 7/oung sired by English male 9S06, grade 5, generation 5 J. Mother. Grade. Gener- ation. Self young. English young of grade — Total English. 2 1 2\ 3 1 2h 1 2| 1 3 2 1 3i 3J 3| 4 4i 1 1 4J 1 1 4J 1 1 1 1 4 5 2 4 1 1 2 10 6844 9871 6369 6622 456 3i 4f 5 5 5 4i 4 3 3f 4i 2 2 1 1 1 5 9 3 5 4 Totals. . . 7 1 4 1 1 3 26 Means. . . . 2.52 4.S6 52 HEREDITY IN RABBITS, RATS, AND MICE. Table 43. — Grade distribution of young sired by English male 1212, grade 5, generation 5. A. MoTHEB Heterozygous English. Mother. Grade. Gener- ation. Self young. English young of grade — Total English. Means higher group. 2 2i 2^ 2f 3 3i •J 2 3f 4 4i 4i 4f 5 5i 9943 455 539 1360 6795 7475 8257 452 9372 230 6841 537 6622 112 1222 596 1613 1860 6369 9363 111 597 4| 4f 4f 4f 4f 4f 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5i 5i 5i 4i 5i 51 3i 4| 4f 5f 4 4| 4i 5J 31 5f 5i 51 4i 6^ 3 5 5f 5| 2 1 7 4 1 2 1 1 2 4 2 1 1 1 1 1 2 1 1 1 1 1 1 2 1 2 1 1 1 1 1 1 1 1 1 1 1 2 1 1 2 1 1 1 1 4 2 5 2 2 2 4 3 3 1 1 2 1 1 5 2 1 2 2 1 * 1 3 5 3 4 7 5 4 5 1 5 5 4 3 2 7 1 3 7 3 3 8 2 ■4.97 ■4.97 >5.15 Totals . . . 29 7 1 5 4 2 2 3 2 11 43 7 87 Means . . . 2.87 4.Q8 B. Mother Homozygous English. Mother. Grade. Gener- ation. English young of grade Total English. If 2 2i oi -2 -4 3 3i 31 3f 4 4i 4^ 4f 5 8682 9953 9684 9970 If 2f 3| 4f 5 4§ 4| 1 1 1 1 4 1 1 1 1 1 2 3 3 2 11 3 3 6 Totals . . . 1 1 1 1 4 2 1 7 5 23 Means. . . . '^..'Sfl 4 7E 1 TABLES. 53 Table 44.— Grade distribution of young sired by English male 6S4, grade SI generations A. Mother HEXERozYaois Knolish. I Mother. Grade. Gener- ation. Self young. English young of grade — Total Englioh. Mean*, higher group. 21 2J 3 1 3 1 2 1 2 3i * 1 1 1 3i 1 2 1 31 1 1 1 4 1 4i 1 4i 4| 1 1 3 1 1 1 1 6 2 1 1 2 3 2 1 5 2 1 1 2 2 2 3 3 3 2 1 3 5i 2 1 2 1 2 1 9684 9943 7475 9871 360 455 584 1360 539 6369 9363 7906 258 452 230 456 537 1222 6841 112 9091 2325 111 597 9592 3J 4-J 4J 4| 4| 4i 4i 41 5 5 5 5 5 5 5 5 5 5 5 5i 5i 5i 5i 5i 4\ 4 5i 4\ 5k 5i 5i 3 5 4 5J 4i 5\ 5} 4i 5f 5i 5i 5f 5f 4i 2 1 2 'o 2 1 1 2 1 1 1 1 1 1 1 1 1 2 ' ■ • • 1 2 1 1 1 1 3 4 2 4 6 2 5 5 5 7 2 6 3 2 4 1 2 2 5 5 6 4 3 5 4.87 5.02 5.06 Totals . . . 21 5 6 11 3 4 3 1 1 9 42 9 94 Means . . . '>'un 4 07 B. Mother Homozygous English. Mother. Grade. Gener- ation. English j'oung of grade — Total English. u H li 2 21 3 3\ 3i ■31 4 4i 4J 4J 5 7814 8682 9953 If 2f 3^ 4f 5 1 1 1 1 2 ' 1 3 1 1 1 1 6 5 3 Totals . . . 1 1 1 1 2 4 2 2 14 Means . . . 2.37 4.69 54 HEREDITY IN RABBITS, RATS, AND MICE. Table 45. — Giade distribution of young sired by homozygous English male 1173, grade S\, generation 5. A. Mother Heterozygous English. Mother. Grade. Gener- ation. English young of grade — Totals. n If 2 2i 2J 2i 3 3i 3i 3f 4 4i 4^ 4f 5 5i 9684 9970 7300 455 584 6369 8257 1222 3i 31 4| 4f 4f 5 5 5 4§ ^ 4i 4 3 4f 5i 1 1 3 2 1 2 1 2 1 2 1 1 1 1 1 2 2 2 1 1 3 1 3 1 3 1 3 1 5 8 8 5 8 5 3 4 Totals . . . 1 4 5 4 2 4 2 2 2 2 2 7 8 1 46 Means . . . 2.42 4 77 B. Mother Homozygous English. Mother. Grade. Gener- ation. English young of grade — Totals. 2 2\ 21 2f 3 7814 9953 2f 3^ 5 1 9 1 2 4 4 6 Totals . . . 1 2 1 2 4 10 Mean .... 2 fi.i^ C. Mother Self. Mother. English young. Totals. 4i ^ 4f 5 5J 7123 7124 8251 9318 1 2 2 4 3 2 2 9 5 1 8 16 2 5 Totals . . . Mean. . . . 