•*€>

STUDIES OF

INHERITANCE IN GUINEA-PIGS AND RATS

BY

W. E. CASTLE and SEW ALL WRIGHT

Published by the Carnegie Institution of Washington

WASHINGTON, 1916

('akxegie Institution of Washington, Publication No. 241

Paper No. 26 of the Station for Experimental Evolution at Cold Spring Harbor, New York

From the Laboratory of Genetics of the Bussey Institution

Copies of this Book

were first issued

SEP 20 1916

43i

PRESS OF GIBSON HROTHERS. INC. WASHINGTON

CONTENTS.

Part I. — An Expedition to the Home of the Guinea-pig and Some Breeding Experiments with Material there Obtained. Bt W. E. Castle.

page.

Introduction 3

Some observations on guinea-pigs in Peru 5

Hybridization experiments with Cavia cutleri 8

Life history of C. cutleri 8

Crosses of C. cutleri males with guinea-pig females 13

Color inheritance among the F2 hybrids 13

(a) Cross 9 albino (race B) X d" C. cutleri 13

(6) Cross 9 albino (race C) X cf C. cutleri 14

(c) Cross 9 brown-eyed cream (race C) X cf C. cutleri 16

(d) Results from (b) and (c) combined 16

(e) Intensity and dilution among the hybrids 17

(/) Significance of the results observed 18

Hybridization experiments with a race of feral guinea-pigs from lea, Peru 20

Origin and characteristics of the lea race 20

Crosses between the lea race and guinea-pigs of race C 23

The F2 generation 24

Summary on the lea race 29

Hybridization experiments with a domesticated guinea-pig from Arequipa 31

d1 1002 and his Fi offspring 31

F2 offspring of d" 1002 33

Back-cross and other offspring of d1 1002 35

Miscellaneous matings of the descendants of cf 1002 36

Summary on the Arequipa domesticated race 41

Size inheritance in guinea-pig crosses 42

Previous work on size inheritance 42

Weights and growth curves of C. cutleri, of various guinea-pig races, and of their

hybrids 43

Skeletal measurements of C. cutleri, of various races of guinea-pigs, and of their

hybrids 47

The C. cutleri hybrids 48

Hybrids of the Arequipa d1 1002 52

The lea hybrids 53

Theoretical explanations of size inheritance and of "blending inheritance"

in general 54

Part II. — An Intensive Study of the Inheritance of Color and of Other Coat

Characters in Guinea-pigs, with Especial Reference to Graded Variations.

By Sewall Wright.

page.

Color and its inheritance in guinea-pigs 59

Skin, fur, and eye colors of guinea-pigs 59

Color of Cavia cutleri 59

Melanin pigment 59

Primary classification of fur colors 59

Yellow group of colors

Dark group of colors "0

Skin colors "1

Eye colors "2

Definitions of fur colors by Ridgway's charts 62

Definitions of eye colors

Heredity of fur and eye color

Color factors of guinea-pigs 63

Classification of color factors 63

iii

jv CONTENTS.

. iad us inheritance in gumearpiga-Gmtinued. PAGE-

II, r. ditv of fur and eve color— Continued ^

( "oior m, white fi4

Lntenatj of genera] color development "*

I >ark M. J ellow color

\ .in.it 1. .us of dark color

1 able of factor combinations

Hereditary factora and the physiology of pigment °'

:> Of experiment-

Material 7,

- • .mat io position

Description of stocks '

1 7 < >1 >lcmB

[nberitanoe of dilution \}

The red-eye factor \}

I Hlul ion ™

The dilution factor • 80

Inheritance of minor variations in intensity 85

Methods and accuracy of grading 85

Variation! in intense guinea-pigs and albinos 85

Multiple allelomorphs • .' ' ' " ^

The relations of imperfect dominance, stock, and age to grades of intensity . . 87

Yariat ions of yellow 89

Varial ions of sepia 92

Variations of eye color 93

Summary 93

Inheritance of variations in the agouti pattern 94

Previous work 95

The inheritance of the agouti of C. rufescens 96

Minor variations 99

inheritance of the agouti of C. cvileri 100

Inheritance of rough fur 100

Classification 102

Previous work 102

Material 103

Problems 104

Inheritance of rough as opposed to smooth 105

Inheritance of major variations 106

Possibilities of linkage among rough and color factors 113

Summary of rough tables 115

>r variations 116

ighnees of Beries II 117

Summary 118

conclusion 119

1 data 121

tables 62 to 137 121

r '" 1 ' iMiiii: BxxmrjBB of Piebald Rats and Selection, with Observations

one Coupling. By W. E. Castle.

PAGE.

Jed with wild rats 163-168

lection of the hooded pattern of rats 168-172

rurth' on t|1(. mutant series 173-174

n yellow rata 175-180

181-187

188-190

191-192

PART I

AN EXPEDITION TO THE HOME OF THE GUINEA-PIG

AND SOME BREEDING EXPERIMENTS WITH

MATERIAL THERE OBTAINED

By W. E. CASTLE

INTRODUCTION.

For several years I have been engaged in studies of heredity in guinea-pigs. In the course of these studies all the common varieties of guinea-pigs have been investigated by the method of experimental breeding and something has been learned concerning their inter- relationships and probable mode of origin. The actual origin of most of these varieties is, however, unknown, as is true also concerning most varieties of domesticated animals. One or two varieties have, however, been made synthetically in the laboratory and it is conceivable that, if we had the original wild stock to work with, from which the domesticated guinea-pig has arisen, some or all of the existing varieties might be synthesized anew and perhaps still others might be obtained, and that in this way something might be learned of the method by which new varieties arise. From considerations such as these I have for several years been seeking to obtain living specimens of the wild species which most closely resemble guinea-pigs. In 1903 I received from Campinas, Brazil, 3 wild-caught individuals referred at the time to the species Cavia aperea, but since found to agree better with the description of C. rufescens. From two of these animals young were obtained, and crosses, the results of which have been described in detail by Dr. Detlefsen (1914), were made with domesticated guinea- pigs. It may be noted that all male Fi hybrids were sterile, but that the Fi females were fertile, and that upon repeated crossing of these with male guinea-pigs, a race of fertile hybrids was at last obtained, these being, in the language of breeders, about f guinea-pig, § rufescens. From this result it seems doubtful whether C. rufescens has any close genetic relationship to the domesticated guinea-pig, although by hybridization it has been found possible to produce races (f or more guinea-pig) which have derived certain characters from a rufescens ancestor.

Cavia aperea from Argentina has been crossed with the guinea-pig by Nehring (1893, 1894) in Berlin, with the production of fully fertile hybrids. This result indicates a closer relationship with the guinea-pig than C. rufescens manifests. Darwin (1876), however, did not regard aperea as the ancestor of the guinea-pig, because he found it to be infested with a different species of louse. I have not myself been able as yet to obtain specimens of C. aperea. Nehring (1889) has argued with much plausibility that Cavia cutleri of Peru is more probably the ancestor of the guinea-pig, for (1) it agrees closely with the guinea-pig in cranial characters and it occurs in a region where guinea-pigs have been for a long time kept in domestication, as is shown by the occurrence of mummified guinea-pigs which had been buried with the dead. Natu- rally I formed a strong desire to secure living specimens of C. cutleri for

3

\ tNTRODUCTION.

mental study, l>ut for several years I was unable to do so. rough pondence with Professor S. I. Bailey, who was at the

tin tor of the Harvard Astronomical Observatory at Arequipa,

oertained thai a wild species of cavy occurred in that locality.

IY< :■ • r Bailey kindly captured some of the cavies and attempted

eatedly to forward them to me, but without success. The steam- ship roiiip.-inii-; n-fn>«'«l to accept them for transportation on the ground that they might lead to detention or quarantining of their vessels, .11 rodents were inspected of being carriers of bubonic plague. 1 yean of waiting and fruitless negotiation with every chance traveler to Peru with whom I came in contact, I resolved to

to Peru myself and get the desired specimens. Through a grant made by the ( Sarnegie Institution of Washington I was enabled, in the fall of 191 1. to carry this resolution into effect.

The < larnegie Institution of Washington and the Bussey Institution ha mther provided means for carrying out the breeding experi- dm 9Ciibed in this paper. I wish to express my gratitude to both

institutions and to thank the director and other officers of the Harvard 1 ege < Observatory for hospitality and generous assistance given me at the Arequipa station. I am indebted also to Professor C. J. Brues for kindly bringing me a stock of guinea-pigs obtained by him near Lim l Peru, in 1912.

SOME OBSERVATIONS ON GUINEA-PIGS IN PERU.

On a midsummer day in December 1911 I arrived as a guest at the Harvard College Observatory in Arequipa, Peru, where I went in search of guinea-pigs, wild and domesticated, to be used in breeding experiments.

The day after my arrival at the observatory I walked a short dis- tance up the highway through a group of adobe cabins, straw-thatched and without chimney or windows, and with a single door. On looking in at the open door of one of the cabins, I was pleased to see a domesti- cated guinea-pig of the common spotted black-and-white sort familiar to lovers of pet-stock throughout the world. In other near-by cabins I found considerable numbers of guinea-pigs were kept, in one as many as 40. They were fed on fresh-cut alfalfa or the green leaves of maize, receiving apparently no other food and no water. At the back or sides of the cabin was a sort of shelf or bench of stone used as a seat or couch, underneath which the guinea-pigs had their home. Their escape through the open door was prevented by a high lintel of stone, perhaps 15 inches (38 cm.) high, over which one has to step in entering. In these cabins were seen most of the common color varieties of guinea- pigs known to us, agouti, black, yellow, and white (albino). None of the colored individuals which I saw was self-colored ; all were spotted with white or with yellow or in both ways. The same predilection for spotting is seen in the other important native domesticated animal, the llama. I saw no llamas except such as were spotted; some were black spotted with white, but the majority were of a soft shade of buff or fawn spotted with white. The common spotted condition of our guinea-pigs is undoubtedly one of long standing; indeed it would seem that the Peruvian natives breed no other variety except such as are either white spotted or all white. The unspotted or " self-colored " varieties now kept by fanciers in Europe and America have probably been produced by selection from stock originally spotted. This is indicated by the great difficulty in securing a self-colored race entirely free from spotted individuals. Most self-colored races, even when bred for many generations from self-colored ancestors exclusively, will pro- duce an occasional individual bearing a few hairs or a patch of hairs of some other color, or of white.

Among the guinea-pigs kept by the natives near Arequipa, I observed an occasional animal having a rough or rosetted coat. This variety is known to fanciers in Europe and the United States under the name Abyssinian. (See Castle, 1905.) It is said, on the authority of Geoffroy Saint-Hilaire, to have been introduced from Peru into Europe about the year 1872 in a rough-coated, long-haired individual received at the Jardin d'Acclimatation, Paris. In conformity with this account

5

6 INHERITANCE IN GUINEA-PIGS.

it may be said that the rough-coated long-haired variety has ever since its introduction been called by fanciers "Peruvian." I saw no long- haired individuals, cither rough-coated or smooth, among the guinea- pigs kept by the natives at Arequipa, and the short-haired rough-coated one- observed had imperfectly developed rosettes, much inferior to the best standard-bred rosetted Abyssinians of fanciers in Europe and the United States. For this reason I infer that no particular attention was given to this character in the breeding of the guinea-pigs which I saw. though this may very likely have been done in other parts of the country. But the unit-character variation which is responsible for the rosetted condition of the coat in Abyssinian guinea-pigs was plainly represented in the stocks kept by the natives in Arequipa and needed only selection to bring it up to the standards of fanciers.

Eight independent mendelizing unit-character variations had been recognized as affecting the coat characters of guinea-pigs up to this time. Six of these were represented among the four or five dozen guinea-pigs which I actually saw in the cabins of natives, the other two unit characters being (1) the long-haired variation which, as already noted, is said to have been brought originally from Peru to Europe; and (2) the brown variation which first came to the notice of fanciers in England about 1900 and was certainly in existence before that time in the United States, as I can state from personal knowledge. It is uncertain whether or not this last variation had already occurred in Pent and was thence transferred to Europe, but it is certain that all the other 7 had done so, and it is very probable that this also originated in Peru. Further, a ninth wholly independent unit-character variation (presently to be described, viz, the pink-eyed variation) has made its appearance in stocks of domesticated guinea-pigs obtained by me at Arequipa in 1911 and by my colleague, Professor C. T. Brues, at Lima, in 1912. So it is clear that this variation also is widely disseminated among domesticated guinea-pigs kept by the natives in Peru and which have never been in the hands of European fanciers at all.

It can be stated, therefore, with probable correctness, that the guhiea- pig has undergone in domestication more extensive variation in color and coat characters than any other mammal, and that this variation has occurred almost if not quite exclusively under the tutelage of the natives of Peru. This conclusion points either to a great antiquity of the guinea-pig :i- a domesticated animal or to more rapid evolution by unit character variation than by other natural processes.

That the natives do give careful attention to the selection of animals f«»r breeding is shown by the following incident: In the cabin near the observatory, where I first saw guinea-pigs in Peru, and where I ulti- mately s. •cured two pairs of animals, one of which I brought back with me, I observed a very large individual which I desired to purchase, and though other individuals were offered me at a very reasonable price,

GUINEA-PIGS IN PERU. 7

this particular one could not be had because, I was assured, he was the "padre" (sire) of the entire family. Size seemed to be the point especially emphasized in the breeding of guinea-pigs in this cabin, as would naturally be the case when the animals formed the meat-supply of the family, as they do now among the native poor of Peru and doubt- less have done since ancient times.

But the chief object of my journey to Peru was the study not of the domesticated guinea-pigs of the country, but of their wild progenitors. Accordingly special efforts were made to secure specimens of the wild cavy, which Professor Bailey had found to be abundant in the locality. Once or twice, when riding along a road between irrigated fields, I had seen a cavy scurry to cover in a pile of rocks; further, I had observed droppings of the animals in the rocky wall of a cattle corral in an alfalfa field. But how to capture the animals alive was a problem which baffled immediate solution. It seemed likely that the natives would know better how to go about this than I did. Accordingly word was passed around among the near-by villages that a good price would be paid at the observatory for wild cavies, either alive or dead. Within a few hours boys began to arrive with the coveted specimens and for the next week I was kept busy preparing skins and saving bones of the animals which were received dead, or making cages and caring for such as arrived alive. In this way 11 cavies (all I could hope to transport safely) and about a dozen skins were soon secured, and preparations were made for the return journey. In due time the journey was accomplished, and with such success that three new races of guinea-pigs were added to our experimental stocks, viz, (1) a wild species, the probable ancestor of the domesticated guinea-pig, identified as Cavia cutleri Bennett; (2) a feral race from lea, probably identical with that described by Von Tschudi; (3) domesticated guinea-pigs, such as are at present kept by the natives of Peru.

8 INHERITANCE IN GUINEA-PIGS.

HYBRIDIZATION EXPERIMENTS WITH CAVIA CUTLERI. LIFE HISTORY OF CAVIA CUTLERI.

The primary object of my journey to Peru was to secure representa- tives of the wild species of cavy, Cavia cutleri Bennett, known to exist there. Four pairs of these animals captured at Arequipa were suc- fullv installed in cages at the Bussey Institution in January 1913.

One of the males soon died without leaving descendants; the other 7 animals 1 4 females and 3 males) produced offspring in captivity, which have continued to breed succesfully, though the stock has at times been seriously reduced by disease in cold weather. Three generations of descendants have been reared from the original stock of 7 animals. Together they number 100 individuals, of which 47 are males and 53 females. All are very uniform in color, size, general appearance, and behavior.

Their color is a dull leaden gray-brown, well adapted to escape notice amid the arid surroundings of their native habitat. The fur is agouti- ticked and the belly light, but the yellow of the ticking and belly is so pale as to resemble a dirty white or very light cream shade. The color

much paler than that of the Brazilian species, Cavia rufescens, studied by Detlefsen. The fur is also finer and softer, in which respect it ■mbles the guinea-pig. The size of C. cutleri is about the same as that of C. rufescens, and between one-third and one-half that of the guinea-pig. The maximum weight of an adult male is about 525 grams ; that of a domesticated male guinea-pig obtained in Arequipa (d*1002) is nearly three times this amount.

In wildness ( 'aria cutleri is very much like C. rufescens. The animals live contentedly in small cages, 2 feet 6 inches square, but invariably retreat under their box or conceal themselves in the hay if anyone approach

The extreme >avageness toward each other of individuals of Cavia cutlrri makes it difficult to rear large numbers of them in captivity. It i- seldom possible to keep more than a single pair in a cage together for any length of time. Two adult males will not live together peace- ably under any circumstances, and if two females are placed together in a cage with one male persecution of one female by the other usually follows. Even when the young are ahWed to grow up in the same cage with their parents, family dissensions are likely to arise as soon as th<- young become mature.

The period of gestation (minimum interval between litters) averages 3

horterthan in guinea-pigs, being 60 to 70 days, and the number

of young to a littervaries from 1 to4. Fifty-three litters bornin captivity

includ tly UK) young, an average of 1.89 young to a litter. The

si/e of litter occurring most frequently is 2, which has been recorded

CAVIA CUTLERI.

9

Table 1. — Number and size of litters -produced by each mother, Cavia cutleri.

Mother and date of her birth.	Date of litter.	Size of litter.	Mother's age at birth of young.	Days since last litter.
9 2 (caught wild) ; born March 1911 (?) 9 3 (caught wild) ; born May 1911 (?) 9 5 (caught wild) ; born Jan. 1910 (?)	Mar. 5, 1913 June 28, 1913 Aug. 29, 1913 Nov. 4, 1913 May 29, 1912 Oct. 3, 1912 Dec. 26, 1912 July 5, 1913 Dec. 15, 1913 July 12, 1912 Sept. 12, 1912 Nov. 15, 1912 Jan. 22, 1913 Sept. 6, 1912 Sept. 26, 1912 Dec. 10, 1912 Feb. 17, 1913 June 30, 1913 Oct. 1, 1913 Aug. 15, 1914 Dec. 2, 1914 July 5, 1912 Sept. 4, 1912 Nov. 4, 1912 Apr. 25, 1913 June 25, 1913 Aug. 25, 1913 June 17, 1913 Oct. 16, 1913 Aug. 3, 1914 Nov. 2, 1914 Julv 26, 1913 Sept. 20, 1913 Aug. 25, 1914 Nov. 2, 1914 July 26, 1913 Nov. 1, 1913 Dec. 27, 1913 June 25, 1913 Aug. 29, 1913 June 28, 1913 Sept. 4, 1913 Nov. 4, 1913 Jan. 5, 1914 Nov. 1, 1913 Dec. 27, 1913 Mar. 1, 1914 Mar. 15, 1914 May 18, 1914 July 20, 1914 Julv 30, 1913 Dec. 8, 1913 Jan. 12, 1915	2 2 2 2 3 3 1 3 1 3 3 3 2 3 1 2 1 3 2 2 1 1 2 2 1 1 1 1 4 2 2 2 2 2 2 2 2 3 2 2 1 2 1 1 3 2 1 1 1 2 1 1 2	months 24 27 29 31 12 17 20 26 32 18 20 22 24 20 4 5 7 12 15 25 29 10 12 14 7 9 11 8 12 22 25 8 10 21 24 6 9 11 5 7 3f ! 6 4 6 4 6 8 6 8 10 9 14 14	62 67 62 64 68 69 61 61 61 61 56 69 56 65 68 62 64 64 63
9 6 (caught wild) ; born Jan. 1910 (?)
911; May 29, 1912
9 15; July 12, 1912
9 26; Sept. 6, 1912 .
9 27; Sept. 6, 1912
9 36; Oct. 3, 1912
9 42; Nov. 15, 1912
9 65 ; Jan. 22, 1913
9 66; Jan. 22, 1913
9 79; March 5, 1913
9 118; June 28, 1913
9124; June 30, 1913
9129; July 5, 1913
9 184; Sept. 4, 1913
9 224; Oct. 16, 1913
9 241 ; Nov. 4, 1913

LO

INHERITANCE IN GUINEA-PIGS.

M times; litten of I have been recorded 18 times, litters of 3, 10 times, md .i litter of 4 once. Factors which influence size of litter are evi- dently age and state of nourishment of the mother. Table 2 shows the relation of age of mother to size of litter. Very young mothers (age •I months Of lees) have only 1 young at a birth. The females become wallv mature a1 a very early age, as do female guinea-pigs. Well- Dourifthed females may breed at 2 months of age, when they are less than half-grown, full growth not being attained until they are 12 or 13 months old Females over 4 months but under 12 months of age produce usually 1 or 2 young at a birth, rarely 3; those which are 1 or Id produce the maximum number of young, usually 2 or 3, rarely 1 or 4. After the age of 2 years the number of young again

Table 2. — Relation between age of mother and size of litter, Cavia cutleri.

Ag«' -f mother in month-	Size of litters and number of each size.	Age of mother in months.	Size of litters and number of each size.
1 in litter.	2 in litter.	3 in litter.	4 in litter.	1 in litter.	2 in litter.	3 in litter.	4 in litter.
4	3 1 ■i 3 2 1 1	2 3 1 1 1 2			12 to 15 16 to 19 20 to 23 24 to 27 28 to 31 32 Total litters. . .	1 1 1 1	3 1 2 6 2	2 2 3 1	1
6. . . .
:
>
in 11
18	24	10	1

decreases to 1 or 2. The oldest female known to have borne young 4 the original stock) had at the time been in captivity over 2 and her estimated age was 32 months. None of the females horn in captivity has given birth to young at a more advanced age than 29 months. Our records accordingly indicate that females rarely ed after fchey have attained the age of 2\ years. The duration of the breeding period in the case of males is more extended. It is prob- able that males •!<> not attain sexual maturity quite so early as females, f<»r females may breed when less than 2 months old, but we have no denoe that males can breed before they are 3 months old.1 But the capacity t.» breed once attained continues indefinitely. One male * 1 caught wild in December 1911 and estimated then to have been tenths old Lb -till siring young, more than 3 years after his capture, being, it U .--timated, nearly 4 years old.

males are capable of breeding again immediately after the birth i. but if they do bo the number of young at the next birth is

Dtion (o a record from his experiments which shows that a male

light infusbo <.f r blood must have been sexually mature at 2J

- the only record known to me of a guinea-pig male breeding when lesa

CAVIA CUTLERI.

11

apt to be less, or the young will be born smaller and less fully developed (with smaller bodies and shorter hair) , and the period of gestation will be shortened, even to 56 days in extreme cases, the normal period being, as in the guinea-pig, between 60 and 70 days. (See table 4.) If the mother is well nourished and has not borne a litter recently, she is more likely to have a large litter of young. The largest litter recorded (4) was borne by a female 1 year old, which had previously had only 1 young, born 4 months earlier. The recorded date of the birth of each litter of young is given in table 1, together with the interval in days between suc- cessive litters by the same mother, except in cases where the interval is obviously greater than the ordinary period of gestation, and it is to

Table 3. — Relation of size of litter and number of litters to time of year.

Born in —	Size of litters and number of each size.	Total Utters.	Total young.
1 in Utter.	2 in Utter.	3 in litter.	4 in litter.
January	1 1 2 1 1 3	2 1 2	1		3 1 3 1 2 5	5 1 4 1 4 7
February March
April
May
June
Total born in first 6 months1 July
9	5	1	0	15	22
2 1 1 1 4	3 5 3 1 5 2	3 2 1 2 1	1	8 6 6 3 8 7	17 11 13 9 17 11
August
September October
November Total born in sec- ond 6 months2. .
9	19	9	1	38	78

1Average, 1.47 young per litter.

2Average, 2.05 young per litter.

be supposed that the mother did not breed again immediately. The variation in these day intervals between litters is shown in table 4, from which it appears that the gestation period ordinarily continues from 61 to 69 days, with 63.3 days as an average. However, the periods as recorded can not be relied upon as accurate, except within limits of about 2 days, for the cages were not inspected daily, but only once or twice a week, and when young more than 24 hours old were found in a cage, the estimated age of the young may differ from the true age by 1 or 2 days. Young less than a day old are readily recog- nized as such by the condition of the umbilical cord.

The 4 original wild-caught females have a somewhat better record of productiveness than their descendants reared in captivity, which indi-

12

INHERITANCE IN GUINEA-PIGS.

rates that laboratory conditions of close captivity are not as favorable for full growth and vigor B8 the freer life and better air of the original habitat. The4 wild-caught females produced 33 young in 14 litters, an average of 2.36 young to a litter. Their daughters or granddaughters, reared in captivity, when of like age, have produced 27 young in 13 litters, an average of 2.07 young to a litter.

Too much emphasis must not be laid, however, on this difference, because productiveness depends largely on food, care, and weather con- ditions, and it is not certain that these were equally favorable for the original females and for their descendants, respectively.

Table 1 shows for each mother how many litters of young she has borne, at what age she bore them, and how many young were contained in each litter. In the case of the 4 females caught wild, the age given for the mother is of course not known; the age recorded is an estimate based on the size of the mother when captured.

Table 3 shows in what month each litter of young was born and what it- size was. This table brings out rather strikingly the effect of the ions and consequent character of food available upon the size and number of the litters.

Taule 4. — Variation in period of gestation {interval since previous litter) in Cavia cutleri. Average, 63.3 days.

Days		Days
between	Cases.	between	Cases.
litter.-.		litters.
56	2	65	1
61	4	67	1
62	3	68	2
63	1	69	2
64	3

In the 6 months from July to December inclusive, litters were born which were conceived under summer conditions, with an abundance of green food available. It will be observed that in this half of the year the litters are numerous (38) and large (average 2.05 young to a litter). The young born in the (i months January to June inclusive were con- i under winter conditions, when the mothers were subsisting largely on a diet of dried or concentrated foods, with a limited amount of green food available. In this half of the year the litters are less numerous 15 and smaller (average size 1.47 young).

uper.it i, re probably does not directly affect the result, as the mak were kept in a he.-, ted house, but purity of the air may possibly the house is much better ventilated in the warmer months. - probably the most important factor, as the condition of the mak changes promptly with change of food, even when other condi- tion- -how no ehani

CROSSES OF CAVIA CUTLERI. 13

CROSSES OF CAVIA CUTLERI MALES WITH GUINEA-PIG FEMALES.

Crosses have been made only between male Cavia cutleri and female guinea-pigs. The reciprocal cross was not undertaken, because the number of cutleri females on hand at any one time has been insufficient and because it seemed probable that a cross with the much larger guinea-pig would be fatal to the cutleri females because of the probable large size of the hybrid offspring. Two races of guinea-pigs were employed in the crosses, these being the purest races available, the genetic properties of which had been long and thoroughly tested.

The race most extensively used may be called race C. It consisted of "brown-eyed cream" individuals or of albinos borne by brown-eyed cream parents. The results of crosses of colored and albino individuals of race C will be described separately. The other race may be called race B. It consisted of intensely black-pigmented individuals or of albinos produced by such black individuals. The results of crosses with the two sorts of individuals will be described separately. A cutleri male bred in captivity (cf 78) was mated with black females of race B, and produced 9 F: young, all colored like C. cutleri, but darker, the ticking of the fur being brick red or yellow instead of creamy white as in cutleri.

Albino females of race B were mated with the same cutleri male (d"78) or with cf 114, another cutleri male reared in captivity, or else with cf 4 or cf 8, which were original cutleri males caught wild. Such matings produced 39 Fx young, all with the same (golden agouti) type of coloration as the young produced by the black mothers.

Females of race C were mated only with the two wild-caught cutleri males (r?4 and c?8). The cream-colored mothers of race C produced 34 young, all golden agouti in color like the young derived from race B crosses, but much lighter. They were, however, darker in color than C. cutleri, the agouti ticking being yellow or reddish, not creamy white as in cutleri. (See plate 3.) Albino females of race C produced by the cutleri males 14 young, indistinguishable in appearance from the young produced by their cream-colored sisters.

The Fi hybrids, whose total number was 96, were all vigorous and large, their adult size nearly or quite equaling that of guinea-pigs. They grew with great rapidity and have proved fully fertile inter se. In wildness and ferocity they are intermediate between the parent races.

COLOR INHERITANCE AMONG THE F2 HYBRIDS.

(a) CROSS 9 ALBINO (RACE B)Xc? CUTLERI.

By breeding inter se certain of the F: hybrids, from the cross 9 albino, race B, X cf cutleri there has been produced a second (or F2) generation of hybrids, which number 75 individuals. As regards color, disregarding minor differences of intensity of pigmentation, these hybrids fall into three classes: golden agouti, black, and albino. Of the

It

INHERITANCE IN GUINEA-PIGS.

utia then are 13, of the blacks 15, and of the albinos 17. The DumericaJ relations ol the classes suggest a dihybrid Mendelian ratio 0| g i, which is in entire agreement with existing knowledge of odor inheritance in guinea-pigs (Castle, 1905; Sollas, 1909). C. cutleri i- evidently homozygous for all Mendelian color factors, since it breeds very true to color. Albino guinea-pigs from a black race are known to possess two independent recessive modifications from this condition, lacking both the agouti factor and the so-called color factor. As

u-ds these factors, then, the wild race, cutleri, forms gametes AC, the albino forms gametes ac, and the Fx hybrids form gametes of the four types \t '. Ac, aC\ and ac. From recombination of such gametes Bhould arise in F2 zygotes as in table 5.

Table 5.

1 AACC

\:iCC

2 A A' !c 4 AaCc

■'>uti

1 aaCC

2 aaCc

3 black

1 AAcc

2 Aacc

3 albino

1 aacc

1 albino

Th' ral kinds of albinos being similar in appearance, the expected

result is 9 agouti. 3 black, 4 albino. The agreement with this expecta- tion is fairly close (see table 6).

Table 6.

	Agouti.	Black.	Albino.
Observed Expected	43 42.19	15 14.06	17 18.75

(b) CROSS 9 ALBINO (RACE C)Xt? CUTLERI.

Ft anini.il> from the cross between an albino of race C and a cutleri male have produced 11 F2 young, which fall into 7 color classes, dis- regarding differences of intensity of pigmentation. These classes and their numerical representation among the 44 young are as follows: golden agouti, 10; black, 1: cinnamon, 8; black-eyed cream, 4; brown- ing; chocolate, t; albino, 14. (Seeplate4.) The occurrence of thee eral classes of 1", young is what previously existing knowl- ; color inheritance among guinea-pigs would have led us to expect, for it was known thai albinos of race C differed from agoutis in the same two factors as the albinos of race B, viz, the agouti factor and the color factor. In addition, the albinos of race C were known to differ from lUtifl in two other factor-, -eon respectively in chocolate and yellow The chocolate race may be considered to have arisen by a recessive modification of the black factor B, and the yellow race by a similar modification of the extension factor E. Accordingly this cross

CROSSES OF CAVIA CUTLERI.

15

was supposed in advance to involve 4 independent Mendelian factors, a supposition which the observed result justifies. The factor differences in the two races are: gametes of cutleri, ABCE; of albino (race C), abce. On this hypothesis the Fi hybrids should form 16 different kinds of gam- etes, the color potentialities of which are indicated in parentheses:

1 ABCE (agouti).	7 AbcE (albino).	12 Abce (albino).
2 aBCE (black).	8 abCE (chocolate).	13 aBce (albino).
3 AbCE (cinnamon).	9 aBCe (black-eyed yellow) .	14 abCe (brown-eyed yellow)
4 ABcE (albino).	10 aBcE (albino).	15 abcE (albino).
5 ABCe (black-eyed yellow).	11 AbCe (brown-eyed yellow).	16 abce (albino).
6 ABce (albino).

From this list it will be observed that 2 different gametic factorial combinations are capable of producing black-eyed yellow, and that the same is true concerning brown-eyed yellow, while 8 different com- binations contain the potentialities of albinos. From these considera- tions it follows that the F2 ratio will be peculiar, since each of the yellow classes that can be distinguished from each other (black-eyed and brown-eyed) will itself be composite, and the same will be true of the albino class. The expected classes and their proportional frequencies will accordingly be :

Golden agouti 81

Black 27

Cinnamon 27

Yellow (black-eyed) 36

Yellow (brown-eyed) 12

Chocolate 9

Albino 64

A cross involving the formation of so many classes of individuals can not be expected to show very satisfactory Mendelian ratios in so small a number of offspring as 44. All the expected classes are represented, although black is represented in a single individual only.

The colored individuals of race C were known in many cases to carry albinism as a recessive character. The albino gametes of such indi- viduals would, in crosses with cutleri mates, form the same kind of zygotes as the albinos of race C, which were used in the cross just described. In considering the results of such crosses, it is therefore proper to include in one category Fi animals derived from both sources. If this is done the F2 young are increased to 108, distributed as shown in table 7, the expected theoretical number in each class being shown in a parallel column.

Table 7.

	Observed.	Expected.		Observed.	Expected.
Golden agouti Black	33 7 13 13 8	34.17 11.39 11.39 15.08 5.07	Chocolate	7 27	3.80 27.00
Black-eyed yellow .... Brown-eyed yellow. . . .	108	108.00

L6

INHl.l.MI \\« I. IN GUINEA-PIGS.

(c) CROSS 9 BROWN-EYED CREAM (RACE C) X c? CUTLERI.

Tl,.- ]• animals produced by crossing brown-eyed cream females of race ( « ith cvtt ri males themselves produced 132 F2 young, distributed among the ■«"»" ~ classes as the albino cross had produced (table 8).

Table S.

	rved.	I oted.		Observed.	Expected.
ek	.-■: 17 10 18 8	48 9 1 16.31 16.31 21 76 7 . 26		5 16	5.44
132	116.00
allow. . . . Brown-cQ ad yellow

1 Kcepl in regard to albinos, the result expected from this cross is the M»itf as that expected from crossing albinos of race C. Accordingly the albinos may be for the moment disregarded. The expectation as regards the colored classes of young may then be stated as shown in the column of "observed" results.

The occurrence of 10 albinos in this F2 generation shows that certain of the Fj pairs were heterozygous for this character, which they obvi- ously derived from the brown-eyed cream parent, not from the cutleri parent; for the brown-eyed cream animals of race C were known in many eases to be capable of producing albinos, whereas the cutleri stock bred true Accordingly such pairs of Fx animals from this cross as pro- duced albino young should be tabulated with the hybrids produced by crossing albinos of race C with cutleri males. If this is done there remain 68 instead of 116 F2 young to be considered. Among these are 3 albino- which it is impossible to transfer to table 7, because it is not known what colored individuals were born in the same litters with them. They were born in a pen containing 2 females which had young simul- taneously, one of which was known to produce albino young, though the other did not. Omitting the 3 albinos, there remain 65 colored young, distributed as Bhown in table 9.

Table 9.

		oted.		Observed.	Expected.
	.1 11 III	27 i- 0 1 1 i i 12 19	Brown-eyed yellow ( Ihocolate	3 2	4.06 3.05
	65	65.00

(d) RESULTS FROM (b) AND (c) COMBINED.

e the expectation is the Bame as regards the relative proportions

al colored classes in all crosses of race C females (whether

albmo or colored with cutleri males, we may legitimately combine all

CROSSES OF CAVIA CUTLERI.

17

the F2 results, omitting only albinos (which have been dealt with already). When this is done we get the results shown in table 10. No class deviates from expectation enough to suggest "linkage" or " coherence" of characters involved in the cross.

Table 10.

	Observed.	Expected.		Observed.	Expected.
Golden agouti Black Cinnamon Black-eyed yellow ....	67 18 18 23	61.59 20.53 20.53 27.37	Brown-eyed yellow. . . . Chocolate	11 9	9.13 6.85
146	146.00

(e) INTENSITY AND DILUTION AMONG THE HYBRIDS.

It has been stated that the Fi young produced by the cross of female albinos of race B with male cutleri were dark golden agouti in color, much darker than the cutleri parent. This darkening of the color per- sisted undiminished into the following generation (F2). Of the 58 colored Fi young derived from this cross none was as light in coloration as the cutleri grandparent. Hence it would appear that the darker coloration introduced by the cross, apparently through the albino parent, does not behave as a simple Mendelian character either domi- nant or recessive; otherwise pale-colored F2 young should have been produced. Whatever factors, Mendelian or otherwise, are responsible for the darkening of the pigmentation are evidently unconnected with the so-called color factor, since they are transmitted by albinos, which lack this factor.

A very different result was observed in crosses of the same cutleri males with females of race C. The colored animals of race C are very pale cream-colored. The Ft young which they produced showed a more intense yellow than either parent, but were much lighter than the hybrids produced in the cross with race B albinos. (See plate 3.)

Among their F2 young appeared some very light-colored individuals, 16 being recorded in a total of 56 young produced by pairs which pro- duced no albinos. The pale-colored young were not confined to any one colored class, but were recorded among the agoutis, blacks, cinna- mons, and yellows. (See table 11.) The proportion recorded is close to one-fourth, from which it would seem that dilution had been intro- duced as a recessive character by the cream guinea-pig grandparents. Since, however, C. cutleri is relatively pale in pigmentation, it is prob- able that some of the animals classified as pale were not "dilute," owing to a factor derived from the guinea-pig ancestor, but because of condi- tions derived from the cutleri ancestor. This statement applies to the young of matings which produced albinos as well as to those which did not. The significant thing is that more pale-pigmented young and those which excelled in paleness were obtained from those matings which did not involve albinsim.

-

IMIKHITANCE IN GUINEA-PIGS.

.,. n DMtAmMm -..ty and dilution among the F2 young derived

from tht cross 9 race CX d'cutleri.

oung.

I

( iiin.'tnion

BUol How .

Brow in jr*d jreltow

la U

Total

in pain pro- ducing no albinos.

Intense.

5

• >

5 2 1

40

Dilute.

8 3 8

2

1G

From pairs pro- ducing albinos.

Intense.

2S

6

13

10

6

7

70

Dilute.

4 1

4 2

11

It was expected thai albinos of race C would produce a much larger

rtion of pale-colored grandchildren, but strange to say this expec-

od was n- >i realised; 81 F2 colored young produced in matings

which yielded albinoe ahowing that the guinea-pig characters had been

eived through albino gametes) included only 11 pale-colored young,

of these is recorded as being paler in color than the cutleri

grandparent. It would appear, therefore, that the albino gametes of

I mothers do not transmit the dilution seen in the cream-colored

animal- of race ( . This would be a puzzling state of affairs had not

Wright 1915) already discovered an easy explanation for it, viz, that

dilution of the cream race is an allelomorph of albinism, and so can

be transmitted in the same gamete with albinism.

I mparing the I', hybrids derived from race C crosses with those

ived from race B crosses, it is certain that the pigmentation of both

darker than that of wild C. cutleri, but the intensity of the race B

hybrid- much exceeds that of the race C hybrids. Among the race B

hybrid- do evidence can be discovered of segregating Mendelian inten-

among the race ( ' hybrids dilution segregates as a simple

Mendelian recessive, precisely as dors albinism, but apparently no

garnet.- transmits both dilution and albinism, for the reason that they

alternative conditions of the same factor. Aside from the factorial

difference in dilution, how does race B differ from race C? Apparently

in no simple factorial way, but iii a genera] way as regards energy of

pigment production, in which hybrids of both races surpass C. cutleri

if differ quantitatively from each other. No Mendelian explanation

• Si difference is at present justified by the observations made.

0) SIGN If 1 OF THE RESULTS OBSERVED.

rmplete fertility of the hybrids produced by crossing wild Cavia

i with the guinea-pig is in striking contrast with the sterility of

bride between C. rvfi and the guinea-pig, as observed by Det-

CROSSES OF CAVIA CUTLERI. 19

lefsen. This indicates that C. cutleri from Peru is the actual wild ancestor of the guinea-pig or closely related to that ancestor. Since, however, Nehring has reported that C. aperea (from Argentina) also produces fertile hybrids with the guinea-pig, it seems likely that these two species are closely related to each other and might interbreed freely if their respective ranges were not completely separated. It seems possible also that both species have contributed to the production of the domesticated form, or that still other species have shared in producing it. Further observations are needed to clear up this matter.

It is evident that the mendelizing unit-character differences, which distinguish one variety of domesticated guinea-pig from another, also exist between guinea-pigs and the wild Cavia cutleri. They are inherited in precisely the same way among the hybrids produced by crossing guinea-pigs with C. cutleri as in crosses of one variety of guinea-pig with another — that is, they mendelize. It is evident that these variations have arisen by a process of retrogressive or loss variation. For example, in the matter of color varieties such as black, brown, yellow, and white, which (in relation to the parent form) are known to breed true without exception, it is evident that these have arisen by loss (or retrogressive modification) of physiological processes which occur in the wild species, since crosses with the wild form bring them back in a heterozygous state, after which they continue to form all possible permutations and recombinations with each other. Thus albinos of race C (which breed true inter se and without crossing with some other variety could pro- duce no other sort) if crossed with C. cutleri (which also breeds true) produce in F2 a definite series of color varieties. This series includes all the color varieties of guinea-pigs more commonly known, such as (1) golden agouti, (2) black, (3) cinnamon, (4) chocolate, (5) black-eyed yellow, (6) brown-eyed yellow, and (7) albino.

The mode of origin of the color varieties of guinea-pigs (and by inference of other domesticated animals also) is therefore clear. These varieties have originated by loss variations or loss "mutations." Is this the means by which species themselves originate? Many biolo- gists have recently advocated this view, as, for example, Lotsy, Baur, and Bateson, but the present case affords rather strong evidence against it. The color varieties of guinea-pigs differ from Cavia rufescens and C. cutleri (undoubtedly distinct species) by the same mendelizing color-factors, but there is no evidence that these two species differ from each other by any color-factor. The two wild species are probably distinct enough to show interspecific sterility, since one is known to form sterile hybrids, the other fertile hybrids, in crosses with the guinea-pig. Their specific distinctness accordingly can not be due to such mendelizing factors as distinguish one domesticated variety from another, but to something more fundamental in character, though less striking in appearance.

QfHUUl kN( B in GUINEA-PIGS.

HYBRIDIZATION EXPERIMENTS WITH A RACE OF FERAL GUINEA PIGS FROM ICA, PERU.

ORIGIN AND CHARACTERISTICS OF THE ICA RACE.

Von Tschudi in 184 L, in bis Fauna of Peru, described, under the name

B wild cavy found occurring in great numbers in the

II, Bays thai the natives call it "cuy del monte," the

. v. and regard it as the original of C.cobaye, the guinea-pig.

Subeequenl writers carefully distinguish the C. cutleri of Von Tschudi

from thai of Bennett, with which my wild cavies from Arequipa agree.

the objects which I hoped to accomplish by the trip to Peru was

to Irani more aboul the cavy which Von Tschudi reported as occur-

ringal lea, and, if possible, to determine its relation to C.etriUri Bennett

and to the guinea-pin.

Through the kindly interesl of Messrs. W. R. Grace & Co. I was able I ire 3 wild-caughl cavies (a male and 2 females) from lea

; to bring them back with me to the Bussey Institution, where they • ■ produced a numerous progeny.

animals were about the Bize of domesticated guinea-pigs, were very timid, and were -elf-colored golden agouti, in every respect similar in appearance to tame guinea-pigs of the color variety named.

animals brought from lea produced 7 golden-agouti young,

all similar to the parents in color, except that one bore a spot of red,

firsl observed indication of contamination of the stock with char-

found in domesticated guinea-pigs. That other indications

re no1 observed in this first mating of the animals was probably due

to the fact thai the male was homozygous for all other color factors, as

subsequent matings of the females with a son of one of them by the

original male proved thai both mother and son were heterozygous in

that variation of the color factor which is seen in "red-eyed" guinea-

pi- I istle, 191 1: Wright, 1915). The same matings with the son

ible 12 proved that one of the two original females (9 503)

tlso heterozygous in the agouti factor and transmitted white-

otting, since she produced a black daughter which had one white foot.

Three other inbred descendants of the original trio of lea animals have

if white; two of them in addition bore spots of red,

tricolors. One of the original trio of animals from lea

when mated with '.">n:>, produced a son (c?575) which was

tly rough-coated. Accordingly we have clear evidence that

the stock derived from [ca was contaminated with at least 5 of the

ily independent unit-factor variations which occur among

doi d guinea-pigs and there can belittle doubt that it really has

1 wholly or in part from domesticated guinea-pig ancestors.

GUINEA-PIGS FROM ICA. 21

Table 12. — Young produced by the three original lea guinea-pigs or their inbred descendants.

A. Both parents golden agouti, one only heterozygous for red-eye (indicated by *).

Father.

501 *505 *505 *505 *533

*533

Mother.

*502 and *503 509 510 530 509

625

Total .

Golden agouti young.

7 4 1 3

7

24

Black young.

One young with spot of red.

One young spotted with red and with white.

B. Both parents golden agouti and heterozygous for red-eye.

Father.

505 505 505 505 505 533 533

Mother.

502 503 504 507 605 529 540

Total .

Golden agouti young.

10 9 1 8 1 5 4

38

Silver agouti young.

4 2 4 5 1 1 1

18

Black young.

One slightly rough.

One black with white foot.

One spotted with red and with white.

One spotted with white.

C. Both parents silver agouti (red-eyed).

Father.

565

565 565 565 565 565

Mother.

527 528 573 593 601 and 604 607

Silver agouti young.

5 8 1 2 3 1

Father.

569 602 79S 798

Mother.

587 and 588 60S 701 872

Total.

Silver agouti young.

31

The question at once arises whether the stock obtained by me from lea was really a feral stock, in origin like the animals described by Von Tschudi, or whether they were present-day domesticated animals concerning whose origin I was deceived. Since I did not myself see the animals captured or see similar animals running at large and did not even visit lea, I can make no positive statement as to their feral origin, but I believe the report made to me by the agents of W. R. Grace & Co., that they were caught feral in the neighborhood of lea, to be correct for the following reasons: (1) The animals were placed

INHERITANCE IN GUINEA-PIGS.

board our steamer during a stop made in the night at Pisco, the the abort line of railway which leads down from lea to the I found them the next morning in the "butcher-shop," con- signed from lea to W . EL ( brace & Co. in Callao. I conclude that they really did come from the ueighborhood of lea. (2) I saw no domesti- ted guinea-pigs in Peru which were self-colored like these animals. All domesticated ones which I saw in Peru, except albinos, were spotted with white, or with yellow, or both. Self varieties are not fancied in P< pu. \ arieties of this aort are not uncommon among the guinea-pigs kept by European and American fanciers, but apparently they have bushed only by careful and long-continued selection from the 1 Meek i iriginally introduced from South America. (3) The spotting ..ugh character which have cropped out as recessives among the descendants of the three lea animals are feebly expressed characters which appear to have been almost obliterated, but which still come to nor feebly under inbreeding. This is what we should expect to find in a feral race acted upon by natural selection, conspicuous nations like spotting tending to disappear.

Tabu 13. —	"an >.	pure lea animals irhose malings are recorded in table 12.
1		ther.	Individual.	Father.	Mother.	Individual.	Father.	Mother.
			9527	505	507	9 601	565	527
9-			9 528	505	503	c?602	565	527
9'-'			9ol".i	505	503	9 604	505	503
9504	601	r 503	9530	505	503	9 605	505	503
</'■	601	602 0 r 503	d"533	505	544	9 607	565	528
9.-	.Mil	:.< 'J	9540	505	509	9 608	565	528
-.09	601	603	cr565	505	507	9 625	533	509
10	.'."I		cP569	505	504	9 701	565	601 or 604
			9573	505	502	6*798	569	587 or 588
			9 5s7	505	507	9 827	798	701
			9 5s^	505	507
			9593	505	504

'( hrigina] stock.

The 3 original lea animals or their inbred descendants mated inter se

>■ produced 1 14 young, of which 62 have been golden agoutis, 49 sil-

routis, and 3 blacks; 4 of the 114 have shown a small amount

of white spotting, '1 have shown yellow spotting, and 1 has shown a

ill amount of roughness of the coat.

'1 he various matingn which have produced these young are classified

in three group* in table 12, and the parentage of each animal which

I pari in a mating i- shown in table 13, from which pedigrees may

lily be drawn tracing back to the original trio. It will be observed

m table L2 that silver agouti was derived from golden agouti as a

and has bred true without exception (31 silver agouti young

produced by silvr agouti parents).

GUINEA-PIGS FROM ICA.

23

CROSSES BETWEEN THE ICA RACE AND GUINEA-PIGS OF RACE C.

An albino male guinea-pig (cf 54) of race C was mated with 5 golden agouti females of the lea stock. It was hoped from this cross to learn as promptly as possible the gametic composition of the lea race, since race C contained a larger number of recessive Mendelian factors than any other race in the laboratory. In this hope we were not disap- pointed. Race C has already been described. It contains two different recessive variations of the color factor, dilution and albinism, which are allelomorphic with each other and with ordinary color, thus forming a system of triple allelomorphs, C, Cd, and Ca, with dominance in the order named (see Wright, 1915). It lacks agouti, black, and extension factors. Visibly the animals of this race are either brown-eyed cream or albino. Male 54 was an albino, bearing the color allelomorph Ca, which is recessive to the color allelomorph Cd found in brown-eyed cream individuals of race C. The mating between cf 54 and the 5 golden agouti females of the lea race produced 13 young, 7 of which were

Table 14.

-Fi result of mating the albino &5J+ of race C with golden agouti females of the lea race.

Mother.	Dark-eyed young. Golden agouti.	Red-eyed young.
Silver agouti.	Sepia.
502 503 504 507 509 Total . . .	3 1 1 2	1 1 X 1	1 2
7	3	3

golden agouti (like the mothers), 3 silver agouti, and 3 a dull black or slate color, which will be called sepia. The silver agouti young were like those produced by lea animals bred inter se. The sepia young represented a new class not previously observed. In common with the silver agoutis they had no yellow in their fur. The ticking and spotting of silver agoutis was of white, as was also the spotting of the sepias, which had no ticking. It seemed probable, therefore, as proved to be the case, that the silver agouti and the sepia young differed from each other only in the presence or absence of the agouti factor. But these two classes of young taken together differ from golden agoutis in lacking yellow pigmentation with which the golden agouti fur is ticked. They also differ from golden agoutis in the intensity of the eye pigmentation, which is very great in golden agoutis and blacks, but ordinarily shows such reduction in silver agoutis and sepias that the eye by reflected light has a deep red glow. It will be convenient to distinguish them as red-

j, IMIKKII W I IN i.l'INEA-PIGS.

,1 ,, being undent 1 thai the red eye is invariably associated with

low in the ooat,

I ,ur of the five [ca mothers which were mated with ^54 had pro-

duced silver agouti red-eyed) young by lea mates. Each of them

pril.j,. l-eyed young by o*54; 1 1 >ge1 her they produced 5 dark-eyed

■ . .l.l.-n agouti) and 6 red-eyed (silver agouti or sepia). The

fifth fca mother 509) had produced 11 golden agouti young when

mated with [ca males known to be heterozygous for silver agouti.

able 12. This is good evidence that she did not carry red-eye as

., character and was accordingly homozygous for dark-eye.

By '64 she produced 2 golden agouti young.

From these several tacts it appears that dark-eyed lea animals capable of producing red-eyedyoung when mated inter se, produce equal Qumbers of dark-eyed and red-eyed young when mated to albinos, but produce qo albinos. This indicates that albinism is recessive both to red-eye and dark-eye, an indication which the F2 result confirms. It will be shown further that the three conditions are mutually allelomor- phic, bo thai a sygote may contain any two of the three, but not more.

d-eye is in fad a fourth member of the albino series of allelomorphs, which include- the following conditions in order of dominance: (1) ordi- nary dark-eye and colored coat, such as is seen in Cavia cutleri and in golden agouti animals of the lea race; (2) dark-eye with dilute coat,

d in colored animals of race C; (3) red-eye and non-yellow coat; l albino. (See Wright, 1915.) For convenience these allelomorphs may be designated by C, C<], Cr, and Ca. The cross of lea females with the albino o*54 involves animals of the formulae CC or CCr mated with an animal of the formula CaCa. The 7 golden agouti young are expected to be of the formula CCa; the 6 red-eyed young of the formula I I We may now compare the experimental with the expected results of breeding such animals in various ways.

THE F, GENERATION.

of the Fj silver agouti males (cf 517) was known from his pedi- • be heterozygous in four characters, viz, red-eye vs. albinism, non-agouti, black vs. brown, and extension vs. restriction. Hie formula was accordingly CrCaAaBbEe, and we should expect him form gametes of 16 different sorts, all equally numerous. This animal was mated with all three kinds of F! females, with the results shown in tabic 15. The golden agouti females produced 25 young, distributed among 10 classes very distinctly different in appearance. -outi females were known from pedigree to be hetero- for the Bame 4 factors as c?517, but to contain a different allelo- morph for albinism. Both he and they carried albinism as a recessive • but whereas he carried red-eye (Cr) as its dominant allelo- rph, they carried dark-eye it he ordinary condition of the color

GUINEA-PIGS FROM ICA.

25

factor, viz, C). His gametes accordingly could transmit either Cr or Ca, but theirs would transmit either C or Ca. Accordingly their young should be in the ratio 2 dark-eyed to 1 red-eyed to 1 albino. The observed numbers were 13:8:4. Each of these three groups might theoretically contain 8 different kinds of individuals, but certain of these would be visibly indistinguishable. The classes visibly different which might themselves be expected to be composite are black-eyed reds and brown-

Table 15. — F<i young from the cross c?54 albino (	race C) X gola	\en	agouti females of lea race
Parents.	Dark-eyed young.
Golden agouti.	Black.	Cinna- mon.	Choco- late.	Black- eyed red.	Brown- eyed red.
d1 517 silver agouti X 9 golden agouti .... Expected	6 5.3	1 1.8	2 1.8	0 0.6	2 2.3	2 0.8
6*517 silver agouti X 9 silver agouti Expected
. . .
cf 517 silver agouti X 9 sepia
Expected
Total observed
6 5.3	1 1.8	2 1.8	0 0.6	2 2.3	2 0.8
Total expected
Parents.	Red-eyed young.	Albinc young	Total.
Silver agouti.	Sepia.	Silver cinna- mon.	Red- eyed choco- late.	Red- eyed white.
0*517 silver agouti X 9 golden agouti Expected	3 2.6	2 0.9	2 0.9	1 0.3	0 1.5	4 6.3	25
0*51 7 silver agouti X9 silver agouti Expected
0 2.8	1 0.9	1 0.9	0 0.3	4 1.7	3 2.3	9
0*517 silver agouti X 9 sepia
5 7.4	5 7.4	1 2.4	1 2.4	8 6.5	15 8.7	35
Expected
Total observed
8 12.8	8 9.2	4 4 2	2 3 n	12 9.7	22 17.3	69
Total expected

eyed reds, each of which might include both agouti and non-agouti animals; also red-eyed whites, which might have either black or brown pigment in the reddish eye, and might transmit either agouti or non- agouti; and finally, the albinos, which might be of as many different sorts as the colored classes, viz, 16.

All except 2 of the 12 visibly different classes expected from this mating were obtained in as small a total number of young as 25. In the two missing classes, chocolate and red-eyed white, the theoretical

IMII.IMIAME IN GUINEA-PIGS.

majbm men only 0.6 and 1.5 individuals respectively, so that their

...a- n<>t surprising-

outi p females, cf517 had 9 young of 4 different color the maximum Dumber of classes expected being 6. «pia T females, cf517 had 35 young, distributed among 6 dif- ,Vnnt color classes, as expected. Summarizing the results from all three kinds of mating*, we 6nd that the F2 young of c?ol7 number 69, distributed among 1 1 of the 12 expected classes of young, the missing rh- l.ri.m one in which the expectation is for 0.6 of an individual, lv more than an even chance for the production of such an indi- vidual in the Dumber of young recorded. (See table 15.)

I >A , , , ] ft— Young product d by red-eyed white parents mated inter se.

		Red-eyed	Albino
Father.	Mother.	white young.	young.
567	571	2	1
.VT<>	726	10	0
77 1	747	5	2
771	758	3	3
771	775	8	0
B42	571	2	0
B49 T	851 otal	2	0
32	6

A word as to the number of classes expected may not be out of place.

The dark-eyed classes expected are 6, identical with those expected

from the cross of race C animals with wild Cavia cutleri. (Compare

p. I I The nui nber of classes expected among the red-eyed young is

1 leas, namely 5, because red-eyed whites which have brown pigment

in the an not be distinguished (except by breeding-test or post

rom those which have black pigment in the eye, the quantity

• present being too small, and the coat in both cases white.

the whole the agreement between expected and observed in this

i n nt is so good as to preclude the idea that any coupling or

associatioD occurs among the 4 unit factors involved in the cross.

This experiment produced 4 color varieties of guinea-pig previously

unknown to me, viz. the 4 red-eyed classes other than silver agoutis,

which had already been obtained from the uncrossed lea race. (See

plat.- 1. _\ and 5. The eve has a similar appearance in all the red-

Bhowing a deep-red glow by reflected light. The silver

■iiti variety, as already explained, differs from golden agouti in the

if the fur. which is white in silver agouti, instead of red or

Aden agouti. Sepia as compared with black has a more

fad.d appearance, approaching chocolate on the sides of the body and

belly, but always darker and unmistakably black above. Silver cinna-

GUINEA-PIGS FROM ICA.

27

mon (or "red-eyed cinnamon," plate 5, fig. 31) differs from silver agouti in having brown hairs ticked with white instead of black hairs ticked with white. It is one of the handsomest of guinea-pig varieties. Red- eyed chocolate is indistinguishable from dark-eyed chocolate, except in eye color. The red-eyed whites all look alike, though they may differ considerably in factorial composition. Their production in this experi- ment was a complete surprise to us and very puzzling until the sugges- tion was made (I think by Mr. Wright) that an essential feature of the red-eyed variation was the absence of yellow color from the fur. It was then realized that a " yellow" animal with red eyes and "non- yellow" fur must of necessity have white fur. This suggestion was immediately put to the test by mating the red-eyed white d"576 with 3 dark-eyed cream females. They produced 12 young, of which 5 were brown-eyed cream, 2 black-eyed cream, 3 red-eyed white, and 2 albino. No young were produced which had coats of any other color than yellow! Hence it is clear that red-eyed whites do not transmit the extension factor.1

Table 17. — Results of mating red-eyed white individuals with albinos.

Father, red- eyed.	Mother, albino.	Red-eyed young.	Albino young.
Silver agouti.	Sepia.	Silver cinnamon.	Choco- late.	White.
567 567 567 567 567 576 576 576	564 568 572 177 711 1430 1439 1446		1 2	1	1 1	4 1	4 2 2 7 1 1 2
T	otal	3	3	1	2	5	19

This same red-eyed white c?576 was also mated with 3 albino females of race B, which carry the extension factor. Both parents, it will be observed, were white, one having red eyes, the other pink eyes. This mating produced 7 young, of which 3 were red-eyed with silver- agouti-colored coats and 4 were albinos. The production of colored young in this case shows that red-eyed white animals may transmit all that is necessary for the production of a colored coat except the exten- sion factor, which the albino parents supplied.

The red-eyed white d"576 was evidently heterozygous for the black factor, since, when he was mated with brown-eyed cream females, he produced both black-eyed and brown-eyed cream young. Another

xAs a further test of red-eyed whites, two other red-eyed white males (615 and 616) were mated with several different red or yellow coated females. They produced 9 red or yellow young, 5 red- eyed young, and 5 albino young, a result completely in accord with that given by c? 576.

2S

INHERITANCE IN GUINEA-PIGS.

red-eyed white male, 507, an F3 descendant of the albino d"54, race C, was found to be homozygous for brown (table 17). What pigment his eyes contained was undoubtedly brown, for w7hen he was mated with 3 albino females descended from the albino cf 54, race C, he produced 1 silver cinnamon and 2 red-eyed chocolate young, besides 5 red-eyed white and 8 albino young. The entire absence of black-colored young indicates that this male, as well as his albino mates, transmitted the capacity to form brown but not black pigmentation. When, however, this same male (567) was mated with an albino derived from race B, which never produces brown individuals, there were obtained 3 sepia- colored young with red eyes, besides 7 albinos, showing that when the mother transmitted black, this male produced black-pigmented young, black being dominant over brown which he himself transmitted.

Table IS. — Results of mating a red-eyed white male xoith brown-eyed cream females.

Father.	Mother.	Young.
Black-eyed cream.	Brown-eyed cream.	Red-eyed white.	Albino.
576 576 576 842	-46 M250 762 870	1 1	2 1 2 1	1 2	2 1
Tot	al	2	6	3	3

Both the males wrhose matings have just been described, viz, 567 and 576, were heterozygous in albinism, since when mated wdth albinos they produced about 50 per cent of albino young. They were evidently of the formula CrCa. If red-eyed white animals of this formula should be mated with each other we should expect individuals to be produced which are homozygous for red-eye, i. e., are of formula CrCr. Twro probably homozygous red-eyed females of this sort have been discov- ered in mating red-eyed white animals inter se. One of them (9 726, table 16) produced 10 young, all red-eyed white, in matings with c?576, known to be heterozygous for albinism. Had this female formed albino gametes she should have produced 25 per cent of albino young in the matings mentioned. It seems probable, therefore, that she did not form Mich gametes. The F3 9 775 (table 16) was probably like- wise homozygous, since her mate is known to have been heterozygous for albinism, but >he produced no albinos in a total of 8 young.

In the foregoing account nothing has been said concerning spotting with white or with yellow; nevertheless spotting of both sorts occurred among certain of the F, and F2 young obtained from the lea crosses. Since the uncrossed lea race contained spotted animals of both sorts, it k n..t surprising that the cross-bred descendants of this race should

GUINEA-PIGS FROM ICA. 29

do the same. Race C, like the lea race, contains only an occasional individual sparingly spotted with white; yellow spotting is of course not visible in a race like C, which contains only yellow or albino indi- viduals. It will suffice to say that the cross-breds, like the parent races, consisted principally of self-colored individuals, and that only an occa- sional dark-eyed individual bore white markings, which in no case were extensive, but were usually limited to a white foot. Among the red- eyed individuals, white spotting was commoner and more extensive, which might seem surprising, unless one remembers that in red-eyed individuals it is impossible to distinguish true white spotting from yellow spotting, since both produce uncolored areas in the coat. Com- plications of this nature make this cross unfavorable for the study of the inheritance of spotting.

SUMMARY ON THE ICA RACE.

1. The "lea race" of guinea-pigs consists of descendants of 1 male and 2 female golden agoutis obtained from the vicinity of lea, Peru, in 1911, and reported to have been caught wild. These animals are supposed to have been descendants of guinea-pigs long since escaped from domestication.

2. This explanation is supported by the observation that within the lea race have cropped out 5 Mendelian variations which are common among domesticated guinea-pigs, viz, (l)the "red-eye" variation, oneof the four allelomorphic forms of the color factor in guinea-pigs; (2) the "non-agouti" allelomorph of the agouti factor; (3) the factor which produces rough coat; (4) the factor for white spotting; and (5) the factor for yellow spotting.

3. An albino guinea-pig of race C differing from wild guinea-pigs by 4 recessive Mendelian characters was crossed with golden agouti females of the lea race. From this cross were obtained in F2 all except one of the expected recombinations of the 4 unit-factor differences between the races crossed. Leaving out of consideration spotting with white and with red, which occurred among some of the hybrids as well as in the uncrossed lea race, there occurred 5 easily distinguishable classes of dark-eyed young and 5 classes of red-eyed young, besides albinos. Only one "expected" class of F2 young was missing, the occurrence of which among other races is well known. There is almost an even chance for its failure to appear in this experiment in the number of young recorded.

4. The four color factors involved in the cross and their allelomorphs

are:

A, a = agouti, non-agouti;

B, b = black, brown;

C, Cr = full color, red-eye;

E, e = extension (of black or brown), restriction.

;;il IMIKHITANCK IX GUINEA-PIGS.

Th) e capable theoretically of forming 16 diflferent combinations, ollows, heterozygous combinations being omitted. The appearance ontaining each of these several combinations is indicated ()pj, pective combinations.

AlK'K. golden agouti = wild type. Single mutations.

\!U •,.. blade-eyed red. AbCF:, cinnamon.

silver agouti. aBCE, black.

Double mutations.

AIH'rr. red-eyed white.

AbCrE, silver cinnamon. !'.. chocolate.

AbCe, brown-eyed red. aBCrE, red-eyed sepia. aBCe, black-eyed red

Triple mutations.

AbCVe, redn yed white, red-eyed white.

abCe, brown-eyed red. abCrE, red-eyed chocolate.

Quadruple mutation. abCre, red-eyed white.

• he I6different combinations, 2 produce black-eyedred individuals indistinguishable except by breeding test; the same is true regarding 1 .r< iwn-eyed reds. Four other combinations identical with these, except for the substitution of Cr for C, produce red-eyed whites, which visibly all alike but which breed differently. Three of the four kinds of red-eyed whites have been identified by breeding test; no doubt the fourth can easily be obtained. The fact that the several classes of red-eyed whites look alike, and that the two kinds of black-eyed reds look alike, and further, that the two kinds of brown-eyed reds look alike, reduces the number of visibly distinguishable classes from 16 to 1 1 , all excepl one of which have been recorded from this single experi- ment. The experiment also produced albinos which theoretically Bhould 1" different formula), if in the formula Ca is everywhere

substituted for its allelomorphs C or Cr. No attempt has been made to distinguish the several expected classes of albinos by breeding tests, the only certain means of identifying them.

Hie cl menl observed between theoretical and recorded

cumbers of V ■ offspring in this cross lends no support to the idea that ciation or linkage occurs among the 4 factorial variations involv

GUINEA-PIGS FROM AREQUIPA. 31

HYBRIDIZATION EXPERIMENTS WITH A DOMESTICATED

GUINEA-PIG FROM AREQUIPA.

While in Arequipa, in December 1911, I purchased in the cabin of a native living near the observatory a pair of domesticated guinea-pigs about one-third grown and perhaps 2 or 3 months old. These animals resembled the ordinary pied guinea-pigs kept for pets or laboratory use in Europe and North America. The female was a tricolor, red, white, and black, and was rough-coated of grade B (Castle, 1905, p. 57). The male was a dilute-pigmented, agouti-marked tricolor (yellow agouti,1 cream, and white), and smooth-coated. This pair of animals was suc- cessfully transported to the Bussey Institution, where they produced 3 litters, of 1, 3, and 2 young respectively. The young of the first 2 litters died at birth; the third litter consisted of 2 males, and as the mother died soon afterward it was impossible to propagate the family farther for lack of females. Of the 6 young produced, 3 were rough-coated and 3 smooth, showing the mother to have been hetero- zygous for rough coat, a dominant character (Castle, 1905). Three were golden agouti and white and three tricolor, one being golden agouti red and white, the other two silver agouti yellow and white.

MALE 1002 AND HIS Fx OFFSPRING.

The father of this family of guinea-pigs (of 1002) proved to be an animal of great vigor and vitality. Although born in Peru (about September 1911) and brought to North America in mid-winter, he has successfully escaped the ravages of disease among our guinea-pigs throughout the rigors of four New England winters and is still vigorous and active. In crosses with other races of guinea-pigs he has sired several hundred young and is now being mated with females which are simultaneously his daughters, his granddaughters, and his great- granddaughters! By repeated back-crosses such as these a race has been established which derives its inherited characters largely from this one animal. This race will be designated the "Arequipa" race.

Crosses of 0*1002 and repeated back-crosses with his female descen- dants have permitted a very full analysis of the factorial constitution of this animal. He possesses either as dominant or as recessive char- acters a majority of the Mendelian variations of guinea-pigs, including one not previously known to occur in any animal other than mice, viz,

xIt should he noted that "silver agoutis" may he of two different sorts: (1) dark-eyed silver agouti with cream-colored hair-tips, and (2) red-eyed silver agouti with white hair-tips. The two varieties resemble each other somewhat and it often requires close observation to discriminate between them, but genetically they are quite distinct. Only the former sort was known to me previous to the Peruvian expedition, and the term "silver agouti" as used in my 1905 paper and by fanciers generally refers to this. It would be better, I think, to use the term cream agouti or yellow agouti for such agouti animals as develop pale yellow in the fur and to restrict the term silver agouti to those which are non-yellow.

INHERITANCE IN (iUINEA-PIGS.

the pink-eye variation with colored coat, first brought to the attention

cientiste in the case of mice through the experiments of Darbishire

1902 . \ similar variation has, however, since been found to occur

in rate | astle,1914). Thenumberof factors in which <? 1002 is hetero-

goua i- surprisingly large and implies doubtless considerable cross-

breeding in the guinea-pig colonies kept by the natives of Arequipa, a

- perhaps connected with the great size and vigor of their animals.

The factorial constitution of 6*1002, as at present understood, is as

follows:

(1) Agouti factor, An. agouti-marked but transmitting non-agouti as a recessive

character. 2 Black factor, BB, homozygous.

Color factor, C,iCr, two different recessive variations, dilution (Cd) being dominant over red-eye (Cr). Both are recessive to ordinary intense color ( ' and dominant over albinism (Ca), the four forming a series of quadruple allelomorphs, us shown by Wright (1915). Extension factor, EE, homozygous.

Dark-eye factor, Pp, heterozygous for the recessive pink-eye (p) variation, with which goes dilution of black or brown pigments, but not of yellow.

(6) As regards the rough variation, this animal is smooth, but nevertheless trans-

mit- occasionally a trace of the rough character, but the character does not crop out among his descendants in any as yet recognizable Mendelian proporl

(7) White spotting, homozygous.

(8) Yellow spotting, homozygous.

cation of young obtained from matings of c?1002 with unrelated guinea-pigs.

Mothers.	Intense dark-eyed.	Dilute dark-eyed.	Red-eyed.
Golden agouti.	Black.	Yellow agouti.	Sepia.	Silver agouti.	Sepia.
5 dnrk-t \ ed aon-agoutia 16 dark-«yed aon-agoutis, heter- ozygous for allunism	4 16	7 8	2 6 2	2 2 5	17 9	6 2
T-.tnl
20	15	10	9	26	8

Male 1002 waa mated with 20 dark-eyed guinea-pigs and 6 albinos

derived either from race B, from a 4-toed race (see Castle, 1906), or

in crossee between the two. Both these races contain only non-

agouti animals. The dark-eyed mothers produced 70 Fj young, the

ilbino mothers produced 18 F! young. Disregarding spotting with

yellow and with white the young of the dark-eyed mothers fall into

elates dark-eyed intense, dark-eyed dilute, and red-eyed— and

h class may be further subdivided into agouti and non-agouti. (See

table i

I he albino mothers, though derived from the same races as the dark-

1,1 mother, produced only two of the three main classes of young,

dark-eyed dilute and red-eyed, which fact confirms Wright's (1915)

GUINEA-PIGS FROM AREQUIPA. 33

conclusion that albinos, even when derived from intense dark-eyed parents, do not transmit intensity to their young in crosses, for the reason that albinism is an allelomorph of dilution as well as of intensity. The fact that c? 1002 produced no albino young, even when mated with albinos, shows that he did not produce albino gametes. The fact that he produced red-eyed young when mated with albinos shows that he transmitted red-eye as a recessive character and that this is dominant over albinism. The fact that he produced no intense dark-eyed young by the albino mothers shows that he lacks intensity and forms only gametes which transmit either dilution or red-eye. All these facts are in harmony with the hypothesis suggested by Wright (1915) that inten- sity, dilution, red-eye, and albinism are allelomorphs of each other, so that a gamete can transmit only one of the four allelomorphs, and a zygote can contain only two of them. Male 1002 is evidently a hetero- zygote of dilution and red-eye (CdCr), both of which are recessive to intensity (C) and dominant over albinism (Ca). Consequently, when he is crossed with albinos, zygotes of two sorts are expected in equal numbers, viz, CdCa and CrCa (dilute and red-eyed), as observed (table 19); and when he is crossed with intense dark-eyed animals carrying albinism as a recessive character (as 15 of the 20 dark-eyed mothers of table 19 did), zygotes of four sorts are expected in equal numbers, viz, CCd, CCr, CdCa, and CrCa, the first two being dark-eyed intense, the third dark-eyed dilute, and the fourth red-eyed. The observed result is in perfect accord with this expectation as regards the classes of young produced, and agrees with expectation sufficiently well as regards the proportions of the classes.

It is expected further that each of the three main classes will be sub- divided about equally into agoutis and non-agoutis. The 6 expected subclasses appear in the experimental results, but there is a considerable excess of agoutis, viz, 56 agoutis to 32 non-agoutis. Whether this depar- ture from the expected equality has any probable significance will be considered further in connection with the F2 generation.

F2 OFFSPRING OF tf1 1002.

For the production of an F2 generation a golden agouti and 4 silver agouti males were selected. The golden agouti male was mated only with a black female, his sister, but the silver agouti males were mated with practically all classes of the Fx females. (See table 20.) They produced altogether 190 F2 young, which, being classified as regards intensity and eye color alone, fall into 6 main classes, viz, (1) dark-eyed intense, (2) dark-eyed dilute, (3) red-eyed, (4) pink-eyed, (5) red-and- pink-eyed,1 and (6) albino. Each of these main groups falls into two

1Animals called red-and-pink-eyed are in reality pink-eyed, but lack yellow in the coat. They transmit in every gamete hoth the factor for pink-eye and the factor for red-eye.

INHKKITANCE IN GUINEA-PIGS.

sulichi.-x - agouti and non-agouti), readily distinguishable, except in the ease of albinos.

Matings in which both parents were agoutis produced 44 agouti to 13 Don-agouti young, expected 43 to 14, which is good agreement. Matings of an agouti with a non-agouti animal produced a considerable os of agoutis, viz, 58 agouti to 42 non-agouti, where 50 to 50 is the expected distribution. The departure from the expected equality of agouti and non-agouti young is, however, not probably significant.

Tmu.k 20. -Classification of the F2 young of &1002, obtained from Fi animals

classified in table 19.

Nature of Fi mating.	Dark-eyed intense.	Dark-eyed ^, , , dilute. Red-eyed.
Golden agouti.	Black.	Yellow agouti.	Sepia.	Silver agouti.	Sepia.
Golden agouti X black	12 2 4	4 4	3 2 4 5 11	i 3 I	1 2 6 7 12
Golden agouti X silver agouti .... Black X silver agouti
Yellow agouti X silver agouti. . . . I >ark-eyed sepia X silver agouti. . Silver agouti X silver agouti Silver agouti X sepia (red-eyed) . . Total	2 10	2 12 28 7
18	8	25	16	53	28
Dark-eye X red-eve
6	4	22	15	18 35	9 19
Red-eye X red-eye
Nature of Fi mating.	Pink-eyed.	Red-and-pink- eyed.	Albino.
Agouti	Non- agouti.	Agouti.	Non- agouti.
< ■< .hlen agouti X black	1	1	3 2	2	4 11 7 11
Golden agouti X silver agouti . . . . clc X silver agouti
Yellow agouti X silver agouti . . rk-eyed sepia X silver agouti . . Silv r agouti X silver agouti gouti X Bepia (red-eyed) . . •al
1	1	5	2	33
Ic-eye x red-eye
1	1	c		15
• ■ r. d-eye
						£		XO

we summarize the matings in which every mother and father is w.i to luivo been capable of producing albinos, we have 96 colored albino young; expected 93 to 31— a very good agreement with bation.

iimmariiring the matings between a red-eyed male and a dark-eyed

male known to have been capable of producing red-eyed young, we

gel Id dark-eyed and 27 red-eyed; expected 33.5 and 33.5. This

GUINEA-PIGS FROM AREQUIPA.

35

apparent deficiency of red-eyed young may have been due to our failure at first to distinguish dark-eyed sepias from red-eyed sepias, which look very much alike when first born. In the summary all sepias are treated as dark-eyed unless a specific record in the ledger indicates that they were red-eyed.

Matings yielding pink-eyed young produced 17 non-pink-eyed and 6 pink-eyed young, which is good agreement with the expected 3 to 1 ratio.

BACK-CROSS AND OTHER OFFSPRING OF ^1002.

Male 1002 was mated with certain of his Fi daughters, producing 90 young of 10 different color classes, as indicated in table 21. He was later mated with certain of the female young produced by the matings last described, these females being both his daughters and his grand- daughters and so "f-blood" Arequipa tracing back to himself. (See

Table 21. — Classification of young of (3*1002 by his Fi daughters (table 19).