1 4 9 16 1 31 4.84 TABLES. 00 Table 46. — Summary of young of selected heterozygous English males by heteroeygous English females. Homo- Hetero- Un- Sire. Gener- Self zygous Mean zygous Mean graded T • • -:: • ,1 ation. young. English young. grade. English young. grade. K..gli.»h young. 1. •C- 2711 4 1 8 12 1.52 11 3.93 3 26 34 50S6 H H 35 20 1.38 44 3.96 64 99 5375 ^ 2^ 25 17 1.20 29 3.79 46 71 5555 4f 2i 65 41 1.36 125 4.40 o 16)> 233 6370 5 3 21 18 1.79 54 4.60 72 93 6420 4| 3 4 10 1.80 19 4.33 29 33 6071 5 3^ 16 22 1.01 32 4.39 54 70 6072 5 3^ 81 75 1.97 159 4.63 1 235 316 6964 5 44- 21 21 1.49 29 4.08 50 71 7699 5 4i 64 75 2.31 149 4.80 8 232 296 9532 5 4k 10 16 2.53 27 4.73 43 53 9S06 5 5^ 7 10 2.52 16 4.86 26 33 1212 5 5 29 26 2.87 61 4.98 87 116 534 5i 5 21 32 2.95 62 4.97 94 115 Totals . . 407 395 817 14 1.226 1,633 Percent . 24.9 24.2 50.0 0.9 75.1 Table 47. — Line of advance in the selection experiment unth heterozygous English males. Sire. No. of homo- zygous young. Mean of homo- zj-gous young. No. of hetero- zygous young. Mean of hetero- zygous young. Advance, homo- zygous young. Advance, hetero- zygous young. 2545 5 12 41 81 75 58 1.05 1.52 1.36 1.97 2.31 2.91 9 159 149 123 2.80 3.93 4.40 4.63 4.80 4.97 ^47 -.16 .61 .34 .60 Li3 .47 .23 .17 .17 2711 5555 6072 7699 1212 and 534 Total advance . . .... Average advance ... . . .... .... 1.86 .37 2.17 .43 Table 48. — Relative frequency of cross-overs among the gametes formed by Fi rats of the two sexes. Series. Fi parent. Dark-eyed young. Red-eyed or pink-eyed yellow young. Total young. Percentage dark-eyed young. Percent Repulsion Coupling Female. . . . Male Female. . . . Male 101 73 699 556 837 703 1,046 731 938 776 1,745 1.287 lO.S 9.4 40.0 43.2 21.5 18.8 19.8 13.5 56 HEREDITY IN RABBITS, RATS, AND MICE. BIBLIOGRAPHY. Castle, W. E., and P. B. Hadley. 1915. The English rabbit and the question of MendeUan unit-character constancy. Proc. Nat. Ac. Sci. 1, pp. 39-42, 6 figs. and John C. Phillips. 1914. Piebald rats and selection. Carnegie Inst. Wash. Pub. No. 195, 56 pp., 3 pi. and S. Wright. 1916. Studies of inheritance in guinea-pigs and rats. Carnegie Inst. Wash. Pub. 241, 192 pp., 7 pi. Haldane, J. B. S., A. D. Sprunt, and N. M. Haldane, 1915. Reduplication in mice. Jour. Genet. 5, 133-135. [Reference from Detlefsen (1918) Genetics, 3, p. 597.] King, Helen D. 1918. The effects of inbreeding on the fertility and on the constitutional vigor of the albino rat. Jour. Exp. Zool. 26, pp. 335-378. Punnett, R. C. 1912. Inheritance of coat-colour in rabbits. Jour. Genet., 2, pp. 221-238, 3 pi. Plate i 4 5 10 11 4a .^^rA- j^ V.i 14 J J 12 ir, 16 17 Grades 1-ls ot Dutch rahl^ils. IK PLATE 2 10 -Ml F- i Jl V"\ii. I'.t, .•! ■ while" Diil.li r.iMiii. 9 7',i:;i, ii\:\i\v I.'., l"in. L'O. :i "(liirk" Diilrli r.iMiii. o'iuOl, gnulc .'). I'ifi. Jl. .1 •t.in " Diilrh r.il>l>it. 9717. uniclc :{. PLATF :i l'"[(;s. l-o. Photographs of KiifiUsh rabbits adopted :is f;r;ulrs l-.">iii classifying tlic variatioii~ observed in the Knfjhsh pattern. Those shown in tigs. 1 and J wrrc hoinozynons. thosi- shown in figs. :\ to ") were heterozygous. Fk;. (). A 'hifili-grade" homozynous Eufihsli ral)bit, 1 17.'5. ^rade 3'2. Kriielically riini|viniblc with a firadc .'> hctcniwjioii^ l-]ntilisli. Compare Hn- •''• v^