Mothers.	Intense dark- eyed.	Dilute dark- eyed.	Red-eyed.	Pink-eyed.	Red-and- pink-eyed.
Golden agouti.	Black.	Yellow agouti.	Sepia.	Silver agouti.	Sepia.	Agouti.	Non- agouti.	Agouti.	Non- agouti.
4 black	1	6	3 2 13 8	3 1 3 7	11 5	3 1 7 2	2 3 2	1 2	1 2	1
1 sepia (dark-eyed). 6 silver agouti 3 sepia (red-eyed) . . Total
1	6	26	14	16	13	7	3	3	1

table 22.) The table 22 matings have produced to date (April 1915) 61 young, distributed in 9 of the 10 classes represented among the table 21 young. The classes of young recorded in tables 21 and 22 are the same as those represented among the F2 young (table 20), with the exception of albinos, which are never produced by & 1002, since he does not transmit albinism, which is recessive to both of its allelomorphs. As regards the characters in which 6*1002 is heterozygous, there is evidence from tables 20 to 22 that in the case of each he forms equal numbers of gametes bearing the dominant and the recessive allelo- morphs respectively. By agouti daughters he has had 50 agouti and 15 non-agouti young; expected, 49 and 16. By non-agouti daughters he has had 45 agouti and 41 non-agouti young; expected, 43 of each. Combining these totals with those recorded in table 19, we find that in all matings with non-agouti animals he has sired 101 agouti and 73 non-agouti young, a not improbable chance deviation from the ex- pected equality of the two classes.

By red-eyed daughters cf 1002 has had 51 dark-eyed and 40 red-eyed young. Adding to this result that recorded in table 19 (last category

INHER1 I \\< i [N '.I [NEA-PIGS.

of matingB), which has the Bame expectation of rod-eyed young (50 per t b total of 68 dark-eyed and 51 red-eyed, fairly good agreement with the expected equality.

Bj dark-eyed daughters which have produced red-eyed young and

bown thai they transmil eit I hm- red-eye or albinism, <? 1002 has

produced 18 dark-eyed and L6 red-eyed young; expected 25.5 and 8.5.

ubtleea other daughters which have not chanced to produce red-eyed young in the litters recorded arc also heterozygous for red-eye, in which

■ ■■ their recorded litters should be added to the foregoing. If this were done, the observed departure in this case from the expected 3 to 1 ratio would doubtless disappear.

Tm m of young of •ftOO& by his granddaughters winch were also his

daughters {table .'!)■

Mother.	I t»t t- 1 , — . ■ daxk- •d.	Dilute dark- eyed.	Red-eyed.	Pink-eyed.	Red-and- pink-eyed.
( roldeti agouti	Black.	Yellow agouti.	Bepia.	Silver agouti.	Sepia.	Agouti.	Xon- agouti.	Agouti.	Non- agouti.
2ri touti 1 Mark	l	l	1 •_> •_> 1 4 3	2 1 1 1 1	1 1 2 1 2 3	1 2 1 1	1 1 3 2	1 3	1 2 1 1
uti 2 nilvrt Kgouti 2 * 2 DOn-SCOUtJ (pink-
2 red-and- pinl tal
4	1	16	6	10	8	7	4	5	0

By pink-eyed daughters, d" 1002 has produced 7 pink-eyed young and B with eyes not pink — complete agreement with the expected equality. By daughters not pink-eyed, but which nevertheless are clearly hetero- lygoue in pink-eye, he has produced 23 pink-eyed and 58 not-pink-eyed young; expected, 20 and 61 an excess of pinks capable of explanation on the Bame ground as the excess of red-eyed young.

dilute-colored daughters o" 1002 has produced 52 dilute-colored

young, bul do intense-colored one-, as expected, since dilution is

ive to intensity. By intense-colored daughters heterozygous for

dilution he has produced in intense and 9 dilute young, equality being

ad.

MISCELLANEOUS MATINGS OF THE DESCENDANTS OFo*1002.

MatingB of the descendants of 71002 beyond the F2 generation were made chiefly with a view- to test further the genetic character of the new Their results are presented in tables 23 to 28 and serve to firm the interpretations already offered.

GUINEA-PIGS FROM AREQUIPA.

37

Matings of red-eyed animals inter se have in most cases produced only red-eyed or albino young, but two matings have also produced pink- and-red-eyed young, i. e., animals which are pink-eyed but develop no yellow in their fur, in which last respect they differ from ordinary pink-eyed and agree with ordinary red-eyed. (See tables 24 and 27.)

Table 23. — Young produced by matings of red-eyed males, descended from c?1002, with dark-eyed females of race B.

Nature of mating.	Dark-eyed.	Red-eyed.	Albino.
Golden agouti.	Black.	Silver agouti.	Sepia.
Silver agouti X black (homozygous) .... Silver agouti X black (heteiozygous in albinism)	1 20	5 7 6	9	7	10
Sepia (red-eyed) X black (homozygous) . . Total
21	18	9	7	10

Most of the red-eyed animals, when bred inter se, produce albino as well as red-eyed young, showing themselves to be heterozygous for albinism and so of the formula CrCa. This is not surprising when we recall that all the Fx red-eyed animals must by hypothesis be of this formula, and that two-thirds of the F2 red-eyed should be of the same sort. In a few matings of red-eyed with red-eyed, which failed to produce albino young (table 24), it is probable that one or both parents

Table 24. — Young produced by matings inter se of red-eyed descendants ofc?1002. (See also table 27).

Nature of mating.	Red-eyed young.	Albino young.
Silver agouti.	Sepia.
	56 18	19 6 9	22 11
Total
74	34	33

were homozygous for red-eye. Matings of red-eyed with albino animals (table 25), which failed to produce albinos in 6 or more young, afford clear criteria for red-eyed animals free from albinism and so of formula CrCr. Only one mating of a red-eyed animal with an albino has pro- duced pink-eyed young. (See table 27.) The red-eyed parent in this case (c?48) was mated with 4 other albinos (all of race B) without producing pink-eyed young, but only red-eyed (13) and albinos (17). The female which produced pink-eyed young was his sister, derived like

INHERITANCE IN GUINEA-PIGS.

himself from paicnti known to transmil pink-eye This indicates that the character pink-eye in guinea-pigB (as in mice) may be transmitted |,v ultiinos. The fart should be emphasized that the pink-eyed young

Tab1 ) . , ... lueedbyred^yed descendants of &1008 matedivith

albinos. [See also tabl 97.)

Nature of red-eyed parent.

Red-eyed young.

Silver agouti.

BQver agouti, hefc as fox albinism 21

. heterosygoufl for albinism

Silver agouti (homoiygouB for red-eye)

,ua for red-eye)

Total.

is

Sepia,

11 31 11 20

39

73

Albino young.

17 34

51

produced m this mating were also red-eyed, i. e., were non-yellow, for

red-eyed animal- may carry pink-eye as a recessive character, and con-

rsely pink-eyed may carry red-eye as a recessive character. How-

,-. if these recessive characters crop out as recessive individuals

from a mating of two like parents with each other, it can in either case

:r only in the form of the double recessive, both pink- and red-eyed.

Xai -• Young produced by pink-eyed descendants of <?1002, mated inter se.

Nature of mating.	Pink-eyed.	Pink-and- red-eyed.	Albino.
Agouti.	Non- agouti.	Agouti.	Non- agouti.
	15 4	2 4	1		1 7
One parent agouti, one non-agouti. . . •:il
19	6	i ! o	8

Pink-eyed animals (with yellow in their fur) have made their appear-

produced by mating dark-eyed animals inter se.

and 27.) In some cases red-eyed young have been

produced by the Bame matings, or pink-and-red-eyed or albinos, for

pin ma to be quite independent of the color factor in its inherit-

eyed animals mated inter se have produced only pink-eyed,

pink-and-red-eyed, and albino young. (See table 26.) Any of these

three fora derived will doubtless be found to transmit pink-eye in

Pink-and-red-eyed animals of whatever origin have been found to

produce when mated with each other) only pink-and-red-eyed young

albinos. Hut the record as regards albinos is doubtful. Two

GUINEA-PIGS FROM AREQUIPA.

39

albino young have been recorded as produced by c?88 mated with his daughters, 9 204 and 9 205; but this same male mated with albino females of race B produced 11 red-eyed young but no albinos, for which reason it seems very doubtful whether he transmits albinism. More probably the two young by pink-and-red-eyed mothers were not albinos, but very pale-colored non-yellow young, possibly lacking the extension factor, in which case their fur would be pure white, their eyes being uncolored because of the pink-eye factor. If so, they would be in appearance indistinguishable from albinos, though behaving very differently in crosses.

Table 27. — Matings of descendants of &1002 which have produced pink-eyed young.

Nature of mating as regards —	Dark-eyed young.	Red-eyed young.
Color factor.	Agouti factor.	Golden agouti.	Black.	Silver agouti.	Sepia.
Dark-eye X dark-eye. . . Dark-eye X red-eye ....	Agouti X agouti Do		4 3	2	1 8	2 4 5
Do
Red-eye X red-eye ....	Agouti X agouti
Red-eve X albino	Non-agouti X non-agouti . .
Total	7	2	9	11
Nature of mating as regards —	Pink-eyed young.	Pink-and-red- eyed young.	Albino young.
Color factor.	Agouti factor.	Agouti.	Non- agouti.	Agouti.	Non- agouti.
Dark-eye X dark-eye. Dark-eye X red-eye . . .		Agouti X agouti	2 1	1 1	1 3	2	7 5
Do
Do Red-eve X red-eye ....	•	Agouti X non-agouti Agouti X agouti
Red-eye X albino	SJon-agouti X non-agouti .
Total	3	2	4	2	12

Nevertheless, it is to be expected that pink-and-red-eyed animals can be produced which are heterozygous for albinism. Such animals necessarily would be heterozygous for red-eye also, which is an allelo- morph of albinism, and so would be of the formula CrCapp, for it is known (1) that pink-eyed animals may transmit albinism; (2) that red- eyed animals may transmit albinism ; and (3) that pink-eye and red-eye are independent of each other in transmission. Consequently, there is every reason to suppose that albinism may exist as a recessive allelo- morph of red-eye in animals which are both pink-eyed and red-eyed.

A pink-eyed animal mated with a dark-eyed one produced 3 dark- eyed young and 1 albino, which adds to the evidence that pink-eye is recessive to dark-eye and may be present in the same zygote as albin- ism. A pink-eyed animal (9 307) mated with a pink-and-red-eyed

40

INHERITANCE IN GUINEA-PIGS.

male (6^140) produced 2 pink-and-red-eyed young. A pink-and-red- eyed male mated with 2 dark-eyed females of race B, which were hetero- zygous in albinism, produced 4 dark-eyed and 4 red-eyed young. Another pink-and-red-eyed male (140) mated with a dark-eyed female descended from d"1002 produced 1 pink-eyed young, in which red-eye was undoubtedly recessive. Most of the foregoing matings are tabu- lated in table 28.

Table 28. — Character of young -produced 'in matings of pink-and-red-eyed descendants of c? 1002.

Character of mate.

Dark-eved

Do

Red-eyed

Do

Pink-eyed

Albino

Pink-and-red-eyed .

Do

Do

Character of mating as regards agouti.

Agouti X agouti

Agouti X non-agouti

Agouti X agouti

Agouti X non-agouti

Do

Do

Agouti X agouti

Agouti X non-agouti

Non-agouti X non-agouti . .

Total .

Dark-eyed young.

Golden agouti.

Black.

Red-eyed young.

Silver agouti.

2 4

19

Sepia.

2

2

10

21

Character of mate.

Dark-eyed

Do

Red-eyed

Do

Pink-eyed

Albino

Pink-and-red-eyed .

Do

Do

Character of mating as regards agouti.

Agouti X agouti

Agouti X non-agouti

Agouti X agouti

Agouti X non-agouti . . . .

Do;

.... Do

Agouti X agouti

Agouti X non-agouti

Non-agouti X non-agouti .

Total .

Pink-eyed young.

Agouti.

Non- agouti.

Pink-and-red- eyed young.

Agouti.

4 1

1 15

21

Non- agouti.

Albino young.

3

1

2

18

G

2(?)

1 2(?)

30

1+4 (?)

Pink-and-red-eyed males mated with albinos of races free from pink- eye produce red-eyed or albino young, but not pink-eyed young, since pink-eye also is a recessive character and becomes visible only when doubly represented in the zygote.

It is clear, accordingly, that when pink-and-red-eyed animals are mated with red-eyed animals (or albinos), young are produced which are red-eyed, but in which pink-eye is recessive; and when the same pink-and-red-eyed animals are mated with pink-eyed animals, young are produced which are pink-eyed, but in which red-eye is recessive. Hence thetwo characters, pink-eye and red-eye, are independent of each other, though red-eye is the dominant allelomorph of albinism, with which pink-eye is wholly unrelated.

GUINEA-PIGS FROM AREQUIPA. 41

SUMMARY ON THE AREQUIPA DOMESTICATED RACE.

1. A domesticated male guinea-pig obtained from the cabin of a native in Arequipa, Peru, has proved to be of great interest because of the large number of color mutations which it either possesses or trans- mits without itself manifesting them. Two wholly new variations (red-eye and pink-eye) were obtained from this animal as recessive characters. The former has been obtained subsequently from the lea race and the latter from a race of guinea-pigs brought from Lima by Professor Brues. Both are probably variations of long standing among the guinea-pigs kept by the natives in Peru, but seem not previ- ously to have been observed among guinea-pigs in Europe or North America.

2. Red-eye is a Mendelian allelomorph of albinism, of dilute pig- mentation, and of intense pigmentation, the four being quadruple alle- lomorphs (Wright, 1915). A gamete may transmit one of the four, but not more; a zygote may contain and transmit (separately) two of the four, but not more. Dominance is in the order of decreasing intensity, viz, (1) intensity, (2) dilution, (3) red-eye, (4) albinism. Intensity and dilution affect all pigments similarly; red-eye and albin- ism inhibit yellow completely, but affect black in very different de- grees, the inhibition of black being nearly complete in albinism, but being partial only in red-eye. Red-eye is a variation unknown as yet in any other animal, but the sooty coat of young Himalayan rabbits possibly is a parallel variation, and the same may be true of Siamese cats.

3. Pink-eye is a variation wholly independent genetically of albinism. It affects only black (or brown) pigments, the intensity of yellow pig- ment being unimpaired in its presence. A similar variation (genetically and physiologically) occurs in mice and also in rats.

4. The new variations (red-eye and pink-eye) have formed, with the previously known unit-character variations of guinea-pigs, many new unit-character combinations, which are here described.

5. From the crosses of the Arequipa domesticated guinea-pig with other guinea-pigs, involving a maximum number of color factors, no evidence is forthcoming that any two of the factors are "coupled" or "linked."

12 [NHERITANCE IN GUINEA-PIGS.

SIZE INHERITANCE IN GUINEA-PIG CROSSES.

PREVIOUS WORK ON SIZE INHERITANCE.

For Beveral yean my pupils and I have been engaged in studying the inheritance of siae among tame or domesticated animals, a subject deserving of careful investigation both because of its economic impor-

ce and because of the light which it may throw on general theories

heredity. \ preliminary study based on skeletal measurements of rabbit* was published in 1909 (Castle etal), which seemed to show that site inheritance is "blending" and does not involve the segregation and ml iin.it ion of distinct Mendelian size-factors. MacDowell (1914), at my suggestion, repeated this work on a larger scale and with similar observational results, establishing, however, the additional fact that

I ' . i.r a back-cross generation, usually shows greater variability in than an 1\ generation, following a cross between animals of unlike

s, though the general result in both cases is the production of inter- mediates. On theoretical grounds MacDowell favored the Nilsson- 1 Ihle view that all variation, even when continuous, is caused by genetic

tors themselves discontinuous, and that blending inheritance in- volves multiple segregating factors. But MacDowell points out that this interpretation is not the only one of w7hich his observations are capable. The genetic purity of his material also, while sufficient to

tblish the general blending character of the inheritance, is not suffi- cient to meet the extreme demands of the multiple-factor hypothesis.

At the same time that MacDowell was making his observations on inheritance in rabbits, Detlefsen (also in my laboratory) made observations on size inheritance in crosses between Cavia rufescens and the guinea-pig. He was unable to rear an F2 generation, because of the <•< implete sterility of the male hybrids, but from a study of repeated backn concluded that "there were no great differences in vari-

ability in the back-crosses of hybrids to guinea-pigs which would indi- cation and recombination of factors for size." This conclu-

D he reached without theoretical bias, for he adds: "The results in DO way controvert the possibility that size may be due to factors which are inherit id in Mendelian fashion; but segregation was not apparent in these classes of matings in this species cross."

My colleague, Dr. John C. Phillips, at about the same time (1912, 1914), undertook crosses of very pure races of ducks, which differed widely in rise, viz, Rouens and Mallards. The two parent races did not overlap in variability in size. The F, offspring were of intermediate

ight, as were also the F2 offspring. The F2 generation of 63 indi- vidual- included only 1 individual which fell outside the range of the 1 individuals, though the standard deviation of the F2 generation was somewhat larger than that of the Fx generation. Four body-

size. 43

dimensions of the Fx and F2 ducks were also studied by Phillips, viz, length of bill, tarsus, neck, and total length. Length of bill and length of neck were slightly more variable in F2 than in F: ; length of tarsus was slightly less variable in F2 than in F1( while total length was more variable in males but less variable in females in F2 than in Fx. It thus appears that F2 is not even uniformly more variable than Fx in size char- acters in this the purest material that had thus far been investigated as to size inheritance among animals. Yet this supposed increase of vari- ability is the only criterion of segregation in size crosses which has been discovered or even suggested. Surely this is a wholly inadequate basis on which to rest a theory that all inheritance is based on discontinuous Mendelian factors.

While the several investigations of size inheritance in rabbits, guinea- pigs, and ducks were in progress, but before their outcome had become apparent, the Peruvian expedition brought to the laboratory material which seemed very favorable for such studies, and I have constantly kept in mind its use in this way. Cavia cutleri from Peru gave us a small race of undoubted purity, less than half the size of the guinea-pig, but which has been found to produce fertile hybrids with it, which per- mits obtaining an F2 generation, a thing impossible with the rufescens hybrids. The lea race and the Arequipa race have also afforded valu- able material for size crosses with our own long-inbred and standardized races of guinea-pigs.

The results which have been obtained, so far as the demonstration of mendelizing size-factors is concerned, are negative, like those previ- ously obtained, though in some respects the material is more satis- factory. But from their bearing on the question whether or not size inheritance depends upon discontinuous Mendelian factors, these ob- servations have, it is believed, several interesting features which will become apparent as the description progresses.

WEIGHTS AND GROWTH CURVES OF CAVIA CUTLERI, OF VARIOUS GUINEA-PIG RACES, AND OF THEIR HYBRIDS.

It will be recalled, from the description of the color inheritance crosses, that cutleri males were crossed with females of two inbred races of guinea-pigs, which we have designated races B and C respectively. Many observations have been made on the weight of race B covering the period from birth to old age. These afford an accurate knowledge of the variability in weight of race B, and of the normal growth rate of animals of this race. Our knowledge of the weight of race C is less complete, though this race is equally inbred and appears not to be more variable in size than race B. Its average size is probably a little greater than that of race B, but the difference is negligible in comparison with the difference of both from the size of C. cutleri. In the hybridization experiments, race C hybrids were bred inter se, as were also the race

•H

INHERITANCE IN I .TINEA-PIGS.

B hybrids, and in no case were hybrids from the two races bred with h other. Nevertheless, the results obtained in the two cases were so

Bimilar that for statistical purposes it was thought best to combine ui. Race K is taken as the standard guinea-pig race with which the

hybrids are compared.

1 or a period <>t about a year and a half all cutleri individuals in the laboratory, whether <>f pun- race or hybrids, were weighed two or three times a month. In this way records were obtained from which growth , ,:, gee of weight, etc., can be deduced. The repeated and

-went weighings allow the detection of periods of depression due to illness or poor feeding. Due allowance has been made for all such observations, as well as for increase in weight of females through pregnancy. Nevertheless, observations on weight are at best not

i

dRaceB

H
		9 Race B.
<-•										6 Cutleri
«•«.'										9 Cutleri

.Day$ 40

SO

120

160

200

2-10

2S0

320

360

400

I 1

rowth-curves of C- cutleri and of race B guinea-pigs, the growth-curve of each sex being shown separately.

altogether satisfactory, since they are subject to fluctuation through conditions of food, accumulations of fat when maturity has been reached, etc. Greater value attaches to the bone measurements of fully adult individuals (over 1 year old) so far as individual varia- bility i> concerned. But the observations on weight afford a basis entirely satisfactory for the determination of average sizes and average growth curve- in different classes of hybrids. Incidentally they afford ontrol on the bone measurements, for they indicate cases of abnormal through disease, fighting, or other cause) and allow of either remedying condition- or rejecting suspicious material.

Pure ctUU n young of both sexes are of about the same average weight

birth, vis, :>() grams (see fig. 1). The females at first grow a little

than the male-, a fact perhaps correlated with their earlier sexual

maturity. At about 50 days of age the two sexes are of practically the

-lit. the male- having again caught up with the females, and

SIZE.

45

subsequently the males are heavier. The average adult weight of a female is about 400 grams, that of a male about 420 grams.

Race B animals of both sexes weigh on the average about 80 grams at birth (see fig. 1), but females grow at first a little faster than males, so that between 10 and 50 days of age females are slightly heavier. But the males soon catch up with the females and from 50 days on are heavier. The same difference between the growth curves of the two sexes is observable here, as in Cavia cutleri. The phenomenon is pos- sibly a general one among mammals. Earlier maturity of the female is attended by more rapid growth, but the ultimate weight attained by males is greater. There is no indication in our observations that the attainment of sexual maturity is followed by any slowing-up of the growth rate in either sex.

In the growth of both C. cutleri and of race B, as in other growth- curves to be described, the curve is at first concave upward, but later becomes convex upward. This agrees with observations on rabbits, fowls, and other organisms, and its significance has been discussed elsewhere (Castle et at., 1909).

1 Weight in								rf Fr Cut x Band C
Oram 800	s		6 Race B.
								dF^Ci	it.xBandC.
600
									S Cutleri.
400
200

Age in Days 40

80

120

160

200

240

280

320

360

400

Fig. 2 — Growth curves of race B and cutleri males and of their male hybrids, both Fi and F2.

Fj hybrid males (from the cross <? cutleri X 9 race B or C, fig. 2) weigh about 85 grams at birth, i. e., they are slightly heavier than the young of either pure race, a lead which they retain throughout subse- quent life. At maturity they weigh about 890 grams, as compared with 800 grams, the average adult weight of race B males, and 420 grams, the average adult weight of pure cutleri males. The females (fig. 3) weigh about the same as the males at birth, or are even a little heavier, but soon begin to grow less rapidly, weighing about 750 grams when 1 year old. The F2 hybrids of both sexes are smaller than the Fi hybrids from birth on, a fact of undoubted significance. (See figs. 2 and 3.) The superior growth impetus which was produced by

INHERITANCE IN I ;UINEA-PIGS.

hybridization ha* no! been retained in the second-generation offspring, which -ink as regards weight to a position intermediate between the part-ni races. N evert hclcss the Fa hybrids are nearer to race B than to cutlery in adult size, which fact suggests that not all the growth ,„,; furnished by hybridization has yet been dissipated. Inform

utli curve the 1 hybrids are also intermediate. The growth curve at first rises rapidly, due in part perhaps to the good milk-giving qualities of their vigorous F, hybrid mothers, but in part probably to inheritance of cutleri qualities, since the cutleri growth curve is a rela- tively steep but low one, indicating rapid growth at first and early turity. The I hybrids also grow rapidly at first, being consider- ably heavier than race B animals until an age of 120 to 150 days has • ii reached. Then they fall below and stay below the weight of race H animals, running a course nearly parallel with that of pure cutleri animals, whereas the growth curve of the Fx animals more nearly approached that of race B animals.

I -i Dayt 40

\ ill curve - of nice R and cutlrri females and of their female hybrids, both Fi and Fj.

\\ bile we are on this subject it may be well to refer to the growth

curves observed in the cross between the Arequipa male, 1002, and

females of race B (or of similar character). (See fig. 4.) The data for

the growth curves of males are more complete in this case than the

data for females and accordingly only the former will be considered.

1 1 nimals are of greal size and vigor, attaining an average adult

weighl of over 1,200 grams. The F2 animals are even larger at birth

than the I animals, a fact which indicates that the size of the mother

mething to do, other than through heredity, with the size of the

mg at birth, for the Fa young rapidly lose the lead which they had

in weight at birth over their F, parents, and subsequent to 40 days of

fall below them in weight. At maturity they weigh less than 1,000

having lost more than half of the gain which the F! animals

showed over race B animals. This difference, it should be stated

phatically, i- not due to environmental conditions of any sort, such

SIZE.

47

as season of the year, food, or the like, for it exists between lots of Ft and F2 animals reared simultaneously and treated exactly alike. It is clear, therefore, that a cross of distinct races, whether wild or domesti- cated, brings a stimulus to growth which leads to the attainment of size considerably beyond that which truly inherited size factors would produce. This stimulus, however, lasts unimpaired for only a single generation. But if it lasts at all into a second generation, and if its persistence is not uniform in amount in all cases, it is evident that it would increase the variability of F2 as compared with Fx. This is a matter requiring careful consideration when the significance of increased variability in F2 is considered.

Weight

Age in Days 40

Fig. 4. — Growth curves of race B males and of the male hybrids, both F! and F2, between the Arequipa male 1002 and females of race B or similar races.

SKELETAL MEASUREMENTS OF CAVIA CUTLERI, OF VARIOUS RACES OF GUINEA-PIGS, AND OF THEIR HYBRIDS.

It has been stated that skeletal measurements of adult animals are considered more reliable criteria of size than total body-weight. For this reason we have carefully preserved for study the skull and the long bones of the right fore leg and right hind leg of each adult animal which died a natural death or was killed, in the races whose size was under investigation. Observations made by MacDowell, Detlefsen, Wright, and Fish (see MacDowell, 1914, appendix) have shown the vari- ous long bones of the legs to be closely correlated in length, so that it seems sufficient for our purpose to measure a single one of these, and we have chosen for this purpose the femur. On this we have taken the

IMIKKH \\< l. IN QTJINBA-PIGB.

length measurement as indicated on MacDowell's figure 5, F. Two observations have also been made of skull dimensions, one of basilar skull length as indicated in MacDowell's figure 1, 0. M., and the other of maximum lygomatic width [Z). The measurements here dealt with found upon repeated measurement to be accurate within 0.1 mm. The observations are combined in classes of 5 mm. range in tables 29 to :•] for the Beveral races and hybrids studied. We may consider first the obeervations on skull length.

THE CUTLERI HYBRIDS.

1 1 ii adult cvthri females have skull lengths distributed as shown in table *-)(.». The range extends over 10 classes; the mean is 51.55 mm., and the standard deviation 13.50 mm. Twenty-eight adult females of race B have a range in skull length of 15 classes; their mean is 58.14 nun., and the standard deviation 19.75 mm. Fr hybrid females between cuth ri males and females of races B and C available for study number 24. with a range of 13 classes, a mean skull length of 57.70 mm., and a standard deviation of 16.85 mm. These figures indicate (like the weight observations) that the Fx hybrids are practically as large as the larger parent race and not more variable. The 33 F2 hybrids studied show a range of 18 classes, with a mean at 54.35 mm. and a standard deviation of 17.20 mm. The Fj mean was about 3 mm. greater than the intermediate between the races crossed, but the F2 mean practically coincides with it. Judged by the standard deviation, 1 _ is not more variable than pure race B and is only slightly more variable than F,. The means show in Fx an increase in size over that we should expect through inheritance, but a loss of this increase in F2.

Observations on the male hybrids from the same cross are recorded in the next four rowrs of table 29. The mean of the Fx hybrids is again greater than that of either pure race and surpasses the inter- mediate point by nearly 4.5 mm. The mean of the F2 hybrids is close to the intermediate, which it exceeds by about 0.6 mm. The vari- ability standard deviation) of F2, however, is considerably greater than that of 1' and even exceeds that of pure race B. Particularly note- worthy is the occurrence of one very large F2 individual, nearly as large as the largesl 1 individual. Compare this with the occurrence of a single very hh.iII 1' female, as small as the smallest pure cutleri female.

The cull, ri hybrids with races B and C show similar results in regard to the other measurements studied — zygomatic width and femur length. t ablee 31 > and 3 1 and] >late6.) In every case F{ exceeds the means of both parent races, but I' approximates the intermediate between them, which n exceeds by a fraction of a millimeter only. In no case is the I mean as great as the race B mean— that of the larger parent race. These facts, like the weight curves, indicate (1) that, so far as heredity oncerned, an exact intermediate between the parent races would

SIZE.

49

Si

•<S>

Hi

-e

s a

a

£

s

6b

a

a.

-2 S

S

OS

a

dig

s

=0

a

C5

«

•noij -BiAap pjBpuB;g

13.50 19.75 16.85 17.20 9.45 15.05 12.15 20.00 12.10 20.65 15.15 17.75 10.00 10.00 11.20 22.65 23.55 24.55

•UBaj^

51.55 58.14 57.70 54.35 52.91 60.35 61.20 57.26 61.92 60.44 64.40 61.86 57.45 60.20 58.84 58.10 62.20 62.17

■JKJOJ,

O X •* CO t^ CO CO -* GO i-h t^ CO X t> f O X t-- ^h CI CI CO tONN rH) T* N U5 i-H S J_,

"o a 03 a a 03 L. U~ a 03 s | 1 73 a> pa CD S3 o

* ' Z9-0 ' Z9

(N

6 99-5 ' 99

^H

f-H

f 99-0 ' 99

6 ' 59-5 ' 59

TJH —4

^'59-059

Tj< rt

■ i-H

6 ' ^9-5 ' f9

i-H CO

• • CO

ff9-Of9

* ■ r

H

rH IO CI

■ rH CO

6 ' S9-5 ' £9

H rH

Mh CO -^

■ i-H

fr"£9~0'89

■ c

1HH CI CO CI io

■ i-H

6 ' S9-5 ' S9

H I— 1

CI -^ CO m

f S9-0 ' S9

• T

r ci

■ CI • IO

• i— 1

6' 19-5 19

1

- CI

CO "O r-l Ol

• • CI

^19-0 19

' • T

f m

Tf CO -H CO

• CI

I— 1 1— 1 t— (

6 09-5 09

■ T

c •

■ t~

- l-

CO rH

• CO

• CI i-H i-H i-H CO

fr' 09-0 '09

• i-

i i— 1

• c

6 CI

■ CI

• in

• • CI

i-H

6 65-5 '65

• c

3 CI

■ c

11 CO

rt rj<

■ t^

• CI CI CO

^H

f 65-0 65

• ^

i 'I1

• t-

- i-h i— i

Tf

•CI 1— 1 • »-H

I-H

6 ' 85-5 ' s'5

• c

3 CI

• 0!

3 ■ rf

CO

• CO

^'85-0'85

• c

3 •*

• c

3 rH t*<

■*

i-H r- 1 rH

6 Z5-5 Z5

< 0O

■ CO

CI

CI

CI

«-H

f ' Z5-0 ' Z5

• c

1 H rH • C

1

• ^h <N

CI CI i-H

6 95-5 ' 95

• CO

• ■ CO

1—4

1—1

fr' 95-0 95

• 1—

CI CI

• CI

6 55-5 ' 55

• C

• rH

• CI

1— 1

^H

*-H

f 55-0 55

• es

-r -r

• I-H

*-^

6W-SW

■ 1—

■ ^-

: ~

*-H

ffr9-0"*9

r- ( O ^H

; ■H

6 85-5 £5

• i—

■ rH CI

f £5-0 85 |

I— 1

: rt

■ ^H

6S5-5 S5 |

CI

• CI CI

f'ZQ-O'-ZQ |

CI

• CO

6' 15-5 15 |

CI

■ i-H CI

tT5-0I5 |

6 05-5 05 |

f 05-0 05 |

<N

6 6fr-5'6fr |

f ' 6^-0 ' 6$

6 ' 8^-5 • 8fr

r-(

• 1— 1

as a as

■cm 03 3 o3 U (h OO

9 Fi cut. X race B or C . . 9 F2 cut. X race B or C . . cf cutleri o71 race B

do O O mm 93 03 O C3 C3 cj XX o o

33 O 03 h X 03 o

m e 33 - X a 43 < - o

pq 03 CJ 03 t- X a 03 H <

- 03 03 X f 03 H M ■b

09 c HH o

d IH o pq 0) 03 03 X 03 03 i — i Ph* o

6 o pq 03 - (H X 09 03 M (N o

c3 O "b

dc (h o fflf 03 03 03 Ih X) 03 03 < 1— 1 H

J -i 0 Q 13 D a < H N H

INHERITANCE IN CJUINEA-PIGS.

-

Ili'U ■

• 55 10 56 11 60 11.45

c o c "0 X OS t- O BC — 1 ~ N i— « i— i i— «

O "5 -r © OS "* •-H

i« o OS •»* i-H i-H rH i-H

9.80 6.30 s 7ii

16.40 12.10 11.80

■111 ...l\

33 . 26

co co cs -r - I~ CI —« CO h- 'O CO CO CO CO

CO CO CO t^ CO CO

o t- **< 00 OS 00 CO CO

36.28 37.13 35.63

38.00 39.26 38.73

I^JOX

-r co — : i

I - CO CO 1" CO CM CM

00 i-H rH T»l

t- CO CM IO

CO t^ f i-H

O00N rH rH

■ a \ E — C CS 00 E V — 1 I ■K t. 1 _

i-H

I P| 0 ri

e n 9 it

i-H

» it (i it-

CM CM

i-H

60I-9M

iH

CO CO

• rH

f- Of-fl Of

»-H

CO tH

i-H CM rH

6 68~9 68

i-H

CO T*

U5 ■*

.

rH rH CO

1 BS n 68

CM CO ■

rH CO

cm u:

■

CM • CO

8 v8~9"88

CM CM •

CO ii

r"

CM CM CO

t '88 0 88

r-i ui

co «:

rH t-

i-H i-H

rH (M

6 £8-9 Z8

1— 1

CO "3 CM

i-H r-

m c rH

■ T-H •

• CM

t .VS-OZS

rH

NNh

cm ic

• cc

i-H CM rH

•

f^ rH

e 9€ .98

CM CM •

i— I CM iH

CM t".

rH CM CM

1 <»8-0"98

CM rH CM

CO Tjl rH i-H

O i-H

i-H rH CO

CO

6 98 V 98

i— t cO

HHIO

CM

CM

r"

1—t

• i-H

f- 98-0 v.

CO CO •

CO

^H

CM

<N

CM

6 n:-9f£

CO iO CO

"*

OS

IQ

fi-HK

•*}< rH CM

CM

■ rH

6' 88-9 "88

■* CO »o

^H

• CM

*'88-0"8S

CO • os

6 Z8-9 SS

l-H (— t ^H X

>H

;-

*'se-o-ze

I-H

i-H

6 18 -'.■ IS

o»

■ i-H

CO

fr 18-0 [8

i-H

■ ^H

6 08-9 08

CM

• <— (

CN

roe oos

(M

6 <w- 9 62

CM

i

C

1 - n c

. i . c :c ■ c c ' : ) . u

5 O . •- a « <X - ^J 3 3 - c - e* - Uh >►<>

* 1 M

: P i

jc . i . i :p ■ < ■ < ■ • . « i> h

- - 2 C qp u c J C 9 - H 1 K ) J -* a £&

j 3 p 3 < 1 1 - ) < j i 3 J • M H r D C

: P 3 J < 0 - r g H n't X c

h e IS J a -• < ) y V -1 CI n t > »

qp IS - - f -

q D J < r IS D C

'I. ' t ; < :p • < • < • < . i : > • i • < it » c

3c H ( 5 ( 3P IS C i < i ■ - ■■ <> a < j < - - hP - C

) H 3 q D < H - < H »- >r ^

. :p • • • * > ■ * u

5c - t a ; qp; IS 0 J C 3 c H i < > a c - - £ps

) 3 1 > i < s 3 H ~» 3

SIZE.

51

to Si

1

1

a

no

"5

I

S

CO

•uori

12.05 12.50 10.35 15.60 8.20 10.70 9.80 14.20 13.35 15.30 10.05 11.80 10.75 9.05 10.85 16.10 17.60 17.80

•m38j\[

38.45 41.16 42.63 40.38 38.77 42.39 43.57 41.32 44.07 42.95 45.16 43.15 42.64 43.49 42.06 42.90 44.63 43.44

•pn<>x

CM © CO CO t- CM Ci ■* COi-H iO •<# GOt^-* Ot^t-^ HNNM iO —l CM CM tjh NlO ,--, S iH

3 o a s 3 O* 5 (H V-i a •— . in V J 'a _g 0) CO 55 e8 5

t'ZH'ZJ-

»— 1

• i-H

6'9H'W

CM

i-H i—l

• i-H

^'9^-0'9^

■ tH -r-l CM

i—l

6'fiHfi^

CM CM 00

i— t 1—1

rS^-O^

1—1

i— 1 T— i io i— i

• f— t

r-l ■ CM

6^-9'^

■ rH HH

Tf CO • IO

r 1-1 tH CO

* ff-0'ff

• f— I

• CO CO

"# »0 IO CO i-H T-l

68fr-9'81?

i— t CO i— 1

• CO IO

i-H CO r-l O • i—l i — 1 CMi-H i—l

^"8^-0'8^

l-H T}H

tPCM'-i CO iO CM t~» cOi-i- CMi-HCM

6'2^-9^

l-H CM CO

CM "# CM CO Tt< i— 1

00 CM CO CO

CM

^2^-0^

>o ■*

O

■ ■ CO CO

CI

l-H

■*

61^-91^

IO CM CO

Ci

CD

CO

■>*

T*l NH ■

nH'i^

co cm co

CO

■*

>o

1— 1 1— 1

t-H

i— l

6'OH'O^

CM (N CM

NHUJ

1— 1

T-H T— 1

i— I

^Of-0'0*

CO

CO ■-!

CM

CM

1— 1

t-H

6 68-9 68

CO CM

CO • i-H

i— l

i— 1

i—l

^•68-068

CO •

iO CM

CM

688-9'88

i-H CO

rt CM

f-88-088

i— 1

CM

6'Z8-9ZS

-

CM

^Z8-0Z8

i—l

6"98-9'98

i— 1

f 98-0 ' 98

i— <

6 98-9 '98

i-f

co c 03

3 3 o o

pq o O

C c m 01 c (4 X 13 o

6 u o pq o o 03 tn X o fa O

a *^ 3 o

pq CO o t-

Q 0 pq co o 33 J- X = u fa"

d c pq CO S3 X c

pq c V X 6 CO H < fa o

PQ CO a a u X D <D i- < ft O

pq «D C 03 t-i X 6 CO <! fa 'b

pq CO a o3 ■- X 6 V <

1 03 u l-H O

a t- o pq CO o s X o fa 0»

c u c pq CO Bl X 99 O l-H c. 0»

oi o l-H

d c pq CO u C3 X 03 c r— (

d c pq CO o o3 « X 03 c l-H fa ^5

i

INHERITANCE IN GUINEA-PIGS.

,1, from the oran, but (2) thai a physiological growth stimulus (not hereditary results in I from the fact that the zygotes produced are formed bj the union of gametee from very dissimilar races, and (3) that the increased F \ Igor is largely, but not entirely, lost in F2. No evi- dence is Found thai it persists in full force in any F2 zygote (with one possible exception), since the upper half of the range of the Fj zygotes u slmosi completely wanting in F,, while the absence of any appreciable increase of variability in Fa shows that any increased vigor due to the cross which persists into Y, persists also very generally among the

otee of that generation, so that practically all are changed in the same Bense and in like amount; otherwise increased variability must result, irrespective of whether size inheritance occurs other than by complete blending.

HYBRIDS OF THE AREQUIPA <? 1002.

( tosses between the Arequipa <?1002 and females similar to those of race B, but averaging a little larger, have yielded an extensive and vigorous race for the study of size inheritance. Among the animals of this crossed race the mortality has been comparatively small, so that good numbers are a\ ailable. The male 1002, sole male ancestor of this race, is b< ill alive, so that his bone measurements are not available; but a female. 1001, secured in the same cabin in Arequipa in 1911, lived unt il fully grown and her bones are available for comparison. Further,

-on of d"1002 and 9 1001 lived until fully grown and his bones also are available. From the measurements of these two and a comparison of the empirical ratio of female to male measurements in the other races Btudied, it is possible to arrive at estimates of the racial size of the Arequipa stock which it is believed are fairly reliable. These are given in table 32, where it is further assumed that the racial size of the animals mated with cf 1002 was substantially that of race B, measurements of the latter 1 icing given for comparison. But whether these assumptions are Bound or not does not affect the validity of the observations on the 1 and 1' hybrids from this cross, which are valuable as regards their interrelations, for the numbers of adult individuals are considerable 13 Fj and 77 1\ animals) and the mortality among them is small. Ft in this experiment (tables 29 to 31, rows 9-12) regularly exceeds the a— umed mid-parental measurement, as in the crosses previously ndered; 1' is in all cases close to the mid-parental, being slightly greater in three cases and slightly less in three cases. As regards the relative variability of the two generations, the standard deviations indicate thai the Fa females (as compared with those of Fx) are con- siderably more variable in skull length (though scarcely more so than race B and are -lightly more variable in skull width and femur length. The male 1 ' , hybrids differ very little in variability from the F, hybrids, the standard deviations being slightly greater in skull measurements but less in femur length.

SIZE.

53

THE ICA HYBRIDS.

Crosses with the lea race were made principally by a male of race C whose measurements are known and which slightly exceed the averages for race B males; but a few crosses with the lea race were also made with race B females (mated with lea males). The measurements given in table 32 for the mates of the lea race are a mean between the meas- urements of races B and C. The standard deviation of the mixed parents should of course exceed that of race B alone, which should increase the variability of Ft and F2, but should not alter the relative variability of these two, since the race B and race C hybrids were bred

Table 32. — Statistical constants derived from tables 23 to 25.

Races

9 cutleri

9 race B

9 Fi, cutleri X B or C . . . 9 F2, cutleri X B or C . . .

cf cutleri

cf race B

c? Fi, cutleri XBorC... cf F2, cutleri X B or C . . .

9 1001, Arequipa

9 race B

9 Fi, Arequipa X race B . 9 F2, Arequipa X race B . C? Arequipa (estimated) . .

cf race B

C? Fi, Arequipa X race B . d1 F2, Arequipa X race B .

9 lea

9 races B and C

9 Fi, lea X race B or C . . . 9 F2, lea X race B or C . . .

cf lea

cf races B and C

<? Fi, lea X race B or C . . . cf F2, lea X race B or C . . .

No. of indi- vid- uals.

10 28 24 33 7 63 26 24

28 18 41

63

27 56

8

7 14 10

17

Skull-length.

Mean.

51.55 58.14 57.70 54.35 52.91 60.35 61.20 57.26 61.10 58.14 61.92 60.44 63.11 60.35 64.40 61.86 57.45 59.00 60.20 58.84 58.10 62.00 62.20 62.17

Standard deviation.

13.50 19.75 16.85 17.20 9.45 15.05 12.15 20.00

19	75
12	10
20	65
15	05
15	15
17	75
10	00
10	00
11	20
22	65

23.55

24.55

Skull-width.

Mean.

30.84 34.68 35.24 33.26 31.63 36.33 37.79 35.24 39.70 34.68 38.37 37.35 41.20 36.33 39.40 38.87 36.28 35 . 00 37.13 35.63 38.00 38.00 39.26 38.73

Standard deviation.

9.35 10.56 11.60 11.45

6.80 11.90 11.70 12.05

10	56
9	40
It	05
11	90
11	95
11	40
9	80
6	30
8	.70
16	.40

12.10 11.80

Femur-length.

Mean.

38.45 41.16 42.63 40.38 38.77 42.39 43.57 41.32 45.50 41.16 44.07 42.95 46.50 42.39 45.16 43.15 42.64 42.00 43.49 42.06 42.90 42.50 44.63 43.44

Standard deviation.

12.05 12.50 10.35 15.60

8.20 10.70

9.80 14.20

12.50 13.35 15.30

10.70 10.05 11.80 10.75

9.05 10.85 16.10

17.60 17.80

entirely distinct from each other and are only tabulated together to secure greater numbers. Again in this cross we are confronted with the same phenomena as regards the skeletal measurements: (1) in Fx a substantial increase in size (least in skull length of Fx male hybrids) ;

(2) while in F2 a return is made toward the mid-parental (mean of the races crossed), in 4 of the 6 measurements it is closely approximated;

(3) the standard deviation meanwhile alters little, not enough to have significance; the lea crosses are of particular interest because the races mated are of nearly the same size. The phenomenon of increased size in F: followed by a prompt loss of the increase in F2 is here observed

INHERITANCE IN GUINEA-PIGS.

ctlv as in the crosses between races of widely different and heritably different sises, but without indication in either case that the size inheri- ]£ other than a simple and permanent blend.

THEORETICAL EXPLANATIONS OF SIZE INHERITANCE AND OF BLENDING

INHERITANCE IN GENERAL.

\\, conclude therefore that, so far as present knowledge goes, the statement made in 1909 that size inheritance is blending and does not mendelise still holds. This does not preclude the possibility that in <ial cases mendelizing factors may exist which affect size. For example, in man brachydactyly is due to such a factor, a simple Mendelian dominant, as was first shown by Farabee (1905), and has been confirmed by Drmkwater in the case of three separate English families. Tbiscbaracter involves a shortening of the skeleton generally, but of the digits in particular. It is transmitted only through affected individuals, the normal offspring of affected individuals producing only normals. Professor James Wilson has stated that the Dexter-Kerry cattle of Ireland differ from ordinary Kerry cattle by a similar men- delizmg factor. If one were to restrict Ins study of size inheritance to cases such as these, he would reach the conclusion that size inheri- tance in general is Mendelian, a wholly mistaken idea. (See Castle, 1914.) Such cases among animals are distinctly rare. Among culti- vated plants they seem to be somewhat commoner, so that many of the inherited size differences studied by botanists involve such factors. One of the commonest of these is involved in the difference between normal (tall) and dwarf habit of growth, a case demonstrated by Mendel for peas in his original experiments; but it is more than doubtful whether Mendelian factors produce the differences in height observed among different races of tall or of dwarf peas respectively. The same is true concerning differences in size or shape of seeds and fruits, as described by Emerson and Gross. It seems almost certain that Mendelian factors are involved in many of the cases studied, but associated with other factors not Mendelian, possibly merely physio- logical, which render the results extremely complex and the variation seemingly continuous in character. To have shown that size inheri- tance ie occasionally affected by Mendelian factors is not by any means to have demonstrated that all size inheritance is due to Mendelian factoi The physiological increase of size due to the crossing of unre- lated races is a fact of far greater economic importance to the animal breeder than the existence of any Mendelian factor affecting size that Par been demonstrated. rhe question may be raised, how are we to account for the increased liability of V as compared with Fu if this is not due to segregation i recombination of multiple factors, as assumed under the Nilsson- 1 hie principle. 1 1 This would be sufficiently accounted for in the

size. 55

case under discussion by an unequal persistence, among the F2 zygotes, of the increased growth stimulus observed in F! and due evidently to the act of crossing, not to inheritance. (2) Increased variability in F2 would also result if a blending occurs in F1? which is imperfect, so that the gametes formed by the Fx individuals are not all the exact mean of the parental gametes, but fluctuate around that mean.

What may we imagine the germinal basis of a blending character to be? Perhaps some substance or ferment which varies in amount, larger amounts producing larger results. If a 5 per cent solution of cane sugar were poured into the same dish with a 10 per cent solution and then sam- ples were dipped from this before the two solutions had been thoroughly stirred together, it might very well happen that the samples would not be of uniform strength. Any other result would be surprising. A char- acter genuinely blending in heredity might be expected to behave in this same way, the quantitatively different conditions found in parent races not blending perfectly in a single generation of association together in an Fi zygote, which therefore would produce gametes less uniform in character than those of the respective inbred parent races. The multiple factor interpretation of size inheritance, besides being superfluous, meets with this serious logical difficulty: If we suppose the difference between two races to depend upon a certain number of independent factors whose action is cumulative, then a less difference must be due to fewer factors, and the fewer factors concerned in a cross, the more obvious is the segregation. But we do not find it easy to detect segregation when races are crossed which differ little in size; the general result is the same as when races are crossed which differ widely from each other. It is difficult to detect any evidences of segregation unless the parent races differ widely from each other, under which condition, if multiple factors are involved, complete segregation should occur least often.

On the whole, the hypothesis of quantitative variations in a blending character presents fewer difficulties as an explanation of size inheritance than the hypothesis of multiple unvarying segregating factors. It is to be preferred on the ground of simplicity alone, but it also accords better with the results obtained in other fields. Jennings now finds, contrary to his earlier observations on Paramecium, which Calkins and Gregory were unable to confirm, that size is a character varying even in asexual reproduction, within what would be a "pure fine" if the theory of factorial constancy were true. My own observations of rats and other rodents (Castle, 1915) may be cited to show that even single Mendelian unit characters are quantitatively variable. If this is so, the hypothesis of multiple factors as a general explanation of variability is quite unnecessary and so should be discarded.

PART II

AN INTENSIVE STUDY OF THE INHERITANCE OF COLOR AND OF OTHER COAT CHARACTERS IN GUINEA- PIGS, WITH ESPECIAL REFERENCE TO GRADED VARIATIONS.

By Sewall Wright, S. D.

COLOR AND ITS INHERITANCE IN GUINEA-PIGS.

The experiments described in the following paper were carried on at the Bussey Institution of Harvard University, between September 1912 and August 1915, under the direction of Professor W. E. Castle. A large number of stocks of guinea-pigs and wild cavies, containing an extensive assortment of variations, were available throughout the experiments, and furnished excellent material for studies on inheritance. The writer wishes here to express his gratitude for the privilege of using freely this material and for the constant encouragement and assistance which Professor Castle has given.

SKIN, FUR, AND EYE COLORS OF GUINEA-PIGS. COLOR OF CAVIA CUTLERI.

The fur color of Cavia cutleri, the probable ancestor of the guinea-pig, is of the agouti type found in most wild rodents, as well as in many other wild mammals. (See plate 3.) The back and sides are slaty black, ticked with yellow (more accurately, cinnamon buff). An isolated hair is of a dull slate color at the base, becoming blacker toward the tip. Near the tip there is a yellow band some 2 or 3 mm. long. The extreme tip for 1 to 2 mm. is black. The belly is cream-colored (more accurately cartridge buff) and is sharply separated from the ticked sides. An isolated hair is pale neutral gray throughout its basal half and cream- colored in the remaining portion. Cavia rufescens of Brazil has a similar ticked coat, but differs in showing less ticking on the back and sides and often in having a ticked belly not sharply separated from the sides. The general appearance is darker. Tame guinea-pigs show a great variety of colors and color patterns and also deviations from the dark skin and black eye color of the wild species.

MELANIN PIGMENT.

The coat colors of mammals are largely due to granular pigments of a kind known chemically as melanin. The pigment in the hair is found principally in the walls of air-spaces in the medulla, but to some extent in the cortex, as described by Bateson (1903) in mice. Melanin pigments are also found in the skin (principally in the epidermis) and in the iris and retina of the eye. A deficiency of pigment in the retina is revealed by a red reflection through the pupil.

PRIMARY CLASSIFICATION OF FUR COLORS.

Three qualitatively distinct melanin pigments are generally recog- nized in mammals, viz, black, brown, and yellow (Bateson, 1903). There are reasons, however, for regarding black and brown as more closely related to each other than either is to yellow. Black and brown granules are acted upon similarly by most hereditary factors which act

59

imii:i:i i \\< i: in GUINEA-PIGS.

,,n either. \ allow pigment, on the other hand, is acted upon very dif- ferently from black and browD by many factors. Accordingly it will be convenient to use a term to include both black and brown pigments, dark pigments. The fur colors fall naturally into two groups, the .lark and vrllow colors, characterized by the predominant presence of dark and yellow pigments respectively.

YELLOW GROUP OF COLORS.

In the ycll-.w group of colors the one of highest intensity is a rich yellow-orange, which matches quite well with ochraceous tawny in Ridgway's <-olor charts (1912). There are all gradations from this ochraceous tawny through cinnamon buff and cartridge buff to white. In this paper it will be more convenient to use the conventional names, red, yellow, and cream, for these grades. In grading the guinea-pigs, three Bamples of hair have been used as standards of grades called red . yellows, and cream^ respectively. White is considered to be creams. All of the yellow colors in guinea-pigs fall into this series, as far as known. In mice, however, Little (1911) has shown that two dilution series between red and white can be distinguished. There is

eries from red to cream resembling in appearance (though not geneti- cally) the guinea-pig series. Another series (the "dilute" reds, yellows, and cream- 1 has a peculiar streaky appearance. The physical relation between these two series is probably similar to that between the sepia

i blue types of dilution among the dark colors, wrhich is discussed

below.

DARK GROUP OF COLORS.

Among the dark colors there are at least three distinct series: ( 1 ) There is the series of neutral grays, passing from black to white. - h colors are shown by the blue rabbits, blue mice, and maltese cats. There are no tame guinea-pigs known whose colors fall distinctly into this series ; but the dull black of the wild Cavia cutleri, especially on the belly, i- a neutral gray quite free from any brown. Examination of the hair of the blue rabbit under the microscope shows dense black pigment masses alternating with colorless spaces, a condition described by Miss Sollas I 1909) in the hair of the blue mouse and apparently comparable to the clumped condition of the black pigment in the Feathers of blue pigeons, described by Cole (1914).

There is a Beries of grades from black through dull brown and tow-color t.i white. This -eries is shown by dilute black guinea-pigs. The various shades of human hair, from black through brown to tow- color, match Bamples from this guinea-pig series very closely. The increase in quantity of pigment in this series in passing up from the lower grades is accompanied by a change in quality. Yellowish-brown pigment gives way to Mack. Dilution of this sort is produced inde-

COLOR. 61

pendently by different factors, the combination of which gives doubly dilute colors, which may still be classed in the same series. Dilute guinea-pigs of this series have been called blue in the literature, but the name is as inappropriate as it would be applied to human brown hair, and, moreover, tends to confusion with the very distinct type of dilution of the blue rabbit. In this paper the colors of this series will be called sepia. Grades of dilution have been represented by numbers, as in the yellow series. White is considered as grade 16. Grading has been done by comparison with standard samples of hair, the colors of which are defined in terms of Ridgway's colors at the end of this section.

(3) The most intense grade of this series is a rich dark brown, such as is found in chocolate guinea-pigs, mice, and rabbits and in liver- colored dogs. This color is not very different from sepia4, but is some- what warmer and less dull. As noted by Miss Durham in the case of brown mice, there seems to be a complete absence of black granules, but a large quantity of brown granules. No intergrades between this brown and black are known. There are dilute browns, each corre- sponding closely to a color in the sepia series. They are often difficult to distinguish from grades of sepia in isolated samples of hair. On the animals, however, the browns seem conspicuously richer than the sepias. There are, further, correlated differences in skin and eye color which are even more conspicuous.

Most guinea-pig colors can be matched fairly well in the sepia, brown, or yellow series, but one other class of variations must be noted. The animals have been graded by the color near the tip of the hair, but while in some blacks, sepias, browns, and yellows the hair is nearly uniform, in most cases the base is much duller than the tip. This gives a somewhat streaky effect to the fur. In the case of dull blacks of this kind, the color at the base of the fur is usually between a neutral slaty black and a dark sepia.

SKIN COLORS.

The color of the skin usually corresponds roughly to the color of the hair which comes from it. Where the fur is thick there is very little pigment in the skin, while exposed places (as ears and feet) are often very strongly pigmented.

Where the fur is yellow the skin in exposed places shows an orange- yellow color, usually with considerable admixture of black. On most of the body the skin is white, with occasional orange-yellow spots. The dilution of fur color is accompanied by dilution of the skin color.

Where the fur is black the exposed parts of the skin are very black, while the rest of the skin is dull black. Where the fur is of the sepia series the color of the ears and feet depends much on the genetic factors responsible for the dilution of the black. In the sepias of the albino

INHERITANCE IN GUINEA-PIGS.

,,-. and feel are quite black, often with intense black blotches. Even in albinos, where the fur is nearly pure white, the ears ..,x be black. In the pink-eyed sepias, on the other hand, th« ery little pigmenl anywhere in the skin.

Brown fur goes with a uniform brown color of ears and feet very different from the dull black of sepias of corresponding intensity of fur color. Dilution in the skin accompanies dilution in the fur. The different skin colors are very conspicuous in animals with tted fur. In these it is easy to find places where the skin spots do not cone-pond exactly to the fur spots. White fur may arise from colored skin and yellow fur from black skin, but the reverse cases do not seem to occur.

EYE COLORS.

The iris ami retina usually contain black and brown pigment. Where there is reduction of pigment in the iris, the pigment tends to disappear first next to the pupil, leaving a dark outside ring. Decreas- .n- grades ol retinal pigment are most easily recognized by the apparent color of the pupil. In black eyes the pupil appears black. Occasion- ally a red reflection can be obtained in strong light. In brown eyes a dark-red reflection is easily obtained by holding the guinea-pig away the light. In the red eye the pupil looks red most of the time and the inner ring of the iris often transmits red light. A pink eye transparent iris and a pink reflection is visible through both iris and pupil in all lights.

The following summary shows the color terms to be used in this per, with their nearest equivalent on Ridgway's color charts (1912). The numbers 15'?, etc., refer to the position in Ridgway's system. For purposes of convenience in defining the color factors, white is included as a member of each color series as well as in a class by itself. In some cases white may be shown to represent extreme dilution of a particular color; in other cases it stands in no relation to particular col<

DEFINITION OF FUR COLORS BY RIDGWAY'S CHARTS.

1 Pigment absent because of factors not belonging to a dilution series. It'/.

Pigmenl present, or absent only because of factors demonstrably belonging

to a dilution sen n. Yellow group.

Red0= 15't, ochraceous tawny.

fellows -16"6, redder than cinnamon huflf, 17"6.

uii, - 19"/, cartridge buff. White. / f>nrk group. (]) Black.

ty black =dark neutral gray. blue ■ neutral gray. White.

COLOR. 63

b. Dark group — Continued.

(2) Black.

Sepia3 = 16"'w, warmer and darker than clove brown, 17"' m. Sepia6 = 16"'Z, warmer and lighter than clove brown, 17'" m. Sepia9 = 17"'% hair brown, slightly purer, however. Sepial2 = 17""6, light drab, somewhat purer. Sepia15 = 17""/, pale drab gray, somewhat purer. White.

(3) Brown =15" m, bister, 15"m, but somewhat warmer and duller. Brown3 = 15"% between army brown, 13"% and buffy brown, \T"i. Brown6 = 17"'fe, somewhat duller than avellaneous, 17'"6. Brown9 = 17""/?

White.

DEFINITIONS OF EYE COLORS.

(1) Black: black iris and pupil.

Dark red : black iris, dark-red pupil in favorable lights. Red : partially transparent iris, red pupil in most lights. Pink : transparent iris, pink reflection through both iris and pupil.

(2) Brown: brown iris, dark-red pupil.

Brown-red: partially transparent brown iris, red pupil. Pink: as above.

HEREDITY OF FUR AND EYE COLOR.

COLOR FACTORS OF GUINEA-PIGS.

Considerable work has been done on the inheritance of color varia- tions in guinea-pigs. The numerous colors which have been listed and several patterns in which these colors may be arranged have been found to be due in the main to relatively few hereditary factors. Some of these factors determine effects which are very easily defined. Thus, any guinea-pig which is homozygous for factor Ca is an albino with pink eyes and white fur, regardless of the presence of any combination of other known factors. On the other hand, certain factors determine nothing except in combination with other factors. Factor E may be present in guinea-pigs of any known color variety whatever. It can only be said that its presence is a necessary condition for the develop- ment of more than a trace of dark pigmentation in the fur. The color which results from a given combination of factors can be made clear most easily by classifying the factors into a series of groups. The following classification is based upon the factors in the rodents which have been most studied, viz, guinea-pigs, mice, rats, and rabbits.

CLASSIFICATION OF COLOR FACTORS.

1. Factors which affect the distribution and intensity of color largely irre-

spective of the kind of color.

A. Factors which govern the distribution of color as opposed to no color

(white) in patterns in the fur, in individual hairs, and in the eyes.

B. Factors which govern the intensity of general color development

within colored areas of fur and eyes.

2. Factors which govern the differentiation between yellow and dark colors

in colored areas of the fur.

3. Factors which determine the kind of dark color in the areas with dark

pigmentation in fur and eyes, without influence on yellow areas.

,,| INHERITANCE IN GUINEA-PIGS.

COLOR VS. WHITE (I A).

Probably dilution of the type of the blue and dilute yellow mice and rabbits' and maltose cats belongs here, rather than in IB, since the effect Beams to be due to the distribution of pigment within the indi- vidual hairs rather than to any effect on the actual pigment granules. Most of the factors which belong in this class, however, are those which determine patterns of white as opposed to areas which are colored under most combinations of other factors. In this class are such fac- tors as mi the one hand determine a self-colored coat, and on the other black-eyed whites, as in mice; white patterns, as in hooded rats, Dutch and English rabbits; or scattered white hairs, as in silvered guinea- pigs. In cases where several independently inherited white patterns have arisen it is evident that there can be no single factor which alone determines self. The "self" allelomorphs of the white-pattern factors can merely be defined as conditions for self. Where more than one white-pattern factor is present in an animal, combination patterns are produced.

( lcar-cnt Mendelian factors which belong to this group are known in mice, rats, and rabbits, but none have been isolated in guinea-pigs, although irregular blotching and silvering with white are common. The >vml >< .1 Z will be used to represent an assemblage of unanalyzed factors.

1 . an assemblage of unanalyzed factors which determine white spotting. INTENSITY OF GENERAL COLOR DEVELOPMENT (IB).

In this group fall albinism and its variations. These factors affect all color, but not wholly irrespective of the kind of color. There are several peculiarities which are discussed more fully in a later section (page 70). The most important is the fact that the level of intensity of the color factor at which yellow can develop at all is higher than the threshold for black or brown. This does not affect the differentiation of the fur into yellow and dark pigmentation areas by factors of group 2, but involves the result that with certain albino series factors, yellow areas appear white, while dark areas are quite strongly colored. Indeed, in albinism itself, dark pigmentation areas can often be distinguished from yellow areas by a slight sootiness in the former, absent in the latter.

' Detenninei the highest intensity of color of skin, fur, and eye which is to be found with a given array of other factors; dominant over Cd, Cr, and Ca, where distinguishable in its effects. In the following table, and in the similar tables under Cd,Cr, and Ca, given the ranges of intensity in the yellow, black, and brown series to which these colors develop when the factor under consideration is present. In the case of black and brown, factor P is assumed to be present. When p is present, black and brown undergo a two-fold dilution. P is also considered present in the case of i j • ■•< "](,r.

"» ellnw series red,, to yellows in guinea-pigs; yeUow3 to cream6 in Cavia

riilleri. Black series blacks to blacks- Brown series browns to browns-

1 •<• color — black, brown.

COLOR. 65

Cd. Determines an intensity of yellow distinctly lower than does C, an intensity of dark pig- mentation usually, but not always lower than does C, and an intensity of eye color rarely distinguishable from that determined by C. More or less dominant over Cr and Ca where distinguishable. (Wright, 1915.)

Yellow series — yellowo to creamy.

Black series — blacko to sepiay.

Brown series — browno (?) to browny.

Eye color — black, brown. Cr. Determines the complete absence of yellow, an intensity of dark pigmentation indis- tinguishable from that determined by Cd and an intensity of eye color lower than that determined by C or Cd- More or less dominant over Ca where distinguish- able. (Castle, 1914a; Wright, 1915.)

Yellow series — white.

Black series — blacko to sepias.

Brown series — browno (?) to browny.

Eye color — red, brown-red. Ca- Determines an absence of pigment, complete with yellow, not quite complete with dark pigments of the fur and skin, but complete in the eyes. (Castle and Allen, 1903; Castle, 1905; Sollas, 1909; Detlefsen, 1914; Wright, 1915.)

Yellow series — white.

Black series — white, dark smudges on nose, ears, and feet.

Brown series — white, brown smudges on nose, ears, and feet.

Eye color — pink.

DARK VS. YELLOW COLOR (2).

Factors of this group affect skin and fur color, but not eye color. In this group come the factors responsible for self yellows, tortoise-shells, and brindles, on the one hand, and self blacks or browns on the other, as contrasted with the ticked or agouti patterns of the wild rodents. Where more than one factor is present which determines a yellow pattern, combination effects are produced, such as in yellow-spotted agoutis among guinea-pigs. The following factors are known in guinea- pigs:

E. A condition for more than a trace of dark pigmentation in the fur; determines dark pig- mentation wherever yellow is not determined by other factors; dominant over e, found in the wild species, all agoutis, blacks, browns, etc., but very rarely in self yellows.

e. Determines the presence of one of the yellow colors in all colored areas of the fur, aside from a slight sootiness; responsible for the yellow in most self yellows, for the white in red-eyed whites, etc. (Castle, 1905, 1907, 1907a; Sollas, 1909; Detlefsen, 1914.)

A. Determines the presence of a yellow color in the light-bellied agouti pattern wherever there is dark pigmentation in which the yellow group ticking may show; dominant over A' and a, found in Cavia culleri and light-bellied agouti guinea-pigs, includ- ing the red-eyed silver agoutis, in which the agouti pattern is in white.

A'. Determines the presence of yellow colors in a more restricted agouti pattern than does A, a pattern usually characterized by a ticked belly not sharply distinct from the sides in color; dominant over a, found in Cavia rufescens and in ticked-bellied agouti hybrids which have rufescens ancestry. (Detlefsen, 1914.)

a. Determines the absence of yellow group ticking in hairs of dark pigmentation; found in blacks, browns, etc. (Castle, 1905, 1907, 1907a, 1913; Sollas, 1909; Detlefsen, 1914.)

Sj/. An assemblage of unanalyzed factors which determine the presence of spots of a yellow color, conditional on factors of group (1) ; found in black and yellow tortoise-shells, black, yellow, and white tricolors, and in some red-eyed black and white bicolors; probably responsible for an occasional self yellow, though never in the writer's experience.

INHERITANCE IN G TINEA-PIGS. \ARIATIONS OF DARK COLOR (3)

if this group are responsible for browns and pink-eyed sepias, as compared with blacks, in guinea-pigs; for browns and pink- , yed sepias in mice, and for the new pink-eyed and red-eyed dilute nations in rats. Where more than one factor of this group or of >up 1H determines dilution, combination effects are produced. Thus we have very pale sepias resulting from the combined effects of two independent dilution factors (CdCdPp).

B Determines ■ oolor of the black-sepia scries wherever dark pigmentation develops, including the eyes; has no influence where yellow pigmentation develops; domi- nant over b, present in the wild species and in blacks, sepias, albinos with black points, black-eyed yellows, etc.

6. Determines a color of the brown series wherever dark pigmentation develops, including the eyes; has no influence where yellow pigmentation develops; present in browns, brown-eyed yellows, etc. (Castle, 1907a, 1908; Sollas, 1909; Detlefsen, 1914.)

/' A inn. lit ion for intense development of dark pigmentation in the fur and for eye colors more intense than pink; not necessary for intense development of yellow; domi- nant over p.

p. Determines a low development of dark colors, i. e., below sepias; has no influence where y« How develops; determines pink eye color. (Castle, 1914a.)

TABLE OF FACTOR COMBINATIONS.

In determining the color which corresponds to a given array of factors the groups of factors must be considered in the order given. Table 33 gives a list of the color varieties corresponding to the combinations of Mendelian factors. At the top and left of the table are indicated, by symbols, the factors present in each of the varieties named in the body of the table. The color of spots produced by 2w and 2y are given below. Only the varieties marked with an asterisk have not yet been synthesized. These include the pink-eyed yellows and creams and a kind of pink-eyed white which is expected to be indistinguishable from an albint > in appearance, though breeding wholly differently. The pink- '1 brown series (bbpp) has not yet been produced and is not included. of the varieties have names given by fanciers which have been used in the literature. In this table, however, it seemed best to use a consistent scheme of naming, indicating at once the color and pattern. Agouti is used as the name for a pattern, the banding of hairs of pre- dominantly a dark color with a yellow color. The names preceding uti give the two colors in each hair. The following table of syno- nym- may lie useful:

Black-red agouti*" golden agouti, i-yellow agouti = yellow agouti.

■ •uti = silver agouti. v. ii-red agouti = cinnamon.

in agouti = light cinnamon. !>ia =bhie. Brown = chocolate.

brown eye (Castle), ruby eye (Sollas).

COLOR.

Table 33.

67

Factors present.	Fur.	Eye.
EA (agouti light-belly). EA' (agouti ticked-belly).	Eaa.	ee (A, A' or aa).
BPC CdCd.. CdCr... CdCa.. CrCr... CrCa- • • CaCa • ■ BppC... CdCd . . CdCr. . . CdCa.. CrCr...	Black-red agouti Dark sepia-yellow agouti Dark sepia-cream agouti	Black Dark sepia Do	Red Yellow Cream . .	Black. Do. Do. Do. Red. Do. Pink. Pink. Do. Do. Do. Do. Do. Do. Brown. Do. Do. Do. Brown-red. Do. Pink.
Dark sepia-white agouti Light sepia-white agouti White (dark points)	Light sepia	Do
Dark sepia Light sepia	White (light points) . Do. .
White (dark points). Pale sepia Verv pale sepia . Do	.Do.
Pale sepia-red agouti Very pale sepia-yellow agouti . Verv pale sepia-cream agouti. .	Red
* Yellow
*Cream . .
Do	Do	. . .*Do
Very pale sepia-white agouti . . Do	....Do Do	* White
. .*Do
CaOa . • bbPC CdCd . . CaCr... CdCa. . CrCr... CrCa... CaCa- - •	White (light points)	White (light points). Brown	. . .*Do
Brown-red agouti	Red. .
Medium brown-yellow agouti . Medium brown-cream agouti . Light brown-cream agouti .... Medium brown-white agouti . . Light brown-white agouti ....	Medium brown . ....Do	Yellow
Cream
Do . .
Medium brown . Light brown ....	White Do
White (It. br. points) .	Do
Factors present.	Sw.	2y	Sw2y.	Eye.
C CdCd.. CdCr... CdCa.. CrCr- - . CrCa...	White spots (clear)	Red spots Yellow spots . . . Cream spots .... Do	Red and white tri- color. Yellow and white. Cream and white tricolor. Do
.... Do .... Do
Do
...Do Do	1 White spots, / often sooty. (Albino)	Sooty and clear, white spots. (Albino)
C aCa- • •	(Albino)

HEREDITARY FACTORS AND THE PHYSIOLOGY OF PIGMENT.

The definitions which have been given for the hereditary factors are based largely on the colors as seen without a microscope. It would be very desirable, however, to correlate color factors accurately with the variations in quality and quantity of the actual pigments and ultimately with the physiology and chemistry of pigment formation.

Considerable progress has been made in recent years in the study of the chemistry of melanin pigments. The melanins are amorphous granular pigments found throughout the animal kingdom. A large number of researches have established the fact that substances which closely resemble the natural melanins can be produced by the action of

((S QfHBBITANCl IN GUINEA-PIGS.

MrUin oxidising enzymes on tyrosin and related aromatic compounds. rosiD is in unportanl constituent of protein molecules and there is much reason to believe that tyrosin and related substances are the chromogens from which the natural melanins are formed. Tyrosinase, .Ul enaione whirl, can oxidize tyrosin to dark substances resembling melanins, has been found very widely among animals, including the <kin- of mammals, a- will be discussed later. _

There have been many theories on the mode of origin of pigment in ,1„. oellfl Early observations indicated that melanin was directly extruded from the nucleus. Recent studies by Hooker (1915) on m vitro cultures of mesenchyme and epithelium of the frog indicate that melanin granules form in the cytoplasm but at the point of known greatest efficiency of the nucleus as an oxidizing agent. Thus, prob- ably chromogen (tyrosin or derivatives) is in the cytoplasm, while oxidizing enzymes are given off by the nucleus.

The color white in the fur of mammals is due to the absence of pigment. The theory of a white melanin seems effectively disproved rtner, 1910). A priori, the presence or absence of pigment might conceived as due either to a deficiency of chromogen or of enzyme. In line with the first view, Gortner (1911) found that the pattern in the elytra of potato beetles is due to a deficiency of chromogen. Fur- ther. ( uenot ' 1903, 1904), in the first attempt to correlate the facts of Mendelian inheritance with the physiology of pigment, suggested pro- visionally that albinos lack the power of producing chromogen, while the different colors which he demonstrated could be transmitted through albinos depend on specific enzymes. On the other hand, recent observations by Onslow (1915) demonstrate that absence of pigment in widely different cases in mammals depends on enzyme differ- ences. He found peroxidases in the skins of gray, black, blue, and brown rabbits, which produce a black pigment from tyrosin in the presence of hydrogen peroxide. In the skins of albino rabbits and mice ami in the white part of the Dutch pattern in rabbits, all recessive whites, he was unable to demonstrate a peroxidase, although there was nothing present which prevented the oxidation of tyrosin to a black pigment when tyrosinase was added. In the white of the English rabbit, a dominant white, he did find an anti-tyrosinase.

Finally, there is strong genetic evidence that albinism in guinea-pigs is not due to absence of chromogen. A diminution in quantity of chromogen should bring about the same diminution in quantity of all pigments, regardless of quality. But in red-eyed guinea-pigs (which we may consider as incomplete albinos, as they have an allelomorph of albinism (r) no yellow develops, leaving white areas where factors of group 2 determine yellow differentiation, but there may be nearly much black a- in normal guinea-pigs. Indeed, in the albino guinea- pi^s and Himalayan rabbits, there is no yellow, but some black.

COLOR. 69

The physical or chemical differences between the pigments of the different fur colors are not wholly clear. According to Onslow (1915) the pigments of black, brown, and yellow rabbits can not be distin- guished, physically or chemically, when isolated. At first sight this seems hardly possible with such apparently different colors. A result thoroughly in line with this view, however, followed the matching of fur colors with Ridgway's charts, much to the writer's surprise at the time. Ridgway distinguishes 72 hues passing from red through orange, yellow, green, blue, and purple, back to red. The yellows, sepias, and browns of guinea-pigs and human brown and red hair all matched colors near hue 17, "orange yellow," in the classification. The differences depended merely on differing amounts of black and white. Bateson (1903), indeed, found that yellow pigment is dissolved from hair by potassium hydroxide very much more rapidly than brown pigment, which dissolved more rapidly than black. This, however, might be due merely to size or density of granules.

This apparent qualitative difference in pigments has been attributed to several causes: (1) to variations in the chromogen acted on by a given enzyme, (2) to interruptions at different stages in the process of oxidation of a given chromogen, (3) to specific enz}mies which in each case can only produce a certain result once the action on the chromo- gen is begun.

Observations of Onslow indicate that for qualitative differences, as well as for the absence of pigment, enzyme and not chromogen differ- ences are responsible. He could find no peroxidase in self yellows, a recessive variation. There must of course have been some peroxidase at some time to produce pigment at all. Perhaps the apparent absence indicates a very low degree of stability in the yellow-producing enzyme. Again, grays differ from blacks by a dominant factor which causes yellow to appear in ticking over the back but white to appear on the belly. Onslow found a tyrosinase inhibitor in the belly and compared the case with that of the dominant white of the English rabbit. As grays differ from self blacks by only one Mendelian factor, it would seem likely that all of the changes in appearance — dorsal yellow tick- ing, ventral white — are to be ascribed to one physiological cause. If black is absent from the belly because of an enzyme inhibitor, it would seem likely that black is replaced by yellow in the dorsal ticking by the presence, for a certain period in the development of a hair, of the same enzyme inhibitor, which, however, is in this case merely an inhibitor of the black-producing reaction, not of the yellow. Reasons for which yellow can appear on the back of rabbits, but not on the belly, when black is inhibited will be discussed later. Thus, a recessive yellow and a dominant yellow-pattern factor are both due to enzyme, not chromo- gen, differences.

70

INHERITAN< B IN GUINEA-PIGS.

II. !id hypothesis that yellow, brown, and black are due to

interruptions of the normal process of oxidation at different stages is difficult to reconcile satisfactorily with the genetic facts in guinea-pigs.

[j brown and black pigments pass through a yellow stage, identical with the final stage of tin' pigment in yellow guinea-pigs, any factor which inhibits the development of yellow must a fortiori inhibit the development of brown and black. We have seen that with factor C, there is complete alienee of yellow pigment, but nearly full develop-

nt of brown and black. We find nearly the converse of this in the effect of factor p. When factor p is present, the development of brown and black is very greatly reduced without the slightest dilution of yellow. This indicates that neither is yellow a stage in the develop- ment of black nor black a stage in the development of yellow. The most satisfactory hypothesis is the third — that there are distinct enzymes which produce yellow and dark pigment.

There are a number of curious facts in connection with the albino series of factors in guinea-pigs which perhaps warrant further specula- tion As has been mentioned, Onslow has shown that albinism is due to the absence of tyrosinase in the skin (and presumably the eye). It

ms reasonable to suppose that the higher allelomorphs are quantita- tive variations in a factor which determines the power of producing tyrosinase. If this is so, we would expect to find that the different zygotic formulas could be arranged in a linear series with respect to their effects on pigments of all sorts. Following are the series with respect to black pigment of eye and fur, and yellow of the fur. (See plates 1 and 2.)

Formula.	Black eye.	Black fur.	Yellow fur.
CC CCd	Black. . . . ... Do	Black	Red. Do. Do. Do. Yellow. ( Iream. Do. White. Do. Do.
Do
	. Do.	Do
CCa	. Do . .	Do
CdCd. . . .	Do	Dark Bepia Do
r	. . Do . .
		Light Bepia Dark Bepia Light Bepia White (sooty) • • •
CrC	Red Do Pink

I he yellow Beries and the less accurately known eye-color series can

ranged in the Bame sequence. There is the striking difference,

however, thai the level of no pigment production is much higher in

yellow than ey.- color. The black of the fur agrees with eye color in

level at which pigmentation becomes evident— between CaCa and

COLOR. 71

CrCa — but the sequence can not be made to agree with either the eye- color or yellow series. CdCa is distinctly lighter than CrCr in black fur but distinctly more intense in eye color while, in yellow fur CdCa is above, CrCr below the threshold of any color. The effects could be explained by a complicated linkage hypothesis. We would need to suppose that there are separate series of allelomorphs acting on yellow, black of fur, and black of eye, respectively, and that Cr and Ca are complexes identical in the yellow-dilution factor, Cd and Cr identical in the black-fur-dilution factor and perhaps C and Cd in the black-eye- dilution factor. But an hypothesis according to which it is a mere accident that the factors which dilute yellow, black of fur, and black of eye are perfectly linked in inheritance can hardly be taken seriously. Another escape would be to suppose that our four factors, Ca, Cr, Cd, and C, are, indeed, variations of the same thing but not linear quantita- tive variations. However, it seems most satisfactory to the writer to attempt to explain the results on the basis of four quantitative gradations of one factor, which determines the amount of the basic color-producing enzyme, if it is in any way possible. Let us see what assumptions must be made to do this. First, it will be convenient to assume with Little (1913) that the basic color-producing enzyme (I) acting by itself on chromogen, produces yellow pigment. The addition of a second substance (II) makes it a black-producing enzyme (I-II). We will further assume that I is relatively unstable and must be produced above a certain rate (that determined by CrCr) in order to reach and oxidize the chromogen in the cytoplasm. United with II it becomes more stable and even produces some effect at the rate of production determined by CaCa. The next assumption is that above the thresh- old for yellow, I-II and the excess of I compete for the chromogen. As a result of partial displacement by the paler color, the intensity of black decreases just above the yellow threshold. CdCa seems paler (and somewhat browner) than CrCr. In the eye, no factor ever brings out a yellow color. There is perhaps never an excess of I here and the intensitj^ of black follows the normal sequence.

Summarizing, the hypothesis to which consideration of the physio- logical and genetic seems to lead is as follows :

(1) There is a basic color-producing enzyme (I) which acting alone on chromogen produces a diffuse or finely granular pigment which appears yellow. It is relatively unstable. Intensity of production and absence or inhibition in parts of fur and eye are determined by the various factors of group 1 — the albino series, " blue "-dilution factors, and recessive and dominant white-pattern factors.

(2) There is a second substance (II) which may unite with I to produce a more stable enzyme, which reacts with chromogen to produce a coarsely granular pigment which appears sepia, brown, or black. When II is present, I is stabilized to such an extent that pigment is

7 > INHERITANCE IN GUINEA-PIGS.

produoed at a lower rate of production of I than is the case of I alone. Uwve the level al which I alone produces yellow, the two kinds of eniymes, yellow- and black-producing, compete with each other for chromogen, producing a mixture of black and yellow, the relative importance depending on the rate at which II is produced. Because of the competition the intensity of black shows two maxima as production increases— one just below the yellow threshold and the other al maximal production of I. Intensity of production or inhibi- tion <>f II in patterns in the fur are determined by various factors group 2) which produce self yellow, yellow spotting, agouti, etc.

There is a third group of substances which, added to the dark- pigment-producing enzyme (II), affect the intensity of dark color pro- duced hut not the power of fixing chromogen in competition with the yellow producing enzyme. They have no effect on the intensity of yellow. In this group are the brown factors of mice and guinea-pigs, and perhaps rabbits and dogs, the pink-eye factor of rats, mice, and guinea-pigs and the new red-eye factor of rats, i.e., the factorsof group 3.

While based to a larger extent on the genetic facts in the albino Beriee in guinea-pigs, the hypothesis explains many cases in other mammals in the sense that apparently complex variations are reduced to a single physiological cause.

In rabbits, single Mendelian factors produce some rather complex variations. A single factor changes a self black to the gray color with a Yrllmv-ticked back but a pure white belly. Another variation changes a self black to a sooty yellow with a black belly. These varia- tions combined in one animal give a white-bellied clear yellow. How can each of these apparently complex color changes be determined by a simple physiological change? Let us suppose that in all rabbits I is produced strongly on the back, but so feebly on the belly that it is below the yellow threshold, but not so feebly that black is greatly affected. Let us suppose that II is likewise more strongly produced on back than on belly. A factor which tends to produce an inhibitor <»f II is added. On the back II (the black-producing enzyme) is inhibited in only a portion of the development of the hair, leaving yellow ticking. On the belly all II is inhibited, leaving white. The result is a gray rabbit. The other factor causes a general slowing up in the production of II. On the back this enables the yellow-pro- ducing enzyme to predominate in competition and sooty yellow results. On the belly— below the yellow threshold — what little black-producing enzyme does develop has no competition and only black can result. We p;et a black-bellied sooty yellow. The combination pattern can only he a white-bellied yellow. In many other mammals color phases ar»- found which can be explained as due either to variations in production of II or I. The red phase of the red fox has a white che The level of production of I is below the yellow threshold

COLOR. 73

but above the black threshold on the chest. Increase in production of II produces the silver phase, nearly self sepia in color, including the chest. The colors of the varying hare seem to be due to variations in production of I determined by environmental causes. The white winter pelage gives way in blotches to a white-ticked sepia; this gives way to yellow-ticked sepia as the intensity of production of the basic enzyme rises above the yellow threshold and in some varieties the full summer pelage is almost self red.

Many other cases could be given in which two color phases of an individual animal or the color patterns of closely allied varieties seem to differ in many respects and yet can be explained on the basis outlined as due to a single physiological change.

In the case of very complex color patterns, it is necessary to suppose that the power of producing the hypothetical enzymes I, II, or III may be distributed in quite complex patterns. But the hypothesis often gives a simple explanation for certain peculiarities in a pattern. In the tiger, the stripes on the back are quite intense yellow and black. The yellow stripes grow paler down the sides, becoming white on the lower sides and belly. The black stripes likewise grow lighter down the sides but at the point at which the yellow becomes white, the black stripes suddenly grow more intense, at least in some individuals, to become paler again on the belly. Again, on the legs, which are white on the inside, yellow on the outside, black stripes are visible on the white part, but either disappear completely or leave merely a streak of sooty red on the yellow part. All of this becomes intelligible if we assume that the basic enzyme (I) is produced at decreasing rates from back to belly and from outside to inside of leg, while the black-produc- ing supplement (II) is distributed in vertical stripes (horizontal on the legs). Two parallel stripes give a remarkable reproduction of the variation in black and yellow in the albino series in guinea-pigs. We have the same change from black and intense yellow to sepia and cream, then to darker sepia and white, and finally light sepia and white, illustrating the different thresholds for the appearance of black and yellow and the reduction in intensity of black above the yellow threshold due to the entrance of competition at this point.

DISCUSSION OF EXPERIMENTS.

MATERIAL. SYSTEMATIC POSITION.

Guinea-pigB belong to the family Caviidae of the hystricomorph division of rodents. There arc throe living genera of Caviidae: Doli- I teem., which contains the large Patagoniancavies; Hydrochcerus i. to which belongs the capybara; and Cavia Pallas, containing the small cavies. Genus Cavia is divided into two subgenera, Cavia proper and Cerodon V. Cuv., distinguished most conspicuously by the greater complexity of the molars in the former. Seven living species are listed under Cavia proper by Trouessart (1904):

C. rufescens Lund, a small dark Brazilian cavy with subspecies in Guiana and

Argentina. c. fvlgida Wagler, a Brazilian cavy probably closely allied to rufescens (Thomas,

1901). C. aperea End., B large pale-colored Brazilian cavy. ( . aeara Wagner, a cavy of Paraguay probably closely alhed to aperea (Thomas,

1901). C. mthri Bennett, a small pale-oolored cavy of Peru. ( \ tschudii Fitzinger, a large, richly colored cavy described from lea, Peru. C. inrcellus Linn., the tame guinea-pig, much larger than at least rvfescens and

cutleri.

DESCRIPTION OF STOCKS.

1 . >ur of these species are dealt with in the experiments to be described, viz, Cavia rufescens, C. cutleri, C. porcellus, and a type which is quite certainly that described as C. tschudii, although it is also quite certain that it is simply feral porcellus. Breeding experiments have been carried on with a fifth species, C. aperea, by Nehring (1894).

The C. rufescens stock was derived from 3 individuals received from Mr. Adolph Hempel, of Campinas, Brazil, in 1903. The history of tins stock is fully described by Detlefsen (1914). When received by the writer, most of the stock consisted of hybrids containing only from • rufescens blood. There were a few § and \ bloods and one ^ blood, 9 A68, which is still alive (August 1915) at the remarkable age of S years 1 month,1 a good illustration of the vigor of the first-generation hybrids. All of the pure rufescens stock has died out. The rufescens hybrids have been crossed with nearly all of the guinea-pig stocks to be described, and most of the color varieties may be found among them. The ticked-bellied type of agouti has been found only among them and in pure rufescens. c. rufescens was not completely fertile with the guinea-pigs Detlefsen, 1914). Detlefsen found that while the female hybrids were fertile, all of the male hybrids obtained were sterile. In the 1 rufescens, derived by crossing the females with guinea-pigs, the males were again all Bterile. Not until the |- bloods were obtained did

I ■ • ■! I >• tot* r L915, aged 8 years, :; months. — W. E. C.

MATERIAL. 75

a few fertile males appear. The percentage of fertile males gradually increased in later generations.

The Cavia cutleri stock was derived from animals captured by Pro- fessor Castle in Peru in 1911 . Like C. rufescens, these are much smaller than the guinea-pig. All show the agouti pattern. The color is described on page 59. Unlike C. rufescens, C. cutleri breeds freely in captivity and crosses readily with the guinea-pig. The male and female hybrids are fertile.

The lea stock of guinea-pigs was derived from 3 guinea-pigs which were obtained by Castle near lea, Peru, in 1911. They were as large as or larger than average guinea-pigs, and of a rich golden agouti color, very different from C. cutleri. Two independent color variations appeared at once in the pure stock, viz, black (aa) and red-eye (CrCr). Such variations are very uncommon among wild species of animals; e. g.. none has occurred within the pure rufescens or cutleri stocks. Both of these variations are found in domesticated guinea-pigs in Peru (Arequipa stock). From the description of Cavia tschudii, quoted in Waterhouse (1848) under the name C. cutleri Tschudi, it seems clear that our lea stock is the same as the former, which was likewise described from lea. In view, however, of the size, color, and possession of recessive color varieties found among tame guinea-pigs of Peru, there can be little doubt that they are feral porcellus.

The Arequipa stock comes from a pair of guinea-pigs brought from Arequipa, Peru, by Castle in 1911. He obtained them from Indians who had them under domestication. Owing to the early death of the only female, no pure stock could be developed, but numerous descen- dants have been derived from the original male 1002, a sepia-cream agouti with white and cream spots, demonstrated to be of constitution EEA%BBPpCdCr, and from a son of the original pair, male 1007, a yellow agouti with white and yellow spots, demonstrated to be of con- stitution EeAaBBPPCdCd. These were crossed mainly with the 4-toe and BW stocks, which are described below. For a full discussion of the origin and nature of the pure cutleri, lea, and Arequipa stocks, see Part I.

The Lima stock comes from 8 guinea-pigs obtained from Indians near Lima, Peru, by Professor Brues in 1913. These guinea-pigs and their descendants have only recently been crossed with other stocks. There have been no agoutis in this stock. The pink-eye and yellow variations, as well as white spotting (but not yellow spotting), have occurred in this stock. A pink-eyed red, the lowest recessive, of this stock is of constitution eeaaBBppCC . There were both rough-furred and smooth- furred individuals in the original stock.

The following stocks come from guinea-pigs obtained from fanciers by Professor Castle and have been maintained for several years at the Bussey Institution.

76

INHERITANCE IN GUINEA-PKJS.

SB Block. A Mock consisting exclusively of very intense blacks. So red or white Bpotting h;ls been observed among them. Unfortu- nately it is a stock of low fortility, and could not be used much to

advantage .

/.Ml' gtodk.— This stock has for years consisted exclusively of very intense blacks and very Booty albinos. The blacks occasionally show i few red hairs or a small red patch. This has been an extremely ful stock, among other thingB, furnishing albinos known to be geneti- cally identical with blacks, except for the albino factor. (Race B of

,r-tn< slock — This is a much-inbred stock, practically all the indi- viduals of which show four good toes on the hind feet instead of the normal three. This stock was developed by selection and inbreeding by Professor ( !astle (Castle, 1906). Most of the individuals are a dull black with dull red blotching and brindling and often with white spots. Albinos appear quite frequently and reds much more rarely.

Table 34. — Genetic formulae of stocks.

• k.	Color.	Roughness.
Mcndelian.	Unanalyzed.	Men- delian.	Unana- lyzed.
( :ivia cutleri ia rufeacens .... Ir:i	E \ B P C E A' B P C E A, a B P C,Cr E,e A,a B P,p C.Cd.Cr E.e a B P,p C E a B P C E a B P C,Ca E(e)a B P C.Ca E a B,b P C E.e a B P Cd,Ca e a b P Cd.Ca E,e A.A'.aB.b P C,Cd,Cr,Ca	(2w2y) 2int + 2w2y 2int + 2w 2int + (2y) 2int + 2w2y 2int - 2w2y (2w2y) 2int - 2int- 2w2y 2int±	r S r S r S R,r s R,r (S)s r s r s R,r s R,r S,s r s r s R,r S,s	2- .... 2- .... 2R .... 2R 2- .... 2- .... 2R 2+ ....
Arrquipa
Lima
BW
4-toe
Tricolor
• ia+cream Brown-eyed cream . . ' rufeaoena hybrids.

/ n ih, tricolor stock the fur is typically a patchwork of red, white, and black. Full-roughs, partial-roughs, and smooths occur among them. The writer has used many guinea-pigs of very mongrel ancestry, which, however, owe their partial rough coat to this stock.

Tl i i picHmdrcn am and brown-eyed cream stocks have been selected

for years for extreme dilution. The former stock consists exclusively

of Bepias, Mack-eyed yellows and creams, and albinos. The latter

siste exclusively of brown-eyed yellows and creams and albinos.

C of Part I.) In the tables, these together are called dilute-

Belection -tuck.

Table :;t shows the Mcndelian factors affecting color and roughness of fur which occur in each stock. Unanalyzed hereditary conditions which affed color and roughness are also included, prefixed by the symbol I. 2w and 2y. as has already been stated, mean hereditary

INHERITANCE OF DILUTION. 77

white and yellow spotting respectively. 2 int+ and 2 int— mean hereditary constitutions which intensify or dilute, respectively, the color associated with a given array of Mendelian factors. 2 + and 2 — in the rough column have a similar meaning with respect to the rough character. 2R means the presence of roughness of a different kind from that analyzed. Where a factor occurs only rarely in a stock, it is inclosed in parentheses.

PROBLEMS.

The inheritance of the discontinuous color variations which are known in guinea-pigs has been solved by previous work. After each factor variation from the wild type (Cavia cutleri) in the definitions of the factors the principal papers on the subject are given. The writer has been concerned mainly with an analysis of inheritance in the contin- uous series of variations by which each of the intense colors — red, brown, and black — grade into dilute colors and ultimately white. A second group of problems concerns the variations in the amount of yellow ticking in agoutis. The writer has worked with the agouti patterns of Cavia cutleri, C. rufescens hybrids, and tame guinea-pigs. The inheritance of variations in the rough coat occasionally found in guinea-pigs is discussed in a later section.

INHERITANCE OF DILUTION.

THE RED-EYE FACTOR.

The experiments with dilution have become closely associated with experiments with certain imported South American stocks (lea, Are- quipa) which are discussed in detail in Part I. A number of hitherto unknown color varieties appeared in these stocks, the inheritance of which could be explained by assuming the existence of a new allelo- morph of albinism intermediate in effect and dominance between albin- ism and its normal allelomorph. More specifically, this new factor is characterized by the production of red eyes, slight dilution of black in the fur, and complete inhibition of yellow pigment development.

The writer has used animals of both the lea and Arequipa stocks in experiments, with results in full agreement with those given in Part I. Crosses 20-1 and 21 to 25 involve red-eye (from lea stock) without also involving dilution. In cross 20-1 a pure lea male, a red-eyed agouti, is crossed with intense guinea-pigs, giving young all intense. This illustrates the dominance of intensity over red-eye.

In cross 21 a pure lea intense male crossed with albinos of intense stock gives both intense and red-eye young. The lea male no doubt was heterozygous for red-eye, but the albinos could not possibly trans- mit red-eye, as they come from a stock in which red-eye has never appeared. This illustrates the apparent reversal of dominance of red- eye whenever albinism is introduced into a cross. A further illus- tration is given in cross 23, in which red-eye by albino of intense

7s IMIKKIIA.V K IN GUINEA-PIGS.

ok gives red-eyes, hut no intense young. In cross 25, red-eyes -. ,1 with jilhinos from various sources give no intense young, but only red-eyes and albinos. One possible explanation of these results would be the supposition that red-eye becomes dominant over its normal allelomorph iii the presence of heterozygous albinism. In this case intense young should appear when such heterozygous red-eyes are crossed together; but, as is shown in cross 24, none such appears. Here red-eyes from cross 21, mated inter se, gave 17 red-eyes, 6 albinos, DO intense. Numerous results of this kind have made it clear that intensity can never be recovered in any generation after a cross of red- eye wit h albino. This means that neither red-eye nor albino can trans- mit the normal allelomorph of the other. Now, the one thing which a recessive variation, of necessity, can not transmit, is its own normal allelomorph. Therefore the normal allelomorphs of red-eye and albino must be identical.

This does not yet demonstrate that albinism, red-eye, and intensity form a series of three allelomorphs. There is still the possibility that red-eve and albinism involve the same recessive allelomorph (Ca) of norma] color (C), but differ by an independent modifying factor. Symbolically we could suppose albinos to be CaCarr, red-eyes to be 1 I IvR (or CaCaEr), intense guinea-pigs of ordinary stocks to be CCrr (or C '( Jarr) , and intense guinea-pigs of lea stock to be CCRR (or CCaRR) . We must suppose the lea stock to be homozygous for the modifier R, to account for the absence of albinos. R must be a unit factor to account for the simple 3 to 1 ratio in cross 24. This hypothesis fits all of the facts given so far. The critical test of its truth is the possibility (as it turns out, impossibility) of producing intense animals (CCaRr) which will give both red-eyes and albinos when crossed with albinos. If intensity, red-eye, and albinism are triple allelomorphs, it should be impossible to obtain such animals. Crosses 21 and 22 are interesting as furnishing just this test. Cross 21 may be represented symbolically as follows according to the two hypotheses:

•w (BW) X intense (lea) = 9 intense + 4 red-eye.

(1) CaCarr X CCaRR = CCaRr CaCaRr.

(2) CaCa X CCr = CCa CrCa.

In oit her case the F, red-eyes crossed inter se should give 3 red-eyes to 1 albino. The result obtained in cross 24 (17 red-eyes to 6 albinos) is in nearly perfect agreement. But the cross of Fx intense with albinos gives very different results under the two hypotheses (cross 22):

-"' X iniensi <!•',) = U\ intense + (0 red-eye) + 25 albinos.

Rr = CCaRr + CCarr + CaCaRr + CaCarr.

- CCa + CaCa-

The complete absence of red-eyes among the 41 young, as well as the

of albinos where an excess of intense is expected, thoroughly

eliminate- the first hypothesis. The results agree reasonably well with

INHERITANCE OF DILUTION. 79

the hypothesis of triple allelomorphs, which we have found to agree with the results of all the other crosses. The only possibility which has not been eliminated is a linkage so close as to simulate a system of triple allelomorphs. Unless exceptions occur which require it, such an hypothesis need not be considered.

Thus the data obtained by the writer are not only in harmony with the theory that albinism, red-eye, and intensity form a series of triple allelomorphs, but can be explained on no other basis, barring the pos- sibility just noted.

DILUTION.

Such color varieties as agouti, black, brown, yellow, etc., are sharply distinct from each other. They segregate from crosses without pro- ducing intergrades and in unforced agreement with Mendelian expec- tation. In contrast with these discontinuous variations are the con- tinuous variations in the intensity of color of each main color variety. Thus, among the yellows, there are all gradations from a pale cream to an intense red. Among the agoutis, there are the pale silver agoutis, the intense golden agoutis, and the intermediate yellow agoutis. There are all grades of dilute blacks known to the fanciers as blues, for which term, as has been explained, sepia is substituted in this paper. Finally, there are all grades of dilute browns and cinnamons. (See plates 1 and 2.)

The existence of these dilute types was noted by Castle (1905) and Sollas (1909), both of whom also recognized that dilution could be transferred from one series to another, e. g., from creams to blacks, giving rise to sepias. They did not, however, suggest any factorial explanation, finding the results of crosses highly irregular. Detlefsen (1914) considered dilution to be recessive, but found the inheritance of dilution very irregular among C. rufescens hybrids. He obtained dilutes in Fy after crossing dilute hybrids with a race of guinea-pigs (brindle or 4-toe), among which dilution had never occurred and which therefore should not carry it as a recessive. It may be remarked in passing that the 4-toe race does contain albinism, which, with present knowledge, satisfactorily accounts for these Fx dilutes.

Thus the difficulties in the way of an understanding of the heredity of dilution have been due (1) to the intergrading of dilute with intense; (2) to data which seemed to indicate that dilution could be due neither to a recessive nor to a dominant unit factor, without complications. Cross 39 gives many examples in which intense by intense has given very dilute young, which seems to indicate that dilution must be recessive if simple Mendelian at all. On the other hand, such cases as that given by Detlefsen are difficult to interpret on this basis. Further, dilute by dilute has often given young much more intense than either parent. Thus, in cross 42-8, we have two medium sepias producing a black. In cross 37 are many cases in which cream by cream

si i INHERITANCE IN (iUINEA-PIGS.

hat produoed yellow. Apparently intense by intense may give any intensity whatever, and almost the same can be said of dilute by dilute. \mi(l this confusion, however, one cross has been found which con- sistently gives a very definite, although unexpected, result. It is found that a dilute crossed with an albino, even of intense stock, never oves intense young, but only well-defined dilutes and albinos. There ■re only a few possible ways in which this result can be explained and, from the results of other crosses, all but one of these explanations have been definitely eliminated, namely, that dilution is an allelomorph of albinism. An allelomorph of albinism was already known to be respon- sible for the red-eyed condition in certain South American stocks I astle, 1914a). It could now be shown that albinism, red-eye, dilution, and intensity are due to a series of four allelomorphs with dominance in the order of increasing pigmentation. A preliminary account of this demonstration has been given in a previous paper (Wright, 1915). In the present paper the demonstration is given in more detail and further steps are taken in the analysis of the variations.

THE DILUTION FACTOR.

The dilute varieties have some resemblance to the red-eyed varieties. The fact that red-eye is due to an allelomorph of albinism suggested that dilution might also be due to a member of the same series of alle- lomorphs. A stock was chosen which was known to carry no dilution. This was the BW stock, which for years has consisted exclusively of the most intense blacks and sooty albinos. The following crosses were designed to eliminate the hypothesis of allelomorphism if incorrect:

(1) Albinos from intense stock were crossed with dilutes:

CaC'all X CCii = CCali.

(2) Albinos from dilute stock were crossed with blacks of intense stock:

CaCaii X CCII = CCali.

If intensity and dilution form a pair of allelomorphs (I, i) which segregate independently of the pair color and albinism (C, Ca), as is the

B in mice and rabbits, these two crosses must give identical results. In each case, color is introduced by one parent, albinism by the other; intensity by one parent, dilution by the other. In fact, identical results should be obtained regardless of whether dilution is due to a unit factor <>r tu multiple factors, or even whether its inheritance is Mendelian or not, provided only that it is inherited independently of albinism. J tosses 16 and 17 and table 35 give the actual results. All cases are included, which involve an intense stock known to carry no dilution.

Among those called dilute below (among the young), none was more intense than sepia* or yellow4. Among the intense, none was more dilute than a dull black comparable in grade but not in color with sepia2, or a red in very few if any cases as dilute as yellow2. There was there- lore no difficulty in drawing a natural line between intense and dilute in these crosses.

INHERITANCE OF DILUTION.

81

It is evident that the two sets of crosses give consistently different results. This difference demonstrates that dilution does not segregate independently of albinism.

An even more striking result follows from a portion of the above data. F: dilutes, one of whose parents was of intense stock, were back-crossed with albinos of intense stock. They gave 9 dilute, 20 albino young, no intense, although these young were at least three-quarters of intense stock. On the other hand, Fi intense, one of whose parents was an albino of dilute stock, were back-crossed with albinos of dilute stock. They gave 5 intense, 7 albinos, no dilutes, although these young were at least three-quarters of dilute stock. It is clear that the hereditary difference between a dilute and an intense can not be transmitted through an albino.

Table 35.

	Intense.	Dilute.	Red-eyed.	White.
cT albino (intense stock) X 9 dilute		56 29	5 10	39 21
9 albino (intense stock) X cf dilute
Total
	85	15	60
d* albino (dilute stock) X 9 intense (intense stock) . . 9 albino (dilute stock) X cfintense (intense stock) . .
47 9			10 2
56			12

It was emphasized above that all the intense animals used in cross 17 came from stocks which have never given dilutes. This was necessary because in other crosses (18, 34, 41) intense by albino has given many dilute young. No such precaution was taken with the dilutes used in cross 16. Any available dilutes were used regardless of ancestry. In fact, 1 1 of them, with 38 young, had one or both parents intense. In none of the other crosses in which dilute has been crossed with albino (19, 27, 38, 44) has any intense young appeared. Thus in crosses with albinos an intense may transmit dilution, but a dilute never trans- mits intensity. From these crosses it seems clear that intensity is dom- inant over dilution. Other crosses on the whole bear this out. The apparent exceptions will be ignored for the present but discussed later.

We have reached the definite conclusion that dilute by albino can never give intense, regardless of ancestry on either side. Since the only thing which a variety of necessity can not transmit is a dominant allelomorph of its essential factor, it follows that dilution and albinism must have the same dominant allelomorph, which we will call intensity.

There are only a few hypotheses which will satisfy this condition. We already know two recessive allelomorphs of intensity, viz, red-eye and albinism. It is conceivable that dilution may be due to the cooperation of an independent factor (or factors) with one or more of the known combinations CrCr, CrCa, and CaCa. If this is not the case,

S> INHERITANCE IN GUINEA-PIGS.

dilution must be due to a new allelomorph in the albino series, let us v.lV i \ modifying factor which shows partial coupling would give

intermediate results.

1 Since dilution and red-eye show considerable resemblance, it would be a plausible hyp.. thesis to assume that they are due to the same allelomorph in the albino series (Cp) but differ by an independent modi- fying factor (D) . With this hypothesis, all stocks used (except the lea and Arequipa) must needs be homozygous for the modifier in order that no red-eyes should appear. Dilutes would be CrCrDD or CrCaDD albinos C.C.DD in these stocks. Thus albinos of these stocks should transmit the modifier and in crosses with red-eyes (CrCrdd) should produce dilutes at least in F2. But in crosses 23 and 25, red-eyes mated with such albinos have given no dilutes, nor have dilutes appeared in Fj in cross 24, among 23 young. Thus an albino can not transmit the hereditary difference between a dilute and a red-eye and the hypothesis is untenable.

(2) Next to be considered is the hypothesis that there is a modifier which converts into a dilute an animal which would otherwise be an all )ino. Dilutes of ordinary stock would be CaCaDD or CaCaDd. In cross 20, dilutes of ordinary stock crossed wdth a pure lea male No. 724, a homozygous red-eye (CrCrdd), produced 5 dilute young which must be of formula CrCaDd. This shows that if the hypothesis is to stand at all, it must be extended, so that the factor which converts an albino into a dilute also converts a red-eye into a dilute. The fact ih.it a dilute may transmit red-eye (crosses 19 and 27) is further evi- dence that this extension is necessary. In this form most of the results can be explained satisfactorily.

(3) The only other hypothesis wmich remains is that dilution is due to a new allelomorph in the albino series making a series of four — C, Cd, Cr, and Ca. The results cited above (crosses 20, 19, and 27) make it evident that dilution is dominant over red-eye. The meaning of a Beries of four allelomorphs can be made clear by considering all of the possible zygotic formulae. Every zygote must have two representatives from the series, but never more than two. Intense guinea-pigs may be homozygous (< C), or carry dilution (CCd), or red-eye (CCr), or albin- ism ('Ca), but can never transmit more than one of the recessive conditions. Dilutes may be homozygous (CciCci) or carry red-eye

or albinism (CdCa), never both. Red-eyes may be homozygous

or carry albinism (CrCa), while albinos can only be homozygous

1 and can never transmit any of the higher conditions.

The critical t est between this hypothesis of four allelomorphs and the

preceding one (that dilute is a modified red-eye or albino), lies in the

possibility or impossibility of producing animals which in crosses with

all linos will transmit more than one recessive condition. If an intense

animal can be obtained which transmits both dilution and red-eye

INHERITANCE OF DILUTION. 83

(CCrDd) or dilution and albinism (CCJDd), or if a dilute can be obtained which transmits both red-eye and albinism (CrCaDd), the hypothesis of modifiers must be adopted. But all attempts to obtain these double heterozygotes have failed. All of the results substantiate the hypothesis of quadruple allelomorphs.

Arequipa male No. 1007 was of formula CrCrDD or CdCd, depending on the hypothesis chosen (see crosses 28 to 34). He was crossed with intense guinea-pigs of BW or 4-toe stock, known to transmit no dilu- tion (CCadd or CCJ . The intense young could only be CCrDd or CCd under the two hypotheses. Five of them were crossed with albinos and gave 13 intense, 20 dilute young, no others (cross 34). Expectation on the hypothesis of a modifier is 16 intense, 8 dilute, 8 red-eye. On the hypothesis of allelomorphs it is 16 intense to 16 dilute. Both the excess of dilutes and the absence of red-eyes point conclusively to the latter.

In cross 18, intense guinea-pigs, each of which had a dilute parent known to transmit albinism and with no lea or Arequipa blood, are crossed with albinos or red-eyes. Under the modifier hypothesis we would expect about half of them to be CCaDd. Under the allelomorph hypothesis, they should be CCd or CCa. As it turned out, there were 6 which gave only intense and dilute (30 intense, 35 dilute) and 8 which gave no dilute young (57 intense to 61 red-eye or albino). Thus there was no intense which had dilute young and also red-eyes or albinos. This result distinctly favors the hypothesis of allelomorphs.

In crosses 19 and 27 dilutes, each from the cross of a red-eye with a stock guinea-pig free from South American ancestry, are crossed with albinos. Under the modifier hypothesis, those which transmit red-eye at all are necessarily CrCaDd, for they must be CraCraD in order to appear dilute; they could get Ca, but not Cr, from the stock guinea-pig parent, and they would necessarily get d from the red-eye parent. Under the allelomorph hypothesis, they should be CdCr, the rest CdCa; 9 gave only dilutes and red-eyes (18 dilutes, 24 red-eyes); 9 others gave only dilutes and albinos (20 dilutes, 16 albinos). There were 3 which had had only 8 dilute young when tabulated. The fact that none of the 9 which had red-eye young also had albinos among 42 young gives a third body of evidence pointing toward the allelomorph hypothesis.

These results make it reasonably certain that the allelomorph hypo- thesis is correct. The only other possibility would involve coupling so close as to simulate multiple allelomorphs. The hypothesis of allelo- morphs has been reached by a method of elimination. It remains to show that all of the data are in harmony with it. In the next section, definite conclusions are reached as to the inheritance of variations in intensity and dilution which make it possible to distinguish intense animals from dilute in all but very exceptional cases.

M

INHERITANCE IN CUINEA-PIGS.

The following table gives a summary of the data bearing on the inheritance of the albino series of allelomorphs based on these conclu- gion8 n will be noticed that animals of every possible formula have been tested by crosses with albinos, the lowest recessives. No attempt has been made bo make all other possible crosses, and several (especially

Tabu 86, -Summary of albino series crosses (crosses 16 to 44).

Parents.

Int. dm x albino.

Dilute X albino.

\,- \ albino

Albino X albino. . Intense x red-eye

Dilute x dilute Intense x intense

Formulas.

X CaCa. • • < '( (I CaCa- • ■ CCr CaCa- • •

CCa CaCa- • •

I il' d X CaCa- • ■ ( VlC'r CaCa- • • CdCa CaCa- • •

CrCr X CaCa- CrCa CaCa •

Int.

40

31

9

64

Dilute X red-eye. .

Red-eye X red-eye Intense X dilute

Ca^a X va^a- • •

°* {§£}■■ °* {§§}•

CdCa

CrCr. CrCa.

CrCa X CrCa-

CC X CdCd. CC CdCa.

CCd CdCd. CCd CdCa. . CC., CdCd- CCa CdC a. .

Dil.

R.E.

36

26 18 98

9 28 25

CdCdX CdCa.

CCd X CCd • • - ('<',1 CCa....

CC CC . . . .

14 12 10 32 12 28

31

15

1 13

24

30 6

W.

71

20

57 75 X

i 40 10 22

15

82

19 35

1 5

17

79

3 X

From crosses-

15

24

17a, 6.

186, 34, 41

21

17c, 17d, 18c, 22

16a, 38a, 44

19,27

166, 16c, 19, 27, 33, 386, 44

25 23,25

Long established.

20

18a, 20 18c

20, 26, 43

20 26,43

24

28 35

29, 40a 32, 36, 40a 28, 40a 36, 406

30, 37, 42 37,42

31,39 39

See rough and Lima crosses

ones involving red-eye) have not yet been made by the writer. The Lasl column refers to the crosses tabulated at the end of the paper. The ratio- expected are obvious from the nature of the matings, except that 39 to 11 were not random crosses of their kind. The appearance of recessive young was used as a criterion of the nature of the parents in these cases. This causes an excess of recessives to be expected.

INHERITANCE OF MINOR VARIATIONS IN INTENSITY. 85

INHERITANCE OF MINOR VARIATIONS IN INTENSITY.

METHODS AND ACCURACY OF GRADING.

The method of grading has been described on page 60. E very- guinea-pig which showed dilute black or yellow in the fur was compared with standard samples of hair within a week of birth. These samples were black0, sepia3, sepia6, and sepia9, in the black series, and red0, yellow.3, and cream6 in the yellow series. Intermediate grades were given by estimate. Grades were taken later in life in many cases in order to determine the relation of age to intensity of pigmentation.

In interpreting the results, it is important to know the accuracy with which the grading could be done and the difficulties met. In some cases the back and belly are fairly uniform in intensity, but usually the belly is considerably the lighter. Tufts of hair for grading have always been taken as near the middle of the back as possible.

In some cases the hair is of fairly uniform intensity from base to tip. In most cases, however, the base is very much lighter than the tip. The color at the tip has been used in grading, although extreme varia- tions in the intensity at the base have also been noted. The color at the tip has most to do with the general appearance of the animal.

The attempt has been made to get both a yellow and a sepia grade for every animal, so that the correlation between the intensities in these series could be determined. This is easy in the sepia and yellow-spotted animals, but in agoutis (where the yellow band of the agouti pattern displaces the sepia near the tip of the hair) determination of the inten- sity of sepia has not been so satisfactory. Several independent determi- nations have been taken in many of these cases. In most cases the same grade was assigned the second time and rarely did the second grade differ from the first by more than one point.

VARIATIONS IN INTENSE GUINEA-PIGS AND ALBINOS.

Before discussing the inheritance of variations among dilutes, it will be well to note briefly the range of variation among guinea-pigs which have the intensity factors (CP). In the BW race the blacks are a very intense black. The base of the hair is only slightly lighter than the tip. In other races, especially the 4-toe stock, the tip of the hair is a dull slaty black and the base a very dull color, often with less pigment than many typical dilutes. The animals have a dull streaky black appearance very different usually from the uniform dark sepia of the darker dilutes. This dull color is not associated with heterozygous albinism. Male M330 was undoubtedly homozygous (CC), having had 9 intense young by albino females and no others; yet he was one of the dullest blacks in stock. On the other hand, nearly all of the intense blacks of the BW race are heterozygous for albinism.

86 INHERITANCE IN GUINEA-PIGS.

This (U1" black can not be due to an allelomorph of albinism between intensity C) and dilution (C(1), since it is a condition which can be tr:m>! i lit ted by albinos. Indeed, the albinos themselves of the BW and 4-toe Btocks differ conspicuously in appearance. The BW albinos have je1 black ears and feet, dark smudges on the nose, and usually M„ on the back. The 4-toe and most other albinos (at the

Bussey Institution I have very much less black on ears, nose, and feet, and the rest of the fur is pure white.

There are parallel variations in the intensity of red in these stocks. 'The occasional red spots in the BW race are of a very intense red

standard redo). In the 4-toe and other dull stocks the red is consider- ably less intense, especially at the base of the hair. The most dilute grade found in tame guinea-pigs known to have factor C is yellow2

D12 cross 35-1).

The wild Cavia cutleri is quite light in color. The black of the fur is a dull slaty color, more like the dull black of the 4-toes than any other color in tame guinea-pigs. The yellow on the back is about yellow3, on the belly cream^ In spite of the resemblance to tame yellow agoutis, Covin cutleri has the intensity and not the dilution factor. When crossed with animals of the BW race, whether blacks or albinos, the young are intermediate in intensity and would be called intense (Part I). Crossed with black animals of the 4-toe race, the young are but little more intense than Cavia cutleri. (See plate 3.)

Summing up : All variations maybe found among intense guinea-pigs, from uniform black0 to a dull slaty black2 and from red0 to yellow2. In the dull grades the hair is especially dull at the base. These variations are hereditary, but have not been analyzed. The hereditary factors for these variations in intense guinea-pigs are responsible for visible differences among albinos. It is to be expected, as indeed is the case, that variations will be found among dilutes, for which these same unanalyzed hereditary differences of different stocks are responsible. Finally, the residual heredity of all tame guinea-pigs has more intensi- fying effect than that of Cavia cutleri, the wild species.

MULTIPLE ALLELOMORPHS.

The ] presence of at least four allelomorphs in the albino series suggests the hypothesis that other allelomorphs in the series may be responsible for the intermediate grades in intensity. It is a tempting hypothesis suppose that the continuous series of variations is correlated with a continuous Beries of allelomorphs, such that each grade of intensity is dominant over all lower grades. If this were the case a stock of dilute-, in which all derive their dilution from a single gamete of one animal, should he fairly constant in their degree of dilution. Again, the cross of dilute by dilute should never give young more intense than the darker parent.

INHERITANCE OF MINOR VARIATIONS IN INTENSITY. 87

However, both of these tests fail. No single gamete stock of dilutes has been found which will not give the entire range of variation when tested. Thus, male D30 red0, an intense which carried dilution as a recessive (CCd), was crossed with red-eyes. His dilute young must all owe their dilution to the same single gamete. They ranged from D340 blackx to D152 sepia5. Yellows which owe their dilution to this same single gamete (derived from male 00 cream6 CdCa, the father of D30) range from D391 yellow2 to 00 cream^. Dilution from a single gamete of A674 sepia6 (CdCa) has given rise to D652 blacki and M306 sepia7, D409 yellow3, and Ml 99 cream7. This last case involves no admixture of South American blood. Inspection of the tables will yield many similar cases. Evidently dilution from a single gamete may appear in dilutes of any grade of intensity. The extreme variations may occur within a single litter (offspring of D30). Again, many examples can be given in which the offspring are much darker than either parent. D652 blacki was the offspring of D215 sepia3 and D106 sepia4. In cross 37 there are 6 cases in which cream6 X cream6 has produced yellow3, with other less extreme cases of this kind. These results do not demonstrate that no more than four allelomorphs in the albino series are present in our stock. They do show that there are other causes producing varia- tion of much more importance than any other allelomorphs which may be present.

THE RELATIONS OF IMPERFECT DOMINANCE, STOCK, AND AGE TO

GRADES OF INTENSITY.

In tables 37 and 38, and diagrammatically in figure 5 and figure 6, all records of grades of dilution at birth are analyzed with respect to genetic constitution and stock. All of those whose genetic constitu- tion was known with complete or nearly complete certainty, either from parentage or from offspring, are put after the proper formula, CdCd, CdCr, etc. All from litters containing two classes are listed separately with the numerical expectation of the classes as (20 CdCd: 32 CdCa), etc. Those in the litters whose formulae were later determined by a test mating are given below in parentheses. These tested individuals are included both among those of certain constitution and in the litters containing two classes. No very close analysis of the influence of stock was possible from the data obtained. However, the following stocks were recognized :

Dil., Dilute selection stock.

Misc., Miscellaneous stocks with but little BW blood and no lea or Arequipa blood.

These contained much dilute selection and 4-toe blood and some C.

rufescens ancestry. hBW, Fx from the cross of miscellaneous with BW stock. \BW, Back-cross of iBW with BW stock. S. Am., All animals with lea or Arequipa blood, in most cases about \ South

American, \ BW, and | miscellaneous, but including pure lea, IcaXBW,

etc.

^

IMM.KITANCE IN GUINEA-PKiS.

In each array Of animals of known constitution and stock the number of animals involved, the mean grade of dilution, and the standard deviation of the frequency polygon are given. It will be noticed that the Btandard deviations decrease as the analysis is made closer. For

Red

White

-Variations in intensity of yellow. Formula and stock printed near mode of each distribution.

mple, the >tandard deviation for all dilute blacks is 1.53, for dilute blacks of formula CdCa is 1.13, and for those of formula CdCa and of South American stock is 1.02. The corresponding numbers for dilute \ellows are 1.10, 0.76, and 0.59, respectively.

INHERITANCE OF MINOR VARIATIONS IN INTENSITY.

89

In tables 39 and 40 are given the mean grades at birth and when more than 4 months old for all guinea-pigs which were graded these two times. These data are arranged by constitution and stock. In most cases the mean grade at birth of the sample graded twice agrees well with the mean grade at birth of the whole array of the same constitu- tion and stock.

Black0 Sep,

Sep

Fig. 6. — Variations in intensity of dilute blacks. Formula and stock printed near mode of each distribution.

VARIATIONS OF YELLOW.

I owe the suggestion that heterozygous albinism may be correlated with extreme dilution of yellow to Professor Castle, who found that attempts to select for a cream stock of maximum dilution led to stocks which invariably gave numerous albinos. The tables and figures con- firm this suggestion in a very striking way. Animals known to be homozygous dilute (CdCd) vary between yellow2 and yellow4 with the mean at yellow2.9. Those known to transmit albinism vary between

QO INHERITANCE IN GUINEA-PIGS.

I uftU W KjJicts of stock and imperfed dominance on intensity of yellow.

list it tit inn.

Stork.

Redg

Red] V2

Yi

Y4

Cr4

<r,

Cn

No.

Mean.

a

Y Cr

( M'.\

I'll

1 3 10

1 2 1 9 2 3 1 3 27 19 (3) (2) 1 7 (1)

2 11 12 13 37 55 8 12 57

3.50 2.64 3.00 4.31 5.51 5.54 I . ss 4.75 4.58

0.50 .77 .41 .46 .60 .71 .33 .43 .59

I >o

J

6

I >•<

3 \ in

! to

4 14 23 7 9 27 20 (3) (4) 13 19

< \K\i

ihi

21 25

4

i to

1

33

l to

'BW

l to

;b\y

I to

S.Am

3 1

1 4

.< r 15 CdC. ( ',!( 'r bv tc-t

Do

.Do. .

( '.!< \ I'V t08t

, .Do

12 ( ',!< ',1' 23 CdCa

Dil .

9 20 (2)

12 18

•

lCd:43CdC». ( ,n ',i t'v teal

Miso

3 (1)

5

4 (1)

5 (2)

!>,,

by teal

.Do. .

(4) m

(2)

lCd:3.5CdCa. 1 MCl

S.Am.BW.

3 14

4 9 36 70

3

Total

7

25 13 169 335

2.88 4.31 5.12 4.72

.65 .46 .76 1.10

6

i . . .

. . .Do

4 80

. . .Do

49

4 7

1 34 10 41

I), I

. ..Do

9

44

128 77

Tahle 38. — Effects of stock and ■	Imperfect dominana.	: on	intensity of black.
Constitution.	Stock.	Bo	Si	s2	B,	s4	s6	S6	s7	s8	No.	Mean.	a
CdCd	Misc S.Am .Do . .	2	1 6 6	6 4 7 2	3		1				11 10 17 4 41 36 15 85 81	2.45 1.40 2.06 1.00 5.51 4.33 3.47 4.20 4.73	0.99 .50 .94 1.00 1.05 .82 .88 1.02 1.32
Do
r	2	1	1
	. Do .
	Misc. .	2 7 9 18 16 9	4 11 2 37 24 6	9 17 3 19 17 8	23 1	3
Do	JBW .
Do	fBW .			1 2
Do	S.Am. .			7 17 13	2 4 7	3
	. . Do . .
I'M: 32 CdCa				9 (2)
. . Do . . .
	. . Do . . .				(1) 1 1 (1) 12 (1) (2) 3	(1)	(3)	(2)
Cd: 8 CdCa	iBW S.Am . . Do . . .		1 1	1 1	9
/l.oCdCd: 1.5 CdCa \CdCa by teal
dCr: 20 CdCa	. . Do . . .		10 (5)	6 (1)	9 (2)	4 (1) (2) 3	6	2
jCdCi by teat	. . Do . . .
	. . Do . . .	(2) 2	(2)
|6CiCr: 11 GCa	..Do...	2 (2)		5 (1)	2
\ CrCr by teat	. .Do...
			(1) 1
	Totnl	2	7 6	10 7 2 3	3 2	1 1				21 17 4 177 81 42 258 303 98 401	1.95 2.06 1.00 4.47 4.73 1.90 4.55 3.95 4.45 4.07	.95 .94 1.00 1.13 1.32 1.06 1.20 1.53 1.54 1.53
		36 16 5 52 66 18 84	54 24 1 78 70 26 96	48 17 2 65 58 20 78	31 17	5 4	3
CdiCdi		2	13	19 3 ::i 6 40
		48 45 19 64	9 10 4 14	3 3 3
Dilute		2 2	20 20
Dil + RE	Do

INHERITANCE OF MINOR VARIATIONS IN INTENSITY.

91

yellow4 and cream7, mean at cream5.i very distinctly paler. Litters which should give both have given the entire range with two modes, at yellow3 and cream5, respectively. It is especially to be noted that among 13 of these, which were given grades before their constitution was known, 4 ranging from yellow2 to yellow4 proved to be homozygotes, while 9 ranging from cream5 to cream7 proved to be heterozygotes. Dilutes known to transmit red-eye (CdCr) have been either yellow4 or cream5, mean at yellow43. These should be compared with those of

Table 39. — Effect of age on intensity of yellow.
Constitu- tion.	Stock.	Mean.	No. in sample.	Mean at birth.	Mean adult.	Dark- ening.
CdCd Do Do Do CdCr	Misc-Dil Misc Dil j-i BW . S.Am . . . . . Do	2.8 5.5 5.5 4.8 4.6 4.3	9 17 9 11 9 5	3.1 5.1 5.2 5.0 4.4 4.8	2.9 5.0 6.0 4.7 4.7 4.2	0.2 .1 - .8 .3 - .3 .6

Table 40. — Effect of age on intensity of black.

Constitu- tion.	Stock.	Mean.	No. in sample.	Mean at birth.	Mean adult.	Dark- ening.
CdCd CdCa	Misc.. . . . . Do	2.5 5.5 4.3 3.5 4.2 4.7 2.1 1.0	8 14 20 6 15 8 16 4	3.0 5.6 4.3 3.3 4.8 4.9 2.2 1.0	2.4 4.6 3.2 2.5 3.3 2.0 1.1 1.0	0.6 1.0 1.1 .8 1.5 2.9 1.1 0
Do Do . Do CdCr	|BW.... fBW.... S.Am. .. Do ... . . .Do
	. . Do

the same stock (S. Am.) which transmit albinism. The difference, yellow4.3 compared with yellow46, is too small to be relied on. Litters which should give both CdCr and CdCa have given a range of yellow4 to cream6, as expected. Thus grade yellow4 may be any sort of a dilute; one more intense is quite certain to be homozygous (CdCd) ; one more dilute is quite certain to transmit either red-eye or albinism.

The influence of stock can only be recognized surely in the case of those known to be CdCa. The numbers are too small among the homozygotes. Among the heterozygotes (CdCa) it is clear that those of dilute and miscellaneous stocks, both with a mean of cream5.5, are distinctly paler than those with an admixture of BW or S.Am, blood with means from cream^ to creamy.

The data in table 39 indicate that yellow undergoes no appreciable change in intensity during the life of an animal, except in the dilute selection stock. In this case there is a change from cream5.2 at birth to creamg.o when adult, among those carrying albinism.

oj INHERITANCE in GUINEA-PIGS.

VARIATIONS OF SEPIA.

We find rather more overlapping of distributions among the sepias tli:m among the yellows when different genetic constitutions are com- pared. Nevertheless there are significant differences in the means. The groups ( !j< ,|. ( \\(\. and CrCr with means from sepiaj to sepia21, nearly black, average distinctly darker than groups CdCa and CrCa with mean- of >epia, - :m(l Bepia4.7, respectively. The case is quite different from yellow dilution in which CdCr and CdCa have the same effect (or nearly bo) contrasting with CdCd. Cr seems to be essentially identical with ( ',; in effect on black, but like Ca in effect on yellow.

1 ( n tint her analysis we must compare stocks. In the miscellaneous stock the average for CdCa is sepia5.5. When this stock is crossed with albinos of BW stock the average of the young— again CdCa — is sepia43. When these are crossed again with BW albinos the average becomes ia,5. The darkening influence of the BW stock is apparent. The South American stock also has a darkening influence with an average of sepia4 o. We find a similar difference between the miscellaneous and South American stocks among the homozygotes.

The comparison of CdCa with CrCa within the same stock (South American) yields a slight but probably significant difference (CdCa, >epia, ,; CrCa, sepia47). Thus there is a difference of 0.5 with a prob- able error of 0.12. It is certain that some of the red-eyed sepias have been paler than any black-eyed sepia.

If there is a real difference here, we wrould expect CrCr to be lighter than (\\(\\ or CdCn but the 4 individuals known to be CrCr give the darkest average of any array. They were not, however, a random sample and, further, were either pure lea or F2 IcaXBW and hardly t<> be compared in stock with those known to be CdCd or CdCr. For the present CdC<j, CdCr, and CrCr may be considered identical in effect • in black fur.

As in the case of yellows, the most critical test of the hypothesis of imperfect dominance is the success of prophecy. In litters which should give both CdCd and CdCa, the 2 darkest tested (sepia2) both proved to be homozygous, while 8 others (sepia4 to sepiaj proved to transmil albinism. Among those which when graded might be either ( ( r or ( ,,( ' . 1 8 were tested. There is some overlapping of ranges, but those which were found to transmit red-eye average very distinctly darker than those which transmitted albinism. Four were tested in an I from red-eve by albino. The 3 dark ones, including 2 which were actually as black as blacks of the BW race, proved to be CrCr, while the other, Bepia*, had albino young and was therefore CrCa.

Table 40 shows thai in the case of the sepias there is a very perceptible darkening with age. This is shown in all groups except the homozy- gous red-eyes, which were practically jet black to begin with. Another interesting point brought out is a race difference in the amount of

INHERITANCE OF MINOR VARIATIONS IN INTENSITY. 93

darkening. The darkening was about 1.0 among 40 animals CdCa without South American blood, although with considerable BW blood in most cases; among 23 animals CdCa or CrCa, with South American blood, the average darkening is 2.0 — twice as much. One case among the latter was very striking. Male D238, a red-eyed sepia, CrCa, was the palest sepia recorded. The tip of the hair was called sepia8; the base was nearly white. When 2 months old, most of the hair was still of this pale color, but there were sharply contrasting areas which were nearly black (sepia2) on the nose, in spectacles around the eyes, in front of the ears, on the feet, and in an asymmetrical patch on the back. At the age of 4 months, most of the fur on the back was sepia^, although the belly remained fairly light. In the lea and Arequipa stocks the dark color always appears first on the nose, feet, and ears. These are the darkest regions generally in all dilutes, a fact which recalls the location of the dark smudges in sooty albino guinea-pigs and Hima- layan rabbits. In adult animals with a large amount of South Ameri- can blood, the darkening with age is so great that 0^0^ can seldom be distinguished from CdrCa, although quite reliable predictions could be made at birth as to the nature of the same animals.

VARIATIONS OF EYE COLOR.

The variations of eye color have not been studied as carefully as those of yellow and sepia fur colors. In intense guinea-pigs (C-) the eye ordinarily appears black (factors B and P of course assumed to be present as throughout the discussion of dilution); in many cases, however, it is possible in the proper light to obtain a red reflection through the pupil. In dilute guinea-pigs CdCd, CdCr, or CdCa, the eye also appears black ordinarily, but a red reflection seems to be obtained more easily as a rule than in intense guinea-pigs. The differ- ence is not great enough to be of value as a criterion. In guinea-pigs which are CrCr or CrCa the pupil appears red in most lights and usually the inner ring of the iris is transparent and also appears red. In very few, if any, cases is the eye so dark that confusion with a dilute or intense is possible. There is much variation in the amount of pigment present. These variations are probably connected with differences in stock and possibly imperfect dominance of Cr over Ca. No pigment has been noted in the pink eyes of albinos. A red-eye can never be confused with a pink-eyed type, unless, of course, factor p is present.

SUMMARY.

1 . First-order effects in the dilution of yellow are due to the presence of various combinations of factors of the albino series of allelomorphs. The red-eye and albino factors (Cr and Ca respectively), produce nearly if not quite identical effects. In the case of black, first-order effects may be due either to different combinations in the albino series or to

!>l

IMIKKITAM K IN < . I I NKA-l'K iS.

independent factors p). I n t he albino series, the dilution and red-eye factors I | and C, respectively) produce nearly if not quite identical effects. In eye pigmentation, as in the black pigmentation of the fur, first-order effects may be due either to different combinations in the albino Beries or toother factors (p); but there is a sharp difference from th«' effects «'ii black fur, in thai the dilution and red-eye factors produce very different effects. In this ease the intensity and dilution factors apparently produce nearly identical effects.

Table 41.

\ bIIow fur.	Black fur.	Black eye.
Formula.	( lolor.	Formula.	Color.	Formula.	Color.
C CdCd CdCra ' !ra' 'ra ■ ■	Red Yellow. . . ( !ream . . White....	C- CdrCdr... CdrCft.... CaCa ....	Black Dark sepia . Light sepia . White	C- cd Cr CaCa ....	Black. Nearly black. Red. Pink.

2. Second-order effects in dilution of yellow, black, and probably eye-color, are due to the unanalyzed residual heredity of different stocks. In the stock at the Bussey Institution BW and South American blood intensify as compared with dilute selection or 4-toe blood. This resid- ual heredity seems to be more important in the case of black than yellow, producing more overlapping of the ranges of the different albino series combinations.

3 . I n only one stock has the intensity of yellow at birth been observed to change appreciably in the lifetime of the animal. In this case, the dilute select ion stock, the creams grow paler as they grow older. Sepia, on the other hand, grows distinctly darker as the animals grow older in all -locks. In the imported South American stocks this darkening is 90 pronounced that adults of any albino series combination, except albinism itself (CaCa), are practically black.

INHERITANCE OF VARIATIONS IN THE AGOUTI PATTERN.

Most wild rodents and many other mammals have a coat color of the agouti type, viz, a predominantly black fur in which each hair has a subterminal yellow band. In many cases, as in the mouse and rat. the entire coat is fairly uniform in appearance. This is not true m .-ill cases, however. The color of Cavia cutleri has been described at the beginning of this paper. It will be recalled that the color of the belly i- Bharply distinct from that of the back, appearing wholly yellow instead of ticked. Tame guinea-pigs of the agouti variety likewise have this so-called light-bellied type of agouti.

VARIATIONS IN AGOUTI PATTERN. 95

The agouti pattern of mice was shown by Cu6not in 1903 to be a unit Mendelian character dominant over its absence as found in blacks. In this and later papers (1903, 1904, 1907) he demonstrated that a white-bellied type of agouti and self yellow are due to members of the same series of allelomorphs. Castle, 1905, demonstrated that guinea-pig agouti is a simple dominant over non-agouti.

This agouti pattern of guinea-pigs is subject to considerable varia- tion. In some cases the belly hairs are entirely yellow, a condition correlated with very broad yellow ticking in the dorsal fur. At the other extreme, the base of the hairs on the belly is black for about half the length, and the dorsal ticking is markedly decreased. This dark type has been produced by repeated crossing with intense blacks (BB race). Although distinctly darker than usual, all of the agoutis from such crosses are distinctly yellow-bellied.

PREVIOUS WORK.

Detlefsen (1914) made experiments with the wild species Cavia rufes- cens of Brazil. This has the agouti pattern, but is somewhat darker than C. cutleri or the tame guinea-pig. The yellow bands in the dorsal fur are narrower and there is usually more black on the belly, which indeed is usually slightly ticked with black. The difference in appear- ance is not very great. Detlefsen found, as he expected, that C. rufescens was homozygous for the agouti factor. In the hybrids between C. rufescens and black guinea-pigs, the agouti behaved as a simple Mendelian dominant. What was not expected was a marked darkening of the agoutis which occurred among the hybrids in many cases. The yellow subterminal bands became so reduced on the back that many of the agoutis appeared more like blacks than guinea-pig agoutis at birth. Black appeared at the ends of the hairs on the belly, and the appearance changed from yellow to ticked. In the early generations the variations in the agouti were exceedingly erratic in their hereditary behavior. Light-bellied hybrids crossed with blacks often gave ticked-bellied young, and ticked-bellied hybrids gave light- bellied young. Nevertheless, as more guinea-pig blood was introduced by repeated back-crosses, the trend was constantly toward the ticked- bellied type. In lines in which the ticked-bellied type had become constant, crosses were made with typical light-bellied agouti guinea- pigs. The ticked-bellied type was found to be recessive and segregated out in later crosses in regular fashion. Detlefsen found that the results in these lines were adequately explained by assuming that the ticked- bellied type is due to an allelomorph of both the light-bellied agouti factor and the non-agouti factor, recessive to the former, dominant to the latter. He used the nomenclature A, A', and a for the tame agouti, wild agouti, and non-agouti factors, respectively.

96

IMll i;ll ANCE IN GUINEA-PIGS.

TH£ INHERITANCE OF THE AGOUTI OF CAVIA RUFESCENS.

The writer ha- had the opportunity of experimenting with the hybrid nil: stock developed by Dr. Detlefsen. As the mode of inherit-

; the fcype of agouti is of special interest in being a character in which two wild species differ, it seemed worth while to obtain additional data. New crosses were made to test out the hypothesis of triple allelomorphs as thoroughly as possible. It may be said at once that the result- obtained completely confirm Detlefsen's hypothesis.

\\ hen received by the writer, there were only 2 light-bellied agouti hybrid- in the stock which had derived their agouti from C. rufescens. These were A606 and A450, \ and f blood hybrids, respectively. They were crossed with black guinea-pigs and one litter was obtained from li 2 bl.uks from A606 and 1 light-bellied agouti and 1 black from A.450. This light-bellied agouti son unfortunately proved to be sterile, so that experiments with light-bellied rufescens agouti came to an end. Only one light-bellied agouti — born dead — has appeared since then which nied to derive its agouti from C. rufescens, and in this case the parentage was doubtful. Thus in the following experiments, rufescens agouti and ticked-bellied agouti are practically equivalent. It must be emphasized that this was not the case in Detlefsen's experiments,

that the following results are simpler than those which he encoun- tered in the earlier generations.

Let us consider first the relations of rufescens agouti and guinea-pig non-agouti. Cross 1 gives matings of non-agoutis with ticked-bellies known to be heterozygous because of a non-agouti parent (table 42).

Table 42.

	Female.	Male.	Agouti light- belly.	Agouti ticked- belly.	Non- agouti.
la 16 le id U	Non-agouti (g. p) Non-agouti (In brid) Non-agouti 'iti ticked-belly uti ticked-belly	X agouti ticked-belly	1	62 17 11 61 10	62 13 12 63 5
X agouti ticked-belly
X A'a (red or white) X non-agouti (g. p.) X non-agouti (hybrid)
1	161	155

The single agouti light-belly was the son of A450, mentioned above, which, though agouti light-belly, is included under agouti ticked-belly ae a rufescens agouti. The cross shows that ticked-belly is a simple dominant over non-agouti. The ratio of agouti ticked-belly to non- agouti is sufficiently close to a 1 to 1 ratio. If ticked-bellied agouti were due to independent modifying factors or to the residual heredity of rw, acting with the same agouti factor as found in C. cutleri and < . porceUus, non-agouti guinea-pigs should possess factors tending

VARIATIONS IN AGOUTI PATTERN.

97

to change ticked-bellied agouti to the typical light-bellied type. The crosses show conclusively that they possess no such tendency. Indeed when it is recalled that, in the early hybrids and C. rufescens itself, light-belly was common, it seems necessary to suppose that guinea-pigs possess a residual heredity which tends to darken agouti.

Ticked-bellied agoutis, known to be heterozygous because of parent- age, were crossed inter se. The results are given in cross 2.

Agtb. X Agtb.

Aglb. . 0

Agtb. 66

Non-ag. 19

This result is sufficiently close to the expected 3 to 1 ratio. One- third of the ticked-bellied young from this cross should be homozygous (A'A') and two-thirds heterozygous (A'a). Several of them have been tested by crosses with blacks (cross 3, table 43) .

Table 43.

Female.

Male.

Agouti light- belly.

Agouti

ticked-

belly.

Non- agouti.

3a 3c 36 3d

7 agouti ticked-belly

Non-agouti

5 agouti ticked-belly

Non-agouti

X Non-agouti

X 9 agouti ticked-belly

X Non-agouti

X 1 agouti ticked-belly

10 40 25 12

11

38

The single agouti light-belly was the one of doubtful parentage men- tioned above. Sixteen heterozygotes were obtained which gave agouti ticked-belly and non-agouti in approximately equal numbers ; 6 possible homozygotes were obtained, rather fewer than is to be expected. The male AA253 with 12 agouti ticked-belly young and 2 females, AA213 and AA217, with 8 agouti ticked-belly young each, were quite certainly homozygous and were used to establish a homozygous ticked-bellied stock. They and their progeny crossed inter se have given only ticked- bellies, 26 in number (cross 4). These homozygous ticked-bellies are indistinguishable from heterozygotes in appearance.

Cross 6 gives matings of homozygous light-bellied agouti guinea-pigs with non-agouti hybrids. The young, 29 in number, are all light- bellied. There is no tendency toward ticked-belly introduced by th#»

W{$S«f / givC. ™^ - ^gxx^mea nybrids (agouti derived from guinea-pigs) with non-agoutis. All of these light-bellies were known to be heterozygous from their parentage. The result, 18 light-bellies, 21 non-agoutis, no ticked-bellies, is in harmony with expectation

(19.5 :19.5). , „ A. _ • ,. „

Crosses 8 and 9 give data on the relation of light-belly to ticked-belly. Homozygous light-bellied guinea-pig by ticked-bellied hybrid gives exclusively light-bellies, 50 in number. Light-belly is thus clearly

ms INHERITANCE IN GUINEA-PIGS.

dominant. Beterxaygous light-belly (with a non-agouti parent) by heterozygous ticked-belly (also with a non-agouti parent) gave 16 light-bellied, 6 ticked-bellied, and 10 non-agouti young where expecta- tion is Hi : S 8.

The results given bo far show that light-belly is dominant or at least epistatic over ticked-belly, that ticked-belly is a simple Mendelian dominant over non-agouti, and that the difference between rufescens :iiul poreeUiu agouti is not a question of residual heredity. The fact thai crossing with guinea-pig non-agouti increases the difference be- tween rufescens and porcellus agouti, instead of destroying it, shows t hat ruf( 8ft ns agouti does not contain the same agouti factor as is found in guinea-pig agoutis. Rufescens agouti must have an allelomorph of guinea-pig agouti, recessive to the latter. This leaves two possibilities. This allelomorph may be (I) the non-agouti factor or (II) a new allelo- morph recessive to the porcellus agouti factor, dominant to non-agouti (Detlef sen's hypothesis). Both of these explanations fit equally well all of the data given so far. Under (I) a guinea-pig light-belly is AAa'a', a non-agouti aaa'a',and a rufescens agouti aaA'A'. Under (II) these three varieties are AA, aa, and A'A', respectively. The critical u si is whether it is possible to produce light-bellies which are double heterozygotes AaAV, capable of having both ticked-bellied and non-

mti young, as well as light-bellies when crossed with non-agoutis. 1 )ctlefsen obtained 5 light-bellied agoutis from the cross light-belly by heterozygous ticked-belly which bear on this point. Each of these had ticked-bellied young, but no non-agoutis. They, therefore, point toward hypothesis (II), which is also more probable a priori. They had, however, only from 3 to 6 young, 21 in all, so that it is not wholly certain that they would have had no non-agouti young if tested further. This point, therefore, seemed to the writer to be one on which additional data would be desirable, and special attention has been paid to it.

The cross heterozygous ticked-belly by heterozygous light-bellies known from their parentage to be free from ticked-belly can be repre- Bented ;i- follows nnder the two hypotheses:

Aglb. AgJb. Aglb. A '/lb. Auib. Nortrng.

\ .' X AaaV = A;. V.-,' + Aaa'a' + aaAV -I- aaa'a'. (II) A'a XAa = AA' + Aa + A'a + aa

-..r.'mti V.'wi-'.'.^v'0 expect 2 light-bellies to 1 ticked-belly to 1 non-

'" ilv A; v ;1) ,,llls, ;,lso have the power of transmitting non-agouti. Under II such light-bellies (AA') should not transmit non-agouti. *f r 1 1 1 hall (.1 the light bellies should be of this type and the other hah should transmit oon-agouti but not ticked-belly (Aa). Thus, if a Lot:. Qumber of young ran be obtained from a light-belly from such acrofi J, which has had ticked-bellied young in crosses with non-agoutis, the presence or absence of non-agouti young is decisive between the

VARIATIONS IN AGOUTI PATTERN.

99

two hypotheses. Cross 10 gives the results of the tests of light-bellied young from such a cross as described (table 44).

Table 44.

	Females.	Males.	Aglb.	Agtb.	Non-ag.
10a	7 aglb . . .	Non-ag . .	17		20
10c	Non-ag . .	4 aglb ....	26		33
106	10 aglb...	Non-ag . .	IS	30
lOd	Non-ag . .	5 aglb. . . .	45	50

In no case has the same animal had both ticked-bellied and non- agouti young. Some of those which have had ticked-bellied young have been quite thoroughly tested. Male M138 had 20 light-bellied and 19 ticked-bellied young. Male B121 had 13 light-bellied and 13 ticked-bellied young. Male M91 had 5 light-bellied and 11 ticked- bellied young. The chance that these can represent 2:1:1 ratios is negligible. Thus hypothesis (I) may be dismissed.

Light-bellied agoutis demonstrated to carry ticked-belly have been crossed inter se (cross 11). They have given 25 light-bellies and 9 ticked-bellies, no non-agoutis. This agrees reasonably well with the expected 3 to 1 ratio. The remaining tables give the results of miscel- laneous crosses. All of them are in harmony with the hypothesis of triple allelomorphs.

MINOR VARIATIONS.

Thus there seems no doubt that the light-bellied agouti of Cavia porcellus, the ticked-bellied agouti of C. rufescens hybrids, and non- agouti form a series of triple allelomorphs. The question remains whether light-bellied Cavia rufescens hybrids possess a different allelo- morph from the ticked-bellied ones, or whether the difference lies simply in the residual heredity. There are no wholly satisfactory data bearing on this point. Nevertheless the fact that the darkening seems associated especially with certain stocks of guinea-pigs seems to favor the second view. The writer has crossed ticked-bellied agoutis re- peatedly with the intense blacks of BB or BW stock. Young have been obtained which were self black, except for a few ticked hairs in the chest and whiskers. One ignorant of their history would probably have classified several of them as blacks. Before they became adult, however, these black ticked-bellies acquired a uniform though very slight yellow ticking throughout the entire fur. On crossing such black ticked-bellies with a dull black stock (4-toe) there is a return to a more strongly developed agouti pattern. The young are uniformly ticked when born. Thus these variations in the agouti pattern seem related to the residual heredity of the stocks, possibly with the same residual heredity which determines the very intense development of pigment,

I DO INHERITANCE IN GUINEA-PIGS.

especially black, in the BB and BW races, the feebler development in the 4-toe race, and the much feebler development in the wild species. If thifl 18 correct, the resemblance of light-bellied rufescens to light- bellied agoutis, like thai of the pale color of C. cutleri to dilute guinea- pigs, is secondary. In both eases the wild species possess a different allelomorph from the guinea-pig in the principal series of factors involved, but owing to different residual heredity, have a superficial smblance.

THE INHERITANCE OF THE AGOUTI OF CAVIA CUTLERI.

The writer has had the opportunity of working with the agouti of Cavia cutleri. Repeated crosses have been made with blacks of the BW race of guinea-pigs to see whether a ticked-bellied agouti could be obtained. While some ventral ticking has been observed in some cases, the J -blood cutleri hybrids are still on the whole good light-bellied lutifl. The cutleri agouti is unquestionably more resistant to dark- ening influences than was rufescens agouti. No results have been obtained yet which serve to differentiate it from the guinea-pig agouti. This is additional evidence that C. cutleri was ancestral to porcellus. The experimental results are given in crosses 68 to 78.

Only one cross has been made between a cutleri and ticked-bellied rufescens hybrids. Male K56, a black | cutleri (f 4-toe), was crossed with two ticked-bellied agoutis, which of course had some rufescens ancestry. There were 6 young (3 blacks and 3 ticked-bellies) of which one was quite light and one was black, except for a few ticked hairs on the chest and whiskers. There was thus no very conspicuous tendency toward light-belly introduced bj' the cutleri hybrid. It seems safe to assume that C. cutleri has a different member of the agouti series of allelomorphs from C. rufescens, but the same or nearly the same as C. porcellus.

Wild species of the same genus seldom differ as much superficially in any one character as do many varieties of domesticated animals. Yet while very large variations in the latter have been shown in many cases to behave as simple Mendelian units in inheritance, the char- acters by which wild species differ usually seem to be highly complex in heredity. Few well-defined Mendelian factors are recorded in the literature of hybridization. It is, therefore, interesting to find that the darker agouti of Cavia rufescens differs from the lighter agouti of C. cutleri by a clear-cut Mendelian factor.

INHERITANCE OF ROUGH FUR.

In the wild species of cavy, and in the ordinary smooth guinea-pigs, the hair shows a definite direction of growth, which is always away from the snout on the body and toward the toes on the legs. This is at least the general tendency of the hair in most mammals, and it is

ROUGH FUR. 101

obviously the most advantageous; the hair lying thus is not ruffled or caught by obstacles when the animal is moving. This direction is not directly imposed on the hair by outside agencies, as might be supposed, but is due to the direction of growth of the hair follicles (Wilder, 1909).

Certain fancy varieties among guinea-pigs, as the long-haired rough "Peruvians " and the short-haired rough " Abyssinians," show a striking deviation from the normal hair direction. In these varieties the coat can be divided into a number of areas, within each of which all hair directions radiate from a definite center. The boundaries of these areas, where contrary hair-currents meet, are marked by crests. The centers with their radiating hair-currents are called rosettes. Many mammals, including man, naturally show rosettes, crests, or other peculiarities of hair direction, but less conspicuously than the rough guinea-pigs. (See plate 7.)

The positions in which rosettes may occur in guinea-pigs are quite definite. Following are the rosettes and irregularities given by Castle (1905), with the addition of L, irregular roughness on the chest.

A. Forehead, unpaired. E. Sides, between shoulder and hip. I. Navel, unpaired.

B. Eyes. F. Hips. J. Front toes.

C. Ears. G. Above the groin. K. Hind toes.

D. Shoulders. H. Mammae. L. Irregular roughness of chest.

In the grading of the young guinea-pigs, large letters, as above, have been used for well-defined rosettes and small letters for feeble rosettes or slight deviations from normal hair direction in an area, indicated only by crests at a boundary. Thus a mid-dorsal crest or mane (e) without any side or hip rosettes is characteristic of a certain grade of partial roughs. The ear rosettes (C) are usually only revealed by a crest between the ears. The shoulder rosettes are seldom well developed. The side rosettes are sometimes doubled in the roughest animals (E,E2).

The number of rosettes present varies from the full set described above, through a continuous series of intermediate grades, to one pair. The variations are not merely haphazard, but may easily be classified. In the first place, it is necessary to distinguish two series. A slight roughness found in certain stocks (the BW, lea, and Arequipa stocks) does not fit into the usual series of variations and will be discussed separately as series II. The roughness of the remaining stocks and also of the fanciers' "Peruvians" and "Abyssinians" we may call series I. In series I all the variations found may be arranged with con- siderable accuracy in a single linear series. Thus Castle (1905) used six grades, passing from rough A with the maximum number of rosettes to rough F, smooth except for the hind toes. These grades will be used in this discussion, with the exception that it has been found con- venient to combine grades C and D, leaving five grades, A to E. These letters used for grades must not be confused with those used to name the rosettes.

102

IMIKHITANCE IN GUINEA-PIGS.

Reversal of hair direction on the hind toes is the most constant feature of the roughness of series I and has been found in all rough guinea-pigs of series I without exception. Following is the usual order of succession of the additional rosettes and irregularities found in pass- ing from a smooth to a full-rough:

Hind toes K

J

il creel e

Side rosttt<->, crest between eara E,C

Forehead, hip. ventral rosettes. . . A,F,H,I,L

Eye rosettes B

Groin, shoulder, second side ro- settes G,D,lv

There i- seldom more than a slight amount of asymmetry. As a rule the paired rosettes are present or absent as pairs. The most common exception is in the side rosettes in low-grade partial-roughs. Among these it is not uncommon to find a good rosette on one side and merely a slight change in hair direction on the other. In classifying, six of the most distinct sets of rosettes have been used as the principal criteria, viz, forehead, eyes, sides, hips, front toes, and hind toes.

CLASSIFICATION.

Hough A. — The forehead and five critical pairs of rosettes must be well developed. In addition, there is always some ventral roughness and a crest between the ears.

Rough B. — This includes various conditions intermediate between rough A and rough C.

Rough C. — There is only one pah' (or half pair) of well-developed r< iee1 1 < s. usually the side rosettes. There is always, in addition, rough- ness on at least the hind toes and usually a crest between the ears.

Rough D. — A mid-dorsal crest is present and roughness of at least the hind toes, but no well-marked rosettes.

Rough E. — Roughness is confined exclusively to the toes, usually to the hind toes.

The following list shows the variations which have been met with in each grade; the letters represent the rosettes:

Rouuli A. ABcEFHIJK to ABCDE1E2FGHIJKL.

Rough B. \<i:.JK, A.KHJK, abFGJK, ABcFHIJKl, ABcDEHIJKl.

Rough C. KK, K.IK, cEK, cEJK

Rough 1). iK. cK, eJK, eJK.

Rough K. K. .IK.

PREVIOUS WORK.

Nehring (1894) made crosses between rough guinea-pigs and the wild species Cama aperea. He described the young as smooth, but noted thai a mane developed along the middle line of the back. Castle (19(C) i demonstrated that rough fur behaves as a Mendelian unit

ROUGH FUR. 103

character, dominant over smooth, and is thus an example of the com- paratively rare class of dominant mutations. He found that the grade of roughness, while fairly constant in some stocks, could be reduced by crossing with smooth guinea-pigs of a particular stock (tricolor), which he described as prepotent smooth. Detlefsen (1914) found that rough fur in hybrids between rough guinea-pigs and the wild species Cavia rufescens continued to be inherited in Mendelian fashion, but that domi- nance ceased to be complete. The writer began experiments in 1913 at Professor Castle's suggestion, to investigate further the heredity of variations in the rough character.

MATERIAL.

Several stocks have been used as material:

(1) l+-toe stock. — A full-rough was crossed with members of the 4- toe stock, and by repeated back-crosses into the latter a stock has been produced which is practically pure 4-toe. No partial-roughs have ever appeared in these crosses . Pure 4-toe smooth animals have been very useful in the experiments, since it has been amply proved that when crossed with full-roughs they never reduce the grade of rough.

(2) Tricolor stock. — Most of the partial-roughs experimented with are of very mongrel stock, with, however, more or less tricolor ancestry. In this section on rough fur the term tricolor stock will be used for convenience for these animals, without implying that all of them actually were tricolors.

(3) Lima stock. — This stock as has been described was derived en- tirely from 8 guinea-pigs (2 nearly full-rough" (rough B) and 6 smooth) brought from Peru in 1913.

(4) Rufescens hybrids. — The writer has worked with a few rough animals descended from Detlefsen's hybrids and containing from \ to ^j- Cavia rufescens ancestry.

(5) Cutleri hybrids. — Crossing with Cavia cutleri has been found by the writer to have a similar effect on the rough character to that described by Nehring (1894) for C. aperea and by Detlefsen (1914) for C. rufescens. The behavior of the roughness in these cutleri hybrids has been investigate •

(0) Miscellaneous smooth guinea-pig stocks have been used in some of the experiments. All used resemble the 4-toe smooths in giving no partial-roughs when crossed with full-roughs.

(7) BW, lea, and Arequipa stock occasionally have shown a slight roughness which is distinct from the usual type and is discussed later under series II. The BW stock has been used in a few cases as a source of smooth guinea-pigs. Aside from the slight roughness of series II all of those used have behaved like 4-toe smooths. Smooth

11)1

INHERITANCE IN GUINEA-PIGS.

Sm

I, ..... on the other hand, behaved like the wild cavies in reducing the roughoi

PROBLEMS.

The following figures show the kinds of roughs which have appeared in the experimenta with the four principal stocks and the nature of the variability in the rough character, the inheritance of which it is desired to analyze. All roughs are included, but only such smooths as had at least one rough parent.

Inspection of figure 7 at once shows striking differences in the variability of the rough character in the different stocks. In the 4-toe stock there is a wide gap between the lowest rough and the smooths which come from the same parents. Only one individual in the 4-toe

stock was graded rough B and this was close to rough A (lacked only groin and hip rosettes) . Most of the individuals were strong rough A. In the Lima stock most of the individuals which were rough at all were rough B or a weak rough A, but 5 were rough C or D. Among the tricolor and cutleri hybrids a continuous series can be formed passing from the best roughs to the smooths. Both show a distinctly bimodal distri- bution of the roughs, the modes being at rough A and rough C. Such a distribution would of course be purely artificial if rough B were a more limited category than rough A or rough C, but the

I i.,. 7.— Distribution of grades of roughness of the definitions show that TOUgh B in- fur in foui M,nK, .,f dudes perhaps the ^ddest range

of possible variation. Further, the large number of rough B's in the Lima stock shows that this class may be piaotically as numerous as rough A under the right hereditary conditions. Thuo thoro are strong intimations that in the tricolor and cutleri stocks, rough A and rough C differ by a unit hereditary difference.

The problem of the inheritance of the variations in the rough char- acter thus seems to resolve itself into three phases: (1) The inheritance of roughness of any sort as opposed to smoothness; (2) the inheritance of a more or less full-rough type averaging about rough A as opposed to a partial-rough type averaging between rough C and D; (3) the inheritance of the variations within the full-rough and partial-rough tyi •

♦•.	1
35	4 toe stock
	9
	164
109
83	Tricolor stock
26	24
		* 1
	♦9
M	Lima stock
24
	6			:
	77
35
»	Cutleri
19	hybrid
12

ROUGH FUR.

105

INHERITANCE OF ROUGH AS OPPOSED TO SMOOTH.

Castle (1905) demonstrated that all roughs differ from smooths by a Mendelian unit-factor and that rough is dominant over smooth. The writer's experience fully confirms this conclusion.

In nearly 3,000 young recorded, smooth by smooth has never given rise to rough (of series I) in spite of much rough ancestry, with one possible exception. This exception was of a kind which was expected and was being tested for when found. One of the smooth parents was undoubtedly like rough E genetically. The case will be discussed later.

Table 45.

Rough X rough.	Rough X smooth.
Formula and stock.	Rough.	Smooth.	Formula and stock.	Rough.	Smooth.
RR X Rr: 4003 (tri)	6 10 88 18 17	0 3 28 7 2	RR X rr: 4003 (tri) Rr X rr: 4-toe	11 29 104 56 60 54	0 32 125 55 63 48
Rr X Rr: 4-toe
Tricolor Lima Cutleri hybrid Total	Tricolor Lima Cutleri hybrid Miscellaneous Total
133 (130)	40 (43)	303 (313)	323 (313)
Expectation Largely Rr X Rr: Miscellaneous	Expectation Largely Rr X rr
46	12	16	12

On the other hand, rough by rough has often given smooth. In the cross of rough by smooths from stocks in which roughs have never occurred, all or half of the young are rough. All wild cavies, for ex- ample, are smooth and have only smooth descendants when crossed with smooth guinea-pigs. They have numerous rough young in F: when crossed with rough guinea-pigs. Thus it is clear that rough is domi- nant. Table 45 is a summary of the rough crosses made by the writer. Grades of roughness are ignored. No special attempt has been made to obtain homozygous roughs. Male 4003, rough E, is the only one which has been adequately proved homozygous. Male R197, rough A (cross 46), is another which is probably homozygous. In the cross Rr X Rr above, only matings are included in which both animals are known to be heterozygous, either because of a smooth parent, or, having had 12 or more young, because of smooth young. Tabulated in this way, the expectation is not appreciably different from 3 rough to 1 smooth. Other litters of rough by rough are tabulated above. Probably most of these are of the type Rr X Rr, although in some cases there were no smooth young. In crosses Rr X rr, the only cases tab-

106

INHERITANCE IN GUINEA-PIGS.

ulated arc those in which the rough is known to be Rr because of

mootfa parent, or because of a smooth young one in 6 or more.

Expectat ion ifl here 1 to 1. The remaining cases of rough by smooth,

in which expectation is still probably not far from 1 to 1, are also

given.

These results are in harmony with the view that rough differs from

smooth by a dominant unit factor.

INHERITANCE OF MAJOR VARIATIONS.

Before giving any hypotheses, it will be well to present the experi- mental results with the immediate deductions which can be drawn from them (tables 46 to 55).

(1) In some stocks there is very little variation in the rough char- acter and there is a wide gap between the lowest rough and smooth. The 4-toe stock is an excellent example of such a stock. It is worthy of note that we get a similar result in the Lima stock if we exclude the litters of LG, L24, L56, and L99. Female L6, smooth, was one of the original 8 in the Lima stock. Female L24, smooth, was her daughter. Female L56, rough C, was the daughter of L24, and male L99, rough C, was the son of L56. Most of the tame guinea-pig stocks (BW, dilute selection) seemed to be like the 4-toe stock in the above respect when- ever rough was introduced into them in a cross.

Table 46.

( r i	Stock and grade.	A	B	C	D	E	Sm
(49) (03)	Four-toe stock : A X A	10 28 7 21 5 4 31	1 5 5 9 4 1				3 32 3 17 21 4 36
A X Sm
Lima stock, except L6, L24, L56, and L99: B X B
A X Sm
B X Sm
Miscellaneous stock: A, B (Lima) X Sm (4-toe) . . A X Sm

Most of those graded rough A or rough B above must be heterozy- 18. As nothing higher than rough A appeared in the crosses A X A

and B X H, it semis clear that the homozygotes in these stocks at

least are no more rough than the heterozygotes, i. e., dominance is

complete.

2 When a wild species of Cavia (smooth), or a smooth of certain

tame stocks is crossed with a full-rough, the rough young are of low

grade — rough ( ' or I ).

ROUGH FUR.

107

As has been mentioned before, Nehring (1894) crossed a rough male guinea-pig with Cavia aperea of Argentina. He described the young as smooth at first, but developing a mane along the back later. This was evidently the dorsal crest of rough D. The reversal of hair direc- tion on the hind toes might easily have been overlooked. Detlefsen (1914) described crosses of C. rufescens of Brazil with full-rough guinea- pigs. His rufescens was undoubtedly a different species from the aperea used by Nehring, since the latter found complete fertility among the hybrids of both sexes, whereas Detlefsen found sterility among all the male hybrids. The rough young were rough D. The skin of one of them (A10) shows roughness on all the toes and a very slight dorsal crest. The writer has crossed rough A guinea-pigs with C. cutleri of Peru with similar results. Nine rough young have been obtained, of which 3 show good side rosettes and are rough C, while the others merely show a dorsal crest (and rough toes) and are rough D. A male of pure lea stock, which being from feral stock probably had consider- able wild ancestry, was tested by crosses with full roughs and also gave only partial-roughs — 5 all rough C. Castle (1905) obtained partial-roughs, C or D, on crossing full-roughs with smooths of tri- color stock. The writer has made further crosses of this kind with similar results. One smooth guinea-pig of the Lima stock, L24, had some partial-rough young when crossed with full-roughs of her stock.

	Table 47.
Cross.	Stock and grade.	A	B	C	D	E	Sm
(68) (67) (71, 74) (65) (59, 60) (51)	Smooth, of wild or feral stock: A X Sm (C. aperia) A X Sm (C. rufescens) .... A X Sm (C. cutleri) A X Sm (lea) Certain tame stocks and wild hybrids: A X Sm fj cutleri) A X Sm (j, fa rufescens) . A X Sm (L6, L24) A X Sm (Tricolor)	10 1 4 6	3 1 1	3 5 10 2 3 3	1 4 6 1 3		7 15 4 23 6 19	Nehring (1894). Detlefsen (1914).

The results given in table 47 show that in partial-roughs from a great variety of sources, the low grade of the roughness is not due to a vari- ation of the rough factor, i. e., to an allelomorph, nor is it due to an inde- pendent duplicate rough factor which produces somewhat similar effects to the factor of full-roughs. These partial-roughs have the identical rough factor present in full-roughs derived directly from the latter.

(3) When a wild cavy is crossed with a partial-rough guinea-pig, rough young of the lowest grade (rough E) are produced.

ins

INHERITANCE IN GUINEA-PIGS.

No rough E young were produced in the crosses under (2) where on.- parent was full-rough.

	Table 48.
	Stock and grade.	A	B	C	D	E	Sm
(69)	CD x sin (C. cutleri)			1	1	9	12

(4) Partial-roughs crossed together may give all grades of roughness from full-rough (A) to the lowest partial-rough (E).

	Tabi	,e 43	.
Cross.	Stock and grade.	A	B	C	D	E	Sm
	cutleri hybrids:
(76)	c xc Tricolor :			1	3	1
(52)	C, D XC	18	6	19	7	12	17
(53)	C,D XE			4	1	6	1
(55)	E XE					4	3

(5) Full-roughs crossed together have never given partial-roughs. Crosses in the 4-toe and Lima stock have already been given. Below are crosses of rough A by rough A in the tricolor and cutleri stocks in which each rough A parent had one or both of its parents partial- rough (C, D) or in the case of the cutleri hybrids, an Fi hybrid.

Table 50.

Cross.	Stock and grade.	A	B	C	D	E	Sm
(46) (75)	Tricolor A X A J cutleri A X A	17 3	2				7

(6) Full-roughs, one or both of whose parents were partial-roughs, have given no partial-rough young in crossing with smooths of 4-toe

stock.

Table 51.

( roa	k ami grade.	A	B	C	D	E	Sm
(50)	I'rh X Sm (J- toe) .... A (} cut) X Sm (4-toe)	19 2					20 2

In (73) two of tin- \ cutU ri rough A did not come from a partial-rough parent, but from smooth \ cutleri hybrids.

ROUGH FUR.

109

(7) Partial-roughs crossed with smooths of 4-toe or a similar stock give partial-rough young, and also, in most cases, full-rough and smooth young.

Table 52.

Cross.	Stock and grade.	A	B	C	D	E	Sm 79 5 21 2 8
(54) (64) (72) (61) (56)	C, D (Tri) X Sm (4-toe, etc.) C (£, j, e? rufescens) X Sm (4-toe, etc.). . C,D (J, £ cutleri) X Sm (4-toe, etc.) .... L 56 C X Sm (Lima) E (Tri) X Sm (4-toe)	34 3 12 2	1	29 2 6 1 13	13 6	1

Crosses (3) to (7) show that full-rough can be recovered from partial- rough usually without trouble, but that partial-rough can not be re- covered from full-rough, either in crossing full-roughs together or with ordinary smooth guinea-pigs (4-toe, etc.). Cross (7) is more significant than appears at first sight. We know that 4-toe smooths transmit nothing which can reduce the grade of roughness. Thus, when we find partial-rough by 4-toe smooth giving partial-rough young, we see that the rough factor and the factor or factors responsible for the low grade of roughness can be transmitted in the same gamete.

(8) Most partial-roughs crossed with full-roughs give a very similar result to the cross partial-rough by 4-toe smooth, except that fewer smooths are produced.

Table 53.

Cross.

(47) (58) (70) (48)

Stock and grade.

A X C (Tri)

B (Lima) X C (Lima)

A (4-toe, etc.) X C, D (§, \ cutleri) A (Tri) X E (Tri)

10

8

B

D

E

Sm

10 4

(9) The lowest grade of roughs (E) very rarely have either a full- rough parent or full-rough young. They also very rarely have a smooth parent of such a stock as 4-toe.

Of the rough young, 507 have been recorded (excluding 8 from mixed bimaternal litters) ; 413 of these had a full-rough (A, B) or a 4-toe smooth parent; yet these include only 3 rough E young. On the other hand, 13 rough E are included in the 67 rough young from C or D X C ; 6 are included in the 11 from C or D X E, 9 are included in the 11 rough young from C X smooth Cavia cutleri and all of the 4 rough young from E X E were rough E. Table 54 shows the matings in which one or both of the parents were rough E. These are repeated from other crosses.

110

INHl.KI IWCE IN GUINEA-PIGS.

The offspring of male 4003 rough E are of special interest. He was undoubtedly homozygous rough; the chance that he was heterozygous jg j .■: [)' =0.0001. iE he was the lowest grade of rough, he very emphatically disproves any necessary relation between homozygosis and high development of the rough character, or between heterozygosis and partial roughness.

Table 54.

(56)

(55)

Stock and grade.

E X Sm (4-toe) .

E X A

E X C,D

E X E

B

13 2 4

D

E

Sm

(10) Occasionally a smooth from a cross which produces rough E will transmit the rough factor, breeding like a rough E. Rough E grades into smooth. From a cross which can produce rough E, 8 smooth animals were tested by crossing with 4-toe smooths to determine whether a smooth can ever be like rough E genetically. One such animal, female R201, was found.

	Table 55.
( p.--	Stock and grade.	A	B	C	D	E	Sm
(57)	7 Sm (C X C) X Sm (4-toe) . . . R 201 Sm (C X C) X Sm (4-toe) . . .			1			32 1

The case is not quite as clear as could be desired, since R201 seemed to show a trace of irregularity on one hind toe when first graded. As an adult she is indistinguishable from a smooth.

These experiments are sufficient, it is believed, to establish the mode of inheritance of the major variations of the rough character.

First , it is clear that partial-roughs do not owe their roughness to an allelomorph of the rough factor or to an independent duplicate rough factor, but to the same factor found in full-roughs. Reasons were given under _' .

Next, any hypothesis is untenable according to which partial-roughs are due to imperfect dominance and hence are necessarily heterozygous either with the ordinary smooth factor of guinea-pigs or with a more potent allelomorph of the latter present in wild cavies and special stocks of tame guinea-pigs. The latter hypothesis was suggested by Det- lefsen 191 1), in the case of the partial-roughs among the rufescens hybrids. Be represented the rough factor by Rf, the ordinary smooth factor by rf. and the smooth factor of Cavia rufescens by rf. He sup- posed that Rf Lb completely dominant over rf, but incompletely domi- nant over rf. Thus Rfrf would be a full-rough, but Rfrf a partia-

ROUGH FUR. Ill

rough. This hypothesis explains very satisfactorily all of the crosses given except those under (7), repeated under (9) and (10). It can not explain the case of male 4003 rough E, who was undoubtedly homo- zygous for the rough factor (RfRf) and yet was the lowest grade of partial-rough. Further, it can not explain the occurrence of partial- rough young coming from the cross partial-rough by 4-toe smooth. The latter are necessarily rfrf under the above hypothesis.

Rfrf X rfrf = Rfrf + rfrf.

Rough C, D, E X smooth = rough A + smooth.

Under this hypothesis, the rough factor Rf and the factor which reduces the grade of rough rf can not be present in the same gamete. But this cross actually gave 70 partial-rough young, 56 from tricolor partial-roughs, 12 from partial-rough cutleri hybrids, and at least 1, probably 2, where the partial-rough parent owed its low grade to Cavia rujescens. Female A606 rough C was \ rufescens. Her parents were 2193, a full-rough guinea-pig, and A63, a smooth rufescens hybrid. The hypothetical factor rf could only have come from the latter. The parents of A63 were A55, a pure Cavia rufescens, and 9586 of BW stock, a stock which has shown no tendency to reduce the grade of full-roughs on crossing with them. It thus seems clear that A606 owes her low grade of rough to her Cavia rufescens grandfather. She was crossed with a 4-toe smooth male, 166, and had two rough young, whose grades unfortunately were not recorded at birth. One, however, A1687, is still living (August 1915), and is a typical rough C. There is reason for believing the other to have been of the same grade. Thus in tricolors, Cavia cutleri and Cavia rufescens hybrids, the same gamete can transmit the rough factor and the factor or factors which limit the full development of the rough character. The formula Rfrf can not, therefore, be used for partial-roughs.

There remains only one line of explanation. Partial-roughs must differ from full-roughs by possessing an independently inherited modi- fying factor (or factors). An incompletely dominant unit modifying factor will explain all of the results satisfactorily. Let us represent the wild condition (aperea, rufescens, cutleri) by rrSS. Let us suppose that the dominant mutation R is necessary for any roughness [of series I] and produces with rare exceptions at least reversal of hair direction on the hind toes (rough E) . The second mutation, s, when heterozy- gous, may permit roughness to extend to grades D or C ; when homozy- gous it permits roughness to reach grades B or A.

rrSS smooth Wild species, lea stock.

rrSs smooth Most tricolor smooths.

rrss smooth 4-toe, BW, BB, dilute stock, most Lima smooths.

RrSS rough E (rarely smooth) R 140, etc.

RRSS rough E (rarely smooth) 4003.

RrSs rough C or D (rarely E or B) . . . Most tricolor partial-roughs. RRSs rough C or D.

RrSS rough A (less frequently B) . . . . 4-toe, most Lima roughs. RRss rough A.

H2 tNHERITANCE IN GUINEA-PIGS.

This explanation fits very well the results which have been given qualitatively. As the numbers are rather small, and as it is necessary besides to assume some overlapping of class ranges, much emphasis Can not be laid on the quantitative results. Nevertheless, the fit is in all cases reasonably close. Let us take up the qualitative results in order.

(1) In some stocks there is very little variation in the rough char- acter and there is a wide gap between the lowest rough and smooth. Evidently if the 4-toe and similar stocks are pure for factor s, only full-roughs and smooths can appear (RRss, Rrss, rrss). Apparently 7 of the original 8 in the Lima stock were ss, while one, L6, was rrSs.

(2) When a wild species of cavy (smooth) or a smooth of certain tame stocks is crossed with a full-rough, the rough young are of low grade, rough C or D. This result necessarily follows from a cross of the type rrSS (wild, lea, tricolor) by Rrss (rough A). The rough young RrSs should be rough C or D.

(3) When a wild cavy is crossed with a partial-rough guinea-pig, rough young of the lowest grade, rough E, are produced.

rrSS X RrSs = RrSs -f RrSS + 2rr Sm C C E 2Sm

(4) Partial-roughs crossed together may give all grades of roughness from full-rough (A) to the lowest partial-rough (E). Most of the partial-roughs handled should be RrSs.

RrSs X RrSs = 3 Rss + 6 RSs + 3 RSS + 4 rr C C 3 A 6C. D 3E 4 Sm

(5) Full-roughs crossed together have never given partial-roughs. A full-rough, whatever its parentage, must be RRss or Rrss. There is no way in which factor S, necessary for partial-roughs, can be trans- mitted by full-roughs.

(6) Full-roughs, one or both of whose parents were partial-roughs, have given no partial-rough young on crossing with smooths of 4-toe stock. Smooths of 4-toe stock are all necessarily rrss and can not transmit factor S.

(7) Partial-roughs crossed with smooths of 4-toe or a similar stock give pari ial-rough young and also, in most cases, full-rough and smooth young.

• rrsa = Rrss + RrSs + 2 rr

C Sm (4-toe) A C 2 Sm

(8) Most partial-roughs crossed with full-roughs give a very similar result to the cross part ial-rough by 4-toe smooth, except that fewer smooth- arc produced.

RrSe x Hi- : Raa + 3 RSe + 2 rr

C A 3A 3C 2Sm

(9) The lowest grade of roughs (E) very rarely have either a full- rough parent or full-rough young. They also very rarely have a

ROUGH FUR.

113

smooth parent of such a stock as 4-toe. The parents and offspring of rough E (SS) must necessarily have at least one S (SS or Ss), while full-roughs and 4-toe smooths are ss. The three exceptional rough E's can only be interpreted as extreme minus fluctuations of type RrSs.

(10) Occasionally a smooth from a cross which produces rough E will transmit the rough factor, breeding like a rough E. The discovery of a smooth (RrSS) which had a rough C young one (RrSs) when crossed with a 4-toe smooth (rrss), apparently violating the dominance of roughness, is the kind of exception that proves the rule, coming as it did where predicted.

The descendants of male 4003 illustrate the theory very well. He was of constitution RRSS by theory.

Table 56.

		A	B	C	D	E	Sm
Pi Fi F2	RRSS X rrss		5	11 8	3	4003 5	4-toe 7
RrSs .
7 Rss + 13 RSs + 7 RSS + 9 rr . . . 7 A, B 13 C, D 7 E 9 Sm	8

Of those called rough B, 2 were close to rough A and 3 were close to rough C.

POSSIBILITIES OF LINKAGE AMONG ROUGH AND COLOR FACTORS.

In the mating just cited, factors R and S enter the cross from the same individual. The excess of full-roughs (Rss), probably 10 where 7 are expected, makes any linkage between R and S very unlikely. Another test is furnished by the cross of double heterozygotes, RrSs with 4-toe smooths, rrss, where it is definitely known whether R and S enter the cross together or apart. Cases which should show coupling if there is linkage are cross 54-1, 2, 4, 5, 6, 12, 15, 16, 17, and cross 72-5, 6, 7. Cases which should show repulsion are cross 61-1, cross 64-1, 3, and cross 72-1, 2, 8, 9, 10.

Table 57.

	A, B Rsrs.	C,D RSrs.	Sm rr —	Cross- overs.	Link- ages.
C, D Sm Coupling — RSrs X rsrs Repulsion — RsrS X rsrs	26 14	29 6	60 16	26 6	29 14
			32	43

The indication of linkage is too slight to be considered significant, especially in view of the excess of cross-overs in the F2 data.

Thus there is probably no linkage between R and S. It is interesting to analyze the data with regard to possible linkages of these factors with

I II

INHERITANCE IN CUINEA-PIGS.

any of the 5 known sots of color factors. Four of the sets of color factors, those in which non-agouti (a), yellow (e), brown (b), and albinism (( JJ arc the lowest recessives, are known to be independent of each other Part 1). As regards the pink-eye factor (p), it is merely known that eross-overs occur between it and non-agouti, albinism, and yellow, and that it is not an allelomorph of brown (i. e., pink-eye X brown give- Mack -eyed intense young).

I )ata on the possible repulsion of R and A are furnished by cross 72- 1, 2, 5, 6, 8, 9, 10 and cross 64-1. Coupling data are to be found in 64-3 and 66-1, 3.

Table 58.

	Ag-Rf ARar	Ag-Sm Arar	B-Rf aRar	B-Sm arar	Cross- overs.	Link- ages.
Ag-Rf B-Sm Coupling — ARar X arar Repulsion — AraR X arar	4 7	7 5	11 7	3 9	18 16	7 12
				34	19

There is an excess of cross-overs. The most probable interpretation is that there is no linkage.

Data on the coupling of A and S are to be found in crosses 70-1 to 9, 71-1 to 6, 72-1, 2, 3, 4, 5, 6, 8, 9, 10, 64-1, and 74-1. Repulsion data is to be found in 64-3 and 72-5.

Table 59.

	Ag-C ASas	Ag-A Asas	B-C aSas	B-A asas	Cross- overs.	Link- ages.
Ag-C, D B-Sm (4-toe) Ag Sm (Jem) B-Rf A Coupling — ASas X asas Repulsion — AsaS X asas	12	16 1	16	16 3	32 3	28 1
				35	29

A and S are quite clearly independent of each other.

Crosses 72-5 to 7 give a few data on the relation of R and S to C. We find in the case of R and C, 4 linkages to 4 cross-overs and in the case of S and C no linkages and 3 cross-overs, indicating probably independence in both cases.

From cross 72-7 we find that cross-overs can occur between R and E, R and B, and S and B. From 61-1 we find that cross-overs can occur between P and R, and P and S.

Summing up: R is quite certainly independent of S and A and cross- over- are known between it and all of the other known factors E, B, C, and P. 8 is quite certainly independent of A and cross-overs are known between it and B, C, and P, but not as yet E. It is hoped that more definite statements can soon be made on these points.

ROUGH FUR.

115

SUMMARY OF ROUGH TABLES.

Table 60, in which smooths are omitted, shows the closeness of fit of the hypothesis to the data as regards the inheritance of variations of the rough character.

Table 60.

Emitting young of L24. 2Omitting young of L6, L24. "Omitting young of A702, A605. 4Some of mothers may be ss or SS.

'Mothers L24, L6. •Mothers A702, A605. 'Including also one J cut.

Cross.	ss X ss	A	B	C	D	E
45 46 49 50 58 *59 260 63 365 66 73 75	A 4-toe A 4-toe	10 27 28 19 7 21 5 4 17 14 2 3	2 1 5 5 9 4 1
A Tri A Tri
A 4-toe Sm 4-toe
A Tri Sm 4-toe
B Lima B Lima
A Misc Sm Misc
A Misc Sm Misc A J cut Sm 4-toe A J cut A j cut
Total
157	27
Expectation ss
184	0	0
Cross.	Ss X ss	A	B	C	D	E
47 451 54 58 559, 60 61 64 665 770 771 72 74	C Tri A Tri, 4-toe Sm Tri A	10 6 34 4 2 3 1 8 10 12	1 1 1 1 1 3 3	5 3 29 3 1 2 2 8 9 6 1	1 3 13 1 2 1 6	1 2
C, D Tri Sm 4-toe, etc Sm Lima . . A, B Lima
C Lima Sm Lima C ruf . hybrid Sm 4-toe, etc Sm ruf. hybrid A
C, D h cut A G. p Sm \ cut .. A G.p
C, D J, £ cut Sm G.p Total
90	11	69	27	3
Expectation ss -4- Ss
100	100	0
Cross.	Ss X Ss	A	B	C	D	E
52 76	C Tri . . . C Tri	18	6	19 1	7 3	12 1
C § cut C J cut
Total
18	6	20	10	13
17	33	17

in.

INHERITANCE IN GUINEA-PIGS. Tabu 60— Continued

18 66 67

( !n>88.

53 69

Cross.

55

S.-i

38

\ Iri E Tri

Bm l-toe E Tri

A l toe, tri Sra pure lea.

\ Sm pure rut

Total

Expectation Sa.

Ss XSS

C, D Tri E Tri

C, D Tri Sm pure rut .

Total

B

Espeetation Ss + SS.

SS X SS

E Tri E Tri

Expectation SS

B

2

13

5

3

23

D

30

D

11

E

15

11

B

D

E

The interpretation given is no doubt open to objections. In some cases the ratios seem rather aberrant. This is in part due to the small numbers, but also to the overlapping of class ranges. In most cases rough B must be considered as full-rough genetically (Rss), but in some cases it is probably partial-rough (RSs). Rough E usually seems to be RSS, but in some cases must be heterozygous (RSs). It has not been demonstrated that factor S of the wild species is identical with the similar factor of the tricolor stock. If not identical, however, the latter stock differs from the wild by two mutations which neutralize each other, while if identical we can consider that the original tricolor stock had simply persisted in the primitive condition, never having had the rough intensifying mutation, s, of the fancier's roughs.

MINOR VARIATIONS.

Probably part of the minor variations in roughness are due to chance irregularities in development which are not hereditary. This is indi- cated by the Blight asymmetry not uncommonly present. This asym- metry seldom amounts to more than the absence of a member of one pair of rose! tee.

No Mendehan analysis has yet been attempted for minor variations, but certain hereditary differences between different stocks are quite

ROUGH FUR.

117

clear. The Lima stock shows a distinctly lower level of development of roughness than is found in the 4-toe stock or even among the full- roughs of tricolor stock. A large part of the variation and overlapping in the remaining experiments in which various stocks have been mixed is made intelligible by assuming that the residual heredity is unfavor- able for roughness in the wild species and especially favorable in the 4-toe stock. If we let 2+ stand for favorable and 2 — for unfavorable residual heredity, the wild species and presumably the primitive guinea-pigs are rrSSS — , while the good fancier's roughs, RRssS-f differ by at least three independent sets of factors, all favorable for roughness.

ROUGHNESS OF SERIES II.

It has been mentioned that irregularities in hair direction have been found in certain stocks which can not be classified by the grades which have been defined. The BW race is a highly inbred race. No indi- viduals of the pure stock have ever been observed to have roughness on the face, back, or toes, but many of them show irregular partings and crests along the chest and belly. It will be remembered that in series I ventral roughness appears only in high-grade roughs — grades A or B. Thus the characteristic roughness of the BW stock is nearly the least characteristic feature of series I.

The only distinction which has been made in these BW roughs is between strong-rough with two or more ridges and poor-rough with only one ridge or a mere trace of roughness. Table 61 shows the principal results.

Table 61.

	Smooth.	Poor rough.	Strong rough.
Smooth X smooth Poor X poor Strong X strong	11 14 5	6 1 5	6 1 16

It is clear that this roughness is due neither to a simple dominant nor to a simple recessive. Aside from this, the results are exceedingly difficult to interpret, since poor X poor gives more smooth than does smooth by smooth. Probably the results will become more harmonious when more data are obtained. It seems safe to conclude at present that this roughness is wholly independent of ordinary roughness in its causation.

Irregularity in hair direction on the back, not resembling anything in series I and not correlated with roughness of the hind toes, has been observed in a few individuals of Arequipa and lea stock. It does not seem to be like the BW roughness, but resembles the latter in the irregularity of its inheritance.

US INHERITANCE IN GUINEA-PIGS.

SUMMARY. The principal results which have been reached may be summarized as follows:

1 . A clasMlieat ion of guinea-pig fur, skin, and eye colors is given with definitions of fur colors in terms of Ridgway's charts (1912).

2. Undent color factors are conveniently classified as follows:

a. Factors which affect the distribution and intensity of color largely irrespective

of the kind of color.

b. Factors which govern the differentiation between yellow and dark colors in

colored areas of the fur. r. Factors which determine the kind of dark color in the areas with dark pigmenta- tion in fur and eyes, without influence on yellow areas.

Definitions of all known guinea-pig color factors are given on this basis and a table of the color varieties arising from combinations of these factors is given.

3. Genetic and biochemical evidence on the physiology of pigment formation suggests the hypothesis that the three groups of factors determine respectively the distribution and rate of production by the nucleus of the following substances:

a. A peroxidase which, acting alone, oxidizes chromogen in the cytoplasm to a

yellow pigment but is so unstable that it must be produced at a relatively high rate to give any pigment at all.

b. A supplementary substance which, united with the first, makes it a dark-pig-

ment-producing enzyme and of such stability that color develops at a much lower level of production of peroxidase than when the supplement is absent. Above the level at which both produce effects, the dark and yellow-producing enzymes compete in the oxidation of chromogen.

c. Additions to the second substance which cause variations in dark color but not

in yellow or in the competition between dark color and yellow.

4. There is a continuous series of variations in intensity of pigmen- tation in the yellow, brown, and black series and in eye color. The ordinary dilute guinea-pigs are found to be imperfect albinos in the sense that dilution is due primarily to a member of the series of allelo- morphs— intensity, dark-eyed dilution, red-eyed dilution, and albinism, with dominance in the order of increasing intensity.

5. A further step in the analysis of the continuous series of variations of intensity is taken in the demonstration that dilution is imperfectly dominant over red-eye and albinism as regards the yellow series of colors, and that dilution and red-eye are imperfectly dominant over albinism, as regards the black series. Smaller effects are due to the residual heredity of different stocks and to age.

I ividence is presented which confirms the hypothesis of Detlefsen (1914) that the light-bellied agouti pattern of tame guinea-pigs, the ticked-bellied agouti of hybrids between the tame guinea-pig and Cavia rufcsccns, and non-agouti (as seen in self blacks or browns) form a Beriee of triple allelomorphs in which light-belly is the highest dominant and non-agouti the lowest recessive. Evidence is presented which

GENERAL CONCLUSION. 119

indicates that Cavia cutleri possesses the same agouti factor as tame agouti guinea-pigs. Light agouti of Cavia cutleri and dark agouti of Cavia rufescens are thus variations in a character in two wild species which differ in heredity by a clear-cut Mendelian factor.1

7. There is a continuous series of variations between smooth fur and very rough or rosetted fur in guinea-pigs. The primary effects in this series are due to two independent pairs of allelomorphs. One factor, discovered by Castle (1905), is essential to any roughness of the common type, and is completely dominant over its allelomorph found in wild cavies and smooth guinea-pigs; the other, an incomplete recessive to its allelomorph in the wild cavies and some tame guinea-pigs, is necessary for the higher grades of roughness. Second-order effects seem to be due to the residual heredity of different stocks, and probably to non-hereditary irregularities in development. There is a roughness of a different type from the usual which is inherited independently.

GENERAL CONCLUSION.

Most of the successful earlier attempts at Mendelian analysis of heredity naturally dealt with variations which were obviously dis- continuous. But in nature such variations are much less common than apparently continuous series of variations. It was thus a common reproach against the Mendelian analysis that it dealt only with excep- tional conditions. The work of Nilsson-Ehle, East, and others has shown how quantitative variation may be brought under a Mendelian explanation. MacDowell (1914) presents data on size inheritance from this standpoint and discusses the literature up to that time. Recently two very interesting papers have been published (Dexter, 1914, Hoge, 1915) which analyze the heredity of certain very variable char- acters in Drosophila by means of linkage relations.

Several of the studies in this paper deal with inheritance in continu- ous series of variations. The only general statement which can be made about the results is that there is no general rule for such cases. Intermediates between varieties which mendelize regularly have been found to follow very definite modes of inheritance, which, however, are very different in different cases and could not possibly be predicted a priori. On the other hand, each mode of inheritance is exactly paralleled by cases among the most diverse groups of animals and plants. It may be interesting to summarize the modes of inheritance of inter- mediates which have been found.

An intermediate condition is sometimes found to be due to an inter- mediate variation of the essential hereditary factor involved, i.e., to an allelomorph. Thus yellows are intermediate between red and albino

1 It should be pointed out, however, that the original stock of Cavia rufescens used in these experi- ments included individuals of the light-agouti character as well as those classed as dark agouti. It seems quite likely that dark agouti arose as a recessive mutation in C. rufescens.— W. k. O.

12() INHERITANCE IN GUINEA-PIGS.

guinea-pigs in appearance, and we find an allelomorph intermediate in dominance between the intensity and albino factor to be responsible for their condition. Sepias are similarly intermediate between blacks and albinos and arc due to the same allelomorph of intensity and albin- ism. The Beries, light agouti of Cavia cutleri, dark agouti of C. rufes- ceru and black, furnishes another example due to triple allelomorphs.

In other cases, the intermediate type is an unfixable one, due to imperfect dominance. Thus cream is the heterozygote between yellow and albino. A "razor back " rough (rough C or D) is the heterozygote between a type smooth except for the hind toes (rough E) and a full- rough (rough A).

\ series of deviations from the original type may depend on the presence of a certain factor necessary for any deviation whose effect is modified to different extents by independently inherited factors. Rough A contains the same rough factor (R) as does rough E, but differs in possessing an independent factor variation (s) favorable for rough- ness. Most of the variation which we have ascribed to residual heredity probably comes under this head.

Deviations from type, which apparently form a natural series, may be due to wholly independent factors whose effects are merely super- ficially similar. A pink-eyed pale sepia superficially seems as good an intermediate between an intense black and an albino as does a black- eyed sepia, yet the former is due to a variation which is wholly inde- pendent of albinism; the latter is due to an allelomorph of albinism. White-spotted animals are sometimes called partial albinos and con- sidered as natural intermediates between the self-colored type and albinos, but genetically they are wholly distinct. Black, agouti, and self yellow form a series which is due to three allelomorphs in mice, but in guinea-pigs two wholly independent sets of factors are involved.

Final ly, we must recognize series of variations in which no Mendelian factors have yet been isolated. The series of white-spotted and yellow- spotted types and the series of polydactylous types are examples in guinea-pigs. Further, in all series of variations, to whatever extent analysis has been carried, there always remains some unanalyzed varia- tion. In many cases such variations are known to be hereditary and can be assigned to the residual heredity of particular stocks. Such unanalyzed variations, however, are probably in general complicated by variation which is not hereditary, due apparently to irregularities in development. If we can measure the importance of such non- hereditary variation by the extent of irregular asymmetry met with, it ery important in white and yellow spotting, in the variations in the development of extra toes on the hind feet, and is noticeable in varia- tion- in roughness.

In the continuous series of variations several of these phenomena have generally been found together. In the series from smooth to full-

TABLES. 121

rough we find a primary unit difference, a modifying factor, imperfect dominance in the effects of the latter, effects of residual heredity, and probably some non-heritable variation. In the series from red through yellow and cream to white we find multiple allelomorphs, imperfect dominance, and small effects due to residual heredity. In the series black through sepia to white, we find independent factors, multiple allelomorphs which show imperfect dominance, and rather prominent effects due to residual heredity and to age. This last series is interesting as at least a close parallel in appearance to the series of variations in human hair — black, brown, tow-color, to white. Thus in each case a complex of the most varied causes underlies an apparently simple continuous series of variations.

EXPERIMENTAL DATA.

EXPLANATION OF TABLES 62 TO 137.

Crosses 1 to 15 include all matings recorded by the writer which involve the inheritance of agouti and in which at least one of the parents had Cavia rufescens ancestry. A large part of the remaining crosses are non-agouti by non-agouti, producing only non-agouti young. All the young in which the agouti factor should produce a recognizable effect, if present, are classified under the heads Lb, Tb, and Non, which mean light-bellied agouti, ticked-bellied agouti, and non-agouti, respectively. Most of these are the typical (black-red) light-bellied or ticked-bellied agouti or black. Those which are not typical, e. g., brown-red agouti light-belly, red-eyed sepia, etc., are described further under the column "Remarks." Those young in which the agouti factor can produce no visible effect, even though present (albinos, reds, yellows, and creams), are described under the column "Unclassified." Thus the exact color of every one of the young from each mating can be found from the tables, with the exception that white and red spotting are not noted. The matings in each cross are numbered in the first column. The number, description, and descent of the mother and father are given in the second and third columns, respectively. As in the case of the young, black-red agouti light-belly or ticked-belly or black, depending on the heading of the column, are understood where no description is given. The descent is indicated in most cases by a reference to the mating from which the animal was derived. Thus 36-4 means mating 4 of cross 36. In other cases the stock is indicated as BB or BW. The symbol ArF2 means F2 from crosses of Arequipa cf 1002 with guinea- pigs. In some cases merely the amount of Cavia rufescens blood is given. Thus M49, in the first cross given, was an ordinary ticked- bellied agouti from mating 9 of cross la. Referring to this mating, we see that his parents were female 84, a black of BB stock, and male All 21, a ticked-bellied agouti with TW Cavia rufescens blood.

122

INHKHITANCE IN GUINEA-PIGS.

( rossee 16 to 44 include all matings recorded by the writer in which there was dilution or red-eye in either parents or offspring, except for a few cases among the Cavia cutleri hybrids and cases of intense by dilute with only intense young. Some other crosses are included for special reasons bearing on the inheritance in the albino series. There is some repetition from matings outside of 16 to 44, but most of those outside are intense by intense, with only intense and, in some cases, albino v( Ring. As in the agouti crosses, all matings are numbered in column 1. The number, description, and descent of the mother and father are given in columns 2 and 3, respectively. All the offspring are classified under the heads Int, Dil, RE, or W, winch stand for intense, dilute, red-eye, and white (albino), respectively. A further description of all except the albinos is given under the column "Remarks." The attempt has been made to give the grade of dilution at birth for every dilute or red-eyed animal where known.

Crosses 45 to 57 give the data on the inheritance of rough fur in the 4-toe and tricolor stocks. As before, the matings are numbered. The young are classified under the heads A, B, C, D, E, and Sm, which refer to the grades of roughness defined in the paper and to smooth.

The parents and offspring were black (usually with red and white blotches) except for a few cases which are all noted. Such a symbol as red-B means a red of grade rough B.

Crosses 58 to 62 give the results in the pure Lima stock and 63 the results in the cross of Lima with other stocks. Where no color is given black is always to be understood.

Crosses 64 to 66 give the matings involving rough fur among Cavia ri'fi scens hybrids which were recorded by the writer.

Cross 67 gives crosses of pure lea with rough A stock.

Crosses 68 to 78 give all the data in matings involving Cavia cutleri ancestry made by the writer.

The following symbols are used :

AL'I.borAg = Black-red agouti, light-belly.

AuTb = Black-red agouti,ticked-belly.

B - Black.

BrAgLb = Brown-red agouti, light-belly. BrAgTb = Brown-red agouti, ticked

belly. Br = Brown.

B = Red (black-eye).

R(Br) = Red (brown-eye).

81 VgLb = Sepia-yellow agouti, light-

bi-Uy. SI Vj'I'b = Sepia-yellow agouti, ticked-

belly. = Sepia.

BrYAgLb = Brown-yellow agouti, light- belly.

BrYAgTb = Brown-yellow agouti, ticked- belly.

LBr = Light brown.

Y = Yellow (black-eye).

Y(Br) = Yellow (brown-eye).

Cr = Cream, used in compounds

like Y.

SAg(R) = Sepia- white agouti (red-eye).

Sep(R) = Sepia (red-eye).

W = White or albino.

Red (p) = Red (pink-eye).

Sep (p) = Sepia (pink-eye).

In such •■xprossiona as S,Y,Ag the numerals stand for the grades denned in the text. In crosses 1 1 6, Lb and Tb are used at the heads of the columns to include any light-bellied or boked-belked agouti Non means non-agouti.

V B, C, D. I ., .Hid Sm are used for grades of roughness and for smooth.

TABLES.

123

Table 62.

Cross 1. — Matings of non-agouti (aa) with ticked-bellied agouti (A'a).

Each of the latter known to be heterozygous because of a non-agouti parent. Expectation: A'a X aa = A'a + aa (1 AgTb : 1 Non-Ag).

la. Mother non-agouti, without rufescens ancestry.

No.

9 Non-Ag.

o^AgTb.

Lb.

TbNon

Remarks.

Unclas- sified.

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15 16

17 18

39 9 49 9 39 9 39 9 29 9 65

29 9 39 9 84 C22 C35 39 9 3W 3 W

BW... 4-toe. . BW.. . 4-toe. . BW. . . BB. .. BW. . . BW... BB. .. Misc.. Misc.. 4-toe. . BW... Misc . .

la-9.

ld-16 ,

M49

....Do...

B5

... .Do...

B27

B30

B69

B191

A1121

A1474 jJg . . . .

Do

B171 la-4.. B117SCrAgTb ld-11 . Do

ld-14 lo-l. la-1. la-5. A---

D44 Sep 16a-3 AA244 Sep 2-12 . BW43W BW.. Sep (R) S.Am

....Do

....Do

D113BCrAgTb 3a-7 AA433a 36-4

6 W 1 W 5 W

1 W

2 W

2 W

4 W

Total .

62

62

3SOAgTb, 2 Sep.... SCrAgTb, BrCrAgTb,

3 Sep.

SCrAgTb

SCrAgTb

SCrAgTb, Sep

3 AgTb, SAgTb(R), 2

Sep(R).

1 W 5 W

1 W

2 W

1 W

No.

10 11

lb. Mother non-agouti, with rufescens ancestry.

9 Non-Ag.

A443

A1390

A1227 W /A1413 \A1291 W

A1309 W

A 1407

A1413

Ml 15 W /M114 \M90 Br

M90 Br

A1330

A

64 • • • •

ii-...

I 3Z. ...

A — A--

16-7..

le '

T2£

cfAgTb.

A469 \ . .

A1050 gSj.

^A781 if.

)A1449 A-

A1513 3^... A1449 ^ . . ....Do .... M189 lc-3.

....Do

Do...

A1331 rlg

Total .

Lb

Tb

17

Non

13

Remarks.

LBr

SCrAgTb .

Br AgTb . Br

SCrAgTb, Sep. BrCrAgTb, Sep

Unclassified.

W

R, R(Br), Cr(Br)

Cr W

1L.M

INHER] I \N< K l\ GUINEA-PIGS.

T vhi.k 62 — Continued.

lr. Male genetically, but not visibly ticked-bellied agouti.
1 8 i B 6 7 >* '.»	. \..i,- U-	cfAgTb.	Lb	Tb	Nod Remarks.	Unclas- sified.
131 \\ G. I-. i '.i < ; n	A.412 R{ Do	Br) ,l» . .			2 3 1 1 3 1 1	Se	p
G p	Do		3 3	s< rAtf] 3 S( "r A	Tb
;,s Bep I >il . 17 3l)('nB) Dil	B42 W Lo-3 Do		gTb
Se SC	p. . .		2 \Y W
■ Hi nil	Do .		1 2 1 1	'rAcTb
M -".'-' \|;(v, X	1 ) 1 8 W lc-5 Do		SCrAgl	n>, 2 Sep
	4 \V
M353 «3.	Dn		Se	p
Total			11	12
h/. Male non-agouti, without rufescens ancestry.
No.	9A*Tb.	d"Non-Ag.	Lb	Tb	Non	Remarks.	Unclas- sified.
1 2 3 J 6 6 7 B 9 10 11 12 13 14 15 16 17 18 19 20 2] 24	a 06 \L'i.ii \	L66 4-to 2966 BB 3013 BB Do .	e			o 1 3 5 2 6 1 2 2 5 1
'A l.Mt Au'I.I) i		'1	4 6 1 7 1
\;in ,'„ A3 1 1 ,'„ 1367 & Aim; j'g A.1068 it A 1171 3', A.1058, A 1 1 7 1 ^ Do
	Do .
	Do
	Do
	Do .
	2996 BB 1357 BW Do .			3 5 2 1 2 2 3 1 1 9 1 3 5 2 2
				2 W
A 1117 .,K Do				SCrAgTb. .
	2996 BB 3013 BB 1357 BW . Do .
Do			4 1 1 1 2 11
A 1450 3't All 17. Al 150 aV A15sl' iA A1583 ei4 A 1C77 a\ A.1678 8'4 B8 ld- B23 ld- B26 1-/ Ml 13 lb-
				w
	2996 BB Do .
	Do . .
-7	Do .
Do .		2 7 1 1 2
-12	. . Do .
11	Do .
-7	C20 Mis 86 W BW	;
Mtl_' BzCrAgTb H^-10....			2 Sop
1	'1	61	63

TABLES.

125

Table 62 — Continued.

le. Male non-agouti (genetically) with rufescens ancestry.
No.	9 AgTb.	o71 Non-Ag.	Lb	Tb	Non	Remarks.	Unclas- sified.
1 2 3 4 5 6 7 8 9 10	A1146 -&	A504 W &		2	2 Sep
B132 ld-3	M293 Y 42-14		2 1	SCrAgTh
B95 ld-4	Do
B24 ld-4	Do		1
B52 la-3	M201 W 42-13		2 2
B33 1<Z-18	Do				W w w
B110 la-1	Do
Bill la-1	Do		1 1 1	2
B128 la-1	Do
B23 ld-12	Do
Total		10	5

xAgLb, but agouti known to be derived from C. rufescens. Summary of Cross 1.

No.

la 16 lc Id le

aa

9 9 Non-Ag (g.p.)

9 9 Non-Ag (hybrid) . .

9 9 Non-Ag

cfcf Non-Ag (g.p.)

C? <? aa ( Y or W ) hybrid Total

cf^AgTb

c?c?AgTb

cfcfA'a (RorW)

9 9AgTb

9 9AgTb

Lb

Tb

62 17 11 61 10

161

Non

62 13 12 63 5

155

Table 63.

Cross 2. — Matings of ticked -bellied agouti (A 'a) with ticked-bellied agouti (A 'a), in which both are known to be heterozygous because of the parentage in each case.

Expectation: A'a X A'a = A A' 4- 2A'a + aa (3AgTb : 1 Non-Ag).

No.

6

7

8

9 10 11 12

12a

13

14

15

16

17

18

19

20

21

9 AgTb.

B15 B58

ld-6.. ld-15.

B59 ld-15.

B68 la-1 . .

A529 BrAgTb & . . . .

A913 gV

A1273SCrAgTb &.

Do

A780 fa .

A1306 jig

A1561 & .

A1566 3S .

A1566

A702

A1450

A1058

A1523

AA176

AA175

M78

mho

D26 SCrAgTb Total . . .

A A A

A

41-4

41-4 9-5. 16-6 lc-4

c? AgTb.

B118

ld-6.

Do

Do .Do

AA15

Do

AA16

A1121

A781

A1307

A1050

Do

A-

A-- A-

JL 64 • •

TS8 '

A-

A--- A---

36-4.

AA15 AA16 AA433a

Do

A1449 & . . .

AA177 SCrAgTb 41-4.

Do

A1161 ^2...

A1170 &...

D33 SCrAgTb lc-6.

Lb

Tb

66

Non

19

Remarks.

3 BrAgTb, SYAgTb.Sep

2 BrAgTb

SCrAgTb

BrAgTb, SCr

AgTb, BrCr AgTb BrAgTb

3 SCrAgTb . . .

BrYAgTb .

BrAgTb, SCr

AgTb, Sep

SCrAgTb....

Br

BrAgTb.. SCrAgTb.

Unclas- sified.

3W

3R.Y 2W 4W

W W W W

W

1 2< i

INHERITANCE IN GUINEA-PIGS.

Taule 64.

M:iinm- of ticked-bellied agouti from cross 2 or 12 (A'A', 2A'a) with non-agouti made in order to test for the presence of homozygotes.

•.,.„ \ \ X aa = A/a (all AgTb)

•; aa = A 'a + aa (1 AgTb : 1 Non-Ag).

3a. Heterozygous females.
1 2 '-» 5 '7 l8	U:Tb.	C? Non-Ag.	Lb	Tb	Non	Remarks.	Unclas- sified.
Ml'.' 2 16.. M203 -' 19. . AA2I1 2-6... A V.'in 12 - AA257 r_'-2.. LA285S< rA-Tl. 12-7.. LA242SYAgTb 12-S.. AA240.AA242 12-8..	393 4-toe ....		2	1 1 2 1 2 1 2 1
Ml 10 Sep 42-11. . .
( 120 Misc. . . .
A1040 fa		2 2 1 1 2
A 1010,356 fa, 4-toe. .■!'.».•: 4-toe I5Sep(R) 21-1
BCrAgTb, 2 Sep .
A1040 fa
7 fl'HKl!<'S		10	11
3b. Possible homozygous females.
No.	9 AgTb.	o" Non-Ag.	Lb	Tb	Non	Remarks.	Unclas- sified.
l o ■\ 4 5	A A 209 2-11.. AA212 BrAgTb 2-6.. . LA213 2-12.. AA217 2-7...	C21 Misc		4 2 8 8 3
C20 Misc
C21 Misc. . . .
Do
\ A298 2 13. .	356 4-toe
5 females		25

'Not certain that both parents were heterozygous (A'a).

TABLES.

127

Table 04. — Continued.

3c. Heterozygous males.
No.	9 Non-Ag.	cfAgTb.	Lb	Tb	Nor	i Remarks.	Unclas- sified.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26	M79 W 3»z fM72 lc-2 . . . \M86 9-1 ... M42 LBr 42-12 . .	M77 BrAgTb 2-16 1 Do		1 2	1 2 1	BrAgTb, Br.. . BrAgTb, Br	3 W
/ AA197 2-10...
M44 Cr 42-1 M99 42-1 MlOl 42-1 {A1407 &..	9,	Do		3 1 2 5 1	SCrAgTb
3?	Do
3'	Do		1 3
	JAA199 SCrAgTb 2-12 AA223 BrYAgTb 2-9 Do
IA1413 3*s S6 sfcr....
S15 2ie. A1659 2ig. B 9 Misc S2 Misc A1665 Misc S22, A1674 (5fs'		2 2 1 2 1 *3	Br.
	Do		2 2
	Do		BrAgTb .
	AA226 2-13 . . .
Do		1 3 2		R
	\....Do
Wis-.. B7 ld-7 . . .	J AA235 12-1 . . . Do	ll
B21 ld-9 B28 ld-1 M168 3^..
4	Do		1		W W W
	Do
M169 J.,	Do			1 3 6 4 1 1 2
M177 lc-2 4 9 9 Misc B28 ld-1		Do		2 2 3 4	BrAgTb, LBr SCrAgTb, Sep, LBr. Br
t	2AA241 SYAgTb 12-8 2AA284 12-7 2AA299 12-8 Do
M183 vW. .	SCrAgTb. .
M261 Sep 7-7.		Sep
AA58 Ve • •	. . Do . . .		2 2
M261.AA58 .	Do
9 males
1	40	38
3d. Homozygous male.
No.	9 Non-Ag.	d"AgTb.	Lb	Tb :	Non	Remarks.	Unclas- sified.
1 2 3	S6 9h S15 5is.	Ai	^.253 SCrAgTb 2-7 . . .		6 4 2		2 BrAgTb
Do
A1659 „1~	Do			2 SCrAgTb . . . .
12

1A litter with an unexpected AgLb. Paternity not wholly certain. 2Not certain that one of parents was heterozygous (A'a).

Summary of Cross 3.

	AgTb, from cross A'a X A'a, tested by cross with Non-Ag.	Lb.	Tb.	Non
3a 36 3c 3d	7 9 9 A'a 5 9 9 A'A'(?) 9c? o71 A'a lc? A'A'	1	10 25 40 12	11 38

1 28

INHERITANCE IN GUINEA-PIGS.

Table 65.

Cross 4. — Ticked-bellied agoutis, known to be homozygous because of test (cross 3). or

parentage (cross 4), crossed together.

tpeetetkm: A'A' X AW	= A'A' (all AgTb).
	9 AgTb.	cTAgTb.	Lb	Tb	Non	Remarks.	Unclas- sified.
l 2 8 4 5 6	LA213 -' 12. . \ V217 2-7. .	AA253 SCrAgTb 2-7 . . Do		9 11 3		2 SYAgTb, SCrAgTb, BrAgTb, BrYAgTb, BrCrAgTb 3 SYAgTb, SCrAgTb 3 SCrAgTb . . .
AA613 4-1.	Do
AA671 4-1.	.. Do .			w
AA577 4-2... \ M)06 SYAgTb 4-2 . . .	AA573 BrAgTb 4-1		1 2		BrAgTb
Do		SCrAgTb
Total
	2G

Table 66.

Cross 6. — Homozygous ticked-bellied agouti (A'A') crossed with heterozygous ticked-bellied

agouti (A'a) or with non-agouti (aa).

Expectation: All AgTb (A'A' or A'a).

No. 1 2 3 i 5	9 AgTb or Non-Ag.	cMgTb.	Lb	Tb	Non	Remarks.	Unclas- sified.
M 181 BrCrAgTb 15-15.. Do	AA253 SCrAgTb 2-7 AA573 BrAgTb 4-1 AA573 BrAgTb 4-1 Ml 16 Sep 42-11. AA670 SCrAgTb 4-1		1 5 1 3 3		SYAgTb 3 BrAgTb, 2 BrCrAgTb BrCrAgTb . . .	W 2 W
M442 BrCrAgTb 1&-10. . M296SAgTb 14-4... Dl94Sep(R) 26-2...
3 SYAgTb . . .
2 SCrAgTb, SAgTb(R)	2 W
Total		13

TABLES.

129

Table 67.

Cross 6. — Matings of non-agouti hybrid (aa) with homozygous light-bellied agouti (AA). Expectation: A A X aa = Aa (all AgLb).

6a. Female non-agouti.
No.	9 Non-Ag.	cTAgLb.	Lb	Tb	Non	Remarks.	Unclassified.
1 2 3	A605 i A642 i	2597 G.p Do	2 2 5 9
A842 i	Do
Total
66. Male non-agouti hybrid.
No.	¥ AgLb.	cf Non-Ag.	Lb	Tb	Non	Remarks.	Unclassified.
1 2 3 4 5	15a G.p . . 3520 Cr(Br) G.p . .	A674 Sep J . . . Do	6 3 7 3 1
		2SCrAg,BrCrAg.	Y(Br),2Cr(Br),W
3a G.p . . 11a G.p.. 3392 G p . .	1040 &.. A504 W ^g . . A1539 &..
		SCrAgLb . .
Total	20

Summary of Cross 6.

No.	aa (hybrid).	AA (g.p.).	Lb	Tb	Non
6a 6b	9 9 Non-Ag .... o71 cf Non-Ag ....	c? c? Ag Lb 9 9 Ag Lb	9 20
Total .	29

130

[NHERITA.M E IN GUINEA-PIGS.

Tabu 68.

Matingfl "f non-agouti (aa) with light-bellied agouti (Aa) of rufescens ancestry, known t«> transmit non-agouti, because of a non-agouti parent.

Expectation: Aa X aa = Aa + aa (1 AgLb : Non-Ag).

No.

10

11

la. Female AgLb.

9 AgLb.

.V

&601

AMI A963 A1310

A bill A1324

M102

Do

3CrAgLb 106-7. . D95S(rAnbb in', s. . D61 SCrAgLb .; 13-5. . . D63 SCrAgLb ^ 13-5...

:,'■: 66-1

D69 SCrAgLb 18-5. M I. '5 SCrAgLb 13-7.

Total

cf Non-Ag.

103 224 A718 166

Do

A719 W A462 W M2

20W

toe. toe.

■1- 4-

4-toe .

&■■

BW

HW36WBW. ...Do

Do

Lb

14

Tb

Non

18

Remarks.

Sep

SCrAgLb

3 SCrAg, 2 Sep

2 SAg(R), Sep, Sep(R)

2 SCrAg, Sep, .

Unclas- sified.

W

3 W

7 W

W

76. Male AgLb.

N...

| 9 Non-Ag.

67 G.p . .

D43 Sep 16a-3

M.'36Sep(R) ArF2. D45Sep 16a-3

cfAgLb.

Lb

Ml 23 13-2.. ' 2

D94 SCrAgLb 106-8 . 1 1 M331 BrCrAgLb 42-10. i. . . DS4 SCrAgLb 106-8.! 1

Total 4

Tb

Non

Remarks.

SCrAg, Sep .

SCrAg, Ag.

Unclas- sified.

W

w

	8i imauy of Cross	7.
No.	' Aa (hybrid).	aa	Lb	Tb	Non
la 71,	9 9 AgLb.... cfcf AgLb....	cfcf Non-Ag. . 9 9 Non-Ag. .	14 4		18 3
Total	18		21

TABLES.

131

Table 69.

Cross 8. — Matings of ticked-bellied agouti (A'a) with homozygous light-bellied agouti (AA). Expectation: AA X A'a = AA' + Aa (all AgLb).

8a. Female AgLb.
No.	9 AgLb.	cfAgTb.	Lb	Tb	Non	Remarks.	Unclas- sified.
1 2 3 4 5	03 G.p	B5 ld-16 Do	5 4 4 8 4
3392 G.p			SCrAg
02,03 G.p	A1155 1*6
	A1474 -L
20a G.p	Do			SCrAg
Total
25
8b. Male AgLb.
No.	9AgTb.	cfAgLb.	Lb	Tb	Non	Remarks.	Unclas- sified.
1 2 3 4 5 6 7 8	B33 ld-18 B36 ld-18	2597 G.p Do	3 5 2 3 6 2 2 2
B37 ld-20	Do
B52 la-3	Do
/B40 W la-3 \B37, B33 above A702 ^	) Do
/ Do
A913 3^	Do
1AA606 SYAgTb 4-2	724 SAe(R} (lea) ....			2 SYAgLb . .
Total
25 50
Total cross 8

!AA606 was A'A'

Table 70.

Cross 9. — Matings of light-bellied agouti (Aa) with ticked-bellied agouti (A'a), both known

to be heterozygous with non-agouti.

Expectation: Aa X A'a = AA' + Aa + A'a + aa (2 AgLb : 1 AgTb : 1 Non-Ag).

No.	9 AgLb.	cfAgTb.	Lb	Tb	Non	Remarks.	Unclas- sified.
1 2 3 4 5 6 7 8	3015 G.p M46BrCrAg 13-3 M57 13-1	A1474 TV A1170 & Do	6 2	2	5		3R W
BrAgLb
1 1 1	1 1
A1310 ^ Do A1311 aSs Do	A1449 3*2			Cr
A1513 g^
A1449 aS, A1513 ^ A1HKO J_	1 5 2
1	2 1
A 1 02fi k1™
Total
16	6	10

132

INHERITANCE IN CUINEA-PICS.

Table 71.

M /t>. — M:vt ihk- of tight-bellied agouti from such crosses as 8 and 9 (AA', Aa) with non- •uti. made in order to test whether light-bellied agouti can transmit both ticked-

bellied agouti and non-agouti.

I ipecUtion: AA' X aa = Aa + A'a (1 AgLb: 1 AgTb). or Aa X aa = Aa -f- aa (1 AgLb : 1 Non-Ag).

10a. Females Aa.
So	9 A*]	cf Non-Ag.	Lb Tb Non	Remarks.	Unclas- sified.
1 a 3 4 5 0 7 B	B120 85-2 B140 Bo-2 Do	C20 G.p M328B-Y 42-17.... 20 W BW Ml 16 Sep 42-11	2 . . 3 .. 3 . .	3 1 2 2
M195 9-7 M217 8o-3 M282 15-12 A1562 W 3'f	2 Br
356 4-toe. . . . Ml 16 Sep 42-11.... AA83 Jj	1 . . 2 . .	2 2 2
Sep
A1691 86-7	Do	6 ..	6
7 females
17 ..	. 20
106. Females AA'.
\"	9 AgLb.	c? Non-Ag.	Lb T	b Non	Remarks.	Unclas- sified.
l 2 3 4 5 fi 7 8 9 10 11 12	ASs'.) Br AgLb ,V . A499 BrAgLb fa . . . . A1688 86-6..	A511W fa AA83 Jj Do	1 . .			2 W,R R
i : 2 : i i 3 : 4	I .... 1 ....	BrAgLb.BrAgTb Br AgTb
A1690 86-7..	Do		3R
B139 8o-2.. Do B141 SCrAgLb 8a-2.. Do	M328B-Y 42-17.... 20 W BW	1 .... 1 .... 1 .... 1 }	3 SYAgLb . . .
SCrAgTb
M328B-Y 42-17.... 20 W BW 356 4-toe 393 4-toe .... Ml 16 Sep 42-11 ... Do	i : 2 2 . 2 2 1	SCrAgLb, 3 SY AgTb 2 SCrAgLb, SCr AgTb	6 W
M25 9-1... M27a 9-1... M82 9-7. M92 Sa-4..
5 ....
i .... 1	SCrAgLb, BrY AgLb, SYAgTb
10 females
18 3	D

TABLES.

133

Table 71 — Continued.

10c. Males Aa.

No.

6

7

8

9

10

11

12

13

14

9 Non-Ag.

75

08 W

09 W M7 M7 AA58 M183 M261 Sep S7

S2

A1665

S7, A1665

B137

B133

B98

D14S W

BB..

4-toe . 4-toe .

A-.-

xV-

&•• l

16 • ■ ■

7a-7.

5S« • •

256

2B6

la-1.

ldr-S.

la-3. lc-8.

cfAgLb.

A581

Do

... .Do M133

16'

J'

Xa-4 .

Do

Do Do

M205

Do

Do

Do

B155

k...Do.

8a-4.

8a- 1 .

D240 BCrAgLbl4-5.

5 males.

Lb

26

Tl)

Non

33

Remarks.

BrAgLb

SCrAgLb, Sep

SCrAg, Sep

Unclas- sified.

4R.Y?

lOd. Males AA'

No.

1 2 3 4 5 6 7 8 9

10

11 12 13 14 15 16 17 18 19 20

9 Non-Ag.

31

M255

M253

M256

M254, M256

AA279

C22, AA278

B9 9

M168

M169

M171

M177

M86

M79W

B31

B53

B54

C29

B9 9

M119

AA174

5 males.

Misc

la-10

la-10

la-10

la-10

3a-3

G.p., 3a-3 (above) . .

A

A

lc-2 . . .

9-1 ... .

A

la-1 . . . la-3 . . . la-1 . . . G.p . . . (above)

A

14-1 . . .

d"AgLb.

M138 9-1. . ....Do... ....Do.... ....Do... ...Do... ....Do.... ....Do .... ...Do... . 8a-4.

M91

....Do.

...Do

M210 15-14.

Do

B121 85-2..

Do

Do

Do

Do

B130 8a-l. Do

Lb

45

Tb

50

Non

Remarks.

SCrAgLb

BrYAgLb, 2 BrCr AgTb

SYAgTb

BrAgLb .

Unclas- sified.

	Summary	of Cross 10.
No.	AgLb (Aa or AA') tested by cross with Non- Ag (aa).		Lb	Tb	Non
10a 10c 10b lOd			17 26		20 33
Total		43		53
		18 45	30 50
Total	63	80

134

IMIKIMTANCE IN GUINEA-PIGS.

Table 72. Cross 11.— Light-bellied agoutis (AA') crossed together.

Botfa parents known to carry recessive ticked-belly by test (except in the case of AA533 with one young).

peetation: AA' \ AA' = AA + 2AA' + A'A' (3 AgLb: 1 AgTb).

1 2 8 4 6 6 7 B	9A*Lb.	o"AgLb.	Lb	Tb	Non	Remarks.	Unclas- sified.
11890 s''-7 M25 9-1	M91 8a-4 Do	4 7 5 4 1 1 1 2 25	2 3
	4SYAgLb
M27a 9-1	. .Do
M25 \! !7a 9-1	. . Do
Hli'i Sa '	Do .
M25 M27 B139 . ...	. .Do	3		BYAgTb
\ 1533 11-1	.Do
\ \>n 11-2	Do	1 9		SYAgTb
Total

Table 73. Cross 12. — Miscellaneous matings of ticked-bellied agouti with ticked-bellied agouti.

No. 1 o 3 4 5 6 7 9 10 11 12	9 AgTb.	cfAgTb.	Lb	Tb	Non	Remarks.	Unclas- sified.
A A_Ui BrCrAgTb 2-5.... M19 2-16...	AA16 A		3 3 1 2 3 3 4 / 1 4 2 1	1	SCrAgTb, LBr	w w R w
Do
AA497 SYAgTb 10d-ll. \ A598 3c-22 A 1523 g'j M203 2-19... AA202 2-7 A A 206 2-9.... Do AA342 12-8...	AA2S4 12-7 Do		1 2	LBr BrAgTb . . . SCrAgTb. .
AA199 SCrAgTb 2-12 AA284 12-7 AA15 &
Sep
2 SCrAgTb 4 SCrAgTb
AA1 77 SCrAgTb lfr-3 AA507 3c-22 Do
A1058 & AA242 SCrAgTb 12-8...	M298 15-16 B 117 SCrAgTb ld-11			BrAgTb . .
SCrAgTb . .

Table 74. Cross 13. — Miscellaneous matings of light-bellied agouti guinea-pig with non-agouti hybrid.

No.

9AgT.b.

G.p 3444 G.p

01 G.p

G.p 3266 BrYAgLb G.p 3220BrAgLb G.p 2718Ag(R) G.p

241 SA R AH

29 9G S R AH- I-', i 21-1

cTNon-Ag.

>A1539

i.i

A426R(Br) ,', .

A678"W ,V

Do

M333 Y

M34 Sep

Mir.c, H M34 Sep

.V •

Kir l .

42 L7.

liic 1

Lb

Tb

Non

Remarks.

SCrAgLb .

7 BrCrAgLb, 2 LBr

2 Br AgLb

4 SCrAgLb

2 Sep

3 SCrAgLb

2 SAg(R)

SCrAg, Sep

Unclas- sified.

3R

4 W

TABLES.

135

Table 75. Cross 14- — Miscellaneous matings of light-bellied with ticked-bellied agouti.

No.	9 AgLb.	cfAgTb.	Lb	Tb	Non	Remarks.	Unclas- sified.
1 2 3 4 5	f20a, 5a G.p. . . . \3015, 3014 G.p 20a, 5a, 3014 G.p AA173 14-1 . . . M82 9-7 4 9 9SAg(R) ArF2.. .	}A412R(Br) & A1474 -^ AA199 SCrAgTb 2-12 . . . A1161 &	3 3 1	1 1 2 1	1		3R
4	SCrAgTb ....
AA508 3d-2 . . .	13	9 SCrAg, BY AgTb

Table 76. Cross 15. — Miscellaneous crosses involving the inheritance of agouti in rufescens hybrids.

No.

9

10

11

12

13

14 15

16

17 18

19

20

21 22

23 24 25 26

27

9 Misc.

AA171 R 14-1 .

M137R 9-1.. Do

M181 BrCrAgTb

A1472 BrAgLb /A1562 W \A1688 AgLb /M27a AgLb \M 19 AgTb

M 175 AgLb

AA242 SCrAgTb /Ml 95 AgLb \M203 AgTb /B122 Agl.b \AA212BrAgTb /M92 AgLb \M106 Y? /M85 R \M82 AgLb / M56 AgLb \M50 AgTb

AA28W

15-15

i

16 • • •

JL

16 • • •

Hb-6. 9-1.. 2-16. 8a-3. 12-8. 9-7.. 2-19. Hb-2. 2-6 . . 8a-4. 10c-2 lb-5. 9-7. 13-1. lb-7.

A--

A ■ •

lb-5

16 • •

iV-

16 • • JL 16 • ■

11-2

A1523 AgTb

A1413B

M84 R(Br) TA556 AgLb JA587W [A533 Y /A495 AgLb \A867 B /AA621 SYALb \AA621 SYALb ....

3 9 9 SAg(R) ArF2

198W ArF2

2 9 9 W ArF2

D125 W la-13

D427 W la-14

D86 W 76-3 .

c^Misc.

AA1161 AgTb^... AA286AgTb 3c-4. M29Br lb-6.

C20B 163 B

G.p..

4-toe

JL 64

l

16

AA83B

393 B

A1040 B

Ml 16 Sep 42-11

C20B G.p..

A1513

AA1 161 AgTb g>5.

Al 170 AgTb ,&. Al 513 AgTb 3V

M83 AgLb 9-7

Do

A1161AgTb j^.

JL 32 ■

Lb

Tb

Non

•104 B

163 B

4-toe .

4-toe .

BW36 W BW.

724SAg(R) lea. M224 BrAgLb 9-2 .

M291 B M2B

133 SAg(R) 133 SAg(R)

3 32-

A-

24-

I20BWAg(R) 24-2.

3 21

1 2

1

Remarks.

BrAgTb . BrAgLb .

BrAgLb, SCrAg Tb, BrCrAg Tb

SCrAgTb

SSYAgTb

10 SAg(R), SAg Tb(R)

SAg(R), Sep(R) SAgTb(R), Sep

(R) 2SAg(R),3Sep

(R)

Unclas- sified.

R,Cr

2R W

R,2Cr

W

3R

W

3 W

136

INHERITANCE IN GUINEA-PIGS.

Table 77. Cross 16. — Matings of dilute with albino of intense stock.

Expectation: CdCd X CaCa - CdCa (all Dil).

CdCr X CaCa - CdCa + CrCa (1 DU : 1 RE). CdCa X CaCa = CdCa + CaCa (1 Dil : 1 W).

16o. Females CdCd.
No.	9 Dil.	o" W (intense stock).	Int	Dil	RE	W	Remarks.
1 a 3 4	\ \fi21 SA'iAg 58 Si-Pi Do	39-4 Dil	BW36 W BW 75 W BW 20 W BW... B42 W la-3		1 2 3 3 9			S6Cr6AgTb 2Sep3 2Sep3, Seps-Crj 3 S6Cr5AgTb
Do
Total
165. Females CdCa.
No.	9 Dil.	c?W (intense stock).	Int	Dil	RE	w	Remarks.
1 2 3 4 5 6 7 8 9 10 11 12 12a 13 14 15 16	15 Sep« 17 Cr, 30Cr6 55 Cr4(Br) /M42 LBr \M44Cr6 M42 LBr AA600 LBr-Crt M357 S4CrtAg B141 S4Y4Ag D43, D44 Sep3 D45 Sj-Cr6 D95 S3Y4Ag D95, M357 D67 Sj-Crj M384 Sep6 M442 BrCr6AgT 1 DG6 Sr-Cr6 Total	Dil Dil Dil	75 W BW B42 W la-3 Do		4 1		3 2 1 4 3 6 4 1 2 2 2 1 1 32	2 Sep3. 2 Sep« Sep6
Dil	Do		1 8 3 1 3 3		SBCr5AgTb 8Sep6 3Sep4 Sep4 S3Y4Ag, S6Cr6Ag Sep 4 SsCrsAgTb. 2 S3 Y4Ag Sepa-Crs Sep3, S3Cr6Ag Sep5, 2 Sep4 Sep4, Sep6 Sep2
42-12. . . 42-12... 42-12 39-19 42-1 39-23 16a-3 16a-3 166-9 . . .	[l5 W BW Do
Do
20 W BW....
Do
Do
Do		1
20 W BW....
. . Do . .		2
16c-4 39-12. . 3 39-12 406-13	Do
86 W BW Do		3 2 1 33
Do . .
16c. Males CdCa.
No.	9W.	c^Dil.	Int	Dil	RE	W	Remarks.
1 2 3 4 5 6	132 W 12a W 5 W B2 W BWll \V (BWir. \Y 120. 21. J ;. 26 W Total ...	4-toe 4-toe BW	A674 Sep« }		4 3 2 2 3 8 22		1 5 13 19	3 Sepj-Cr^, Sep 3 Sep8-Cr8 2 Sep6 S3-CrB, SsCr6Ag Tb 2 S5Cr6AgTb Sep< 2 S6Y«Ag. 3 S4- Y4, 2 S4, SB
M34 Sep«-Cr4 16c-l Do
BW BW BW 22-	B117S«CriAgTb 39-14 l....Do
/ 13Cr,(Br) Dil

TABLES.

137

Similar Matings from Crosses 19, 27, and 33, and Summary of Cross 16.

No.	Dilute.	W (intense stock) .	Int	Dil	RE	W
16a 166 16c	5 9 9 Dil (CdCr) .... 3 c? & Dil (CdCr)	W BW....		9 3 5 4 9 33 22	10 5	"l 2 32 19
. ..Do
7 9 9 Dil (CdCa) .... 1 c? Dil (CdCa) .... 9 9 Dil (CdCd) .... 9 9 Dil (CdCa) .... d" o71 Dil (CdCa)	. ..Do
. ..Do
W
. . .Do
. . Do .
Total		85	15	60

Table 78.

Cross 17. — Matings of intense from intense stock with albinos from dilute stock or from

two dilute parents.

Expectation: CC X CaCa = CCa (all Int).

CCa X CaCa = CCa + CaCa (1 Int : 1 W).

17a. MaleCC.
No.	9 W.	cflnt.	Int	Dil	RE	W	Remarks
1 2 3	M117 W 42-11 M327 W 42-17	3013 B BB	2 2 2				2B 2B 2B
Do
D 37 W Dil	. . Do . .
Total
6
176. Female CC.
No.	9 Int.	c?W.	Int	Dil	RE	W	Remarks.
1 2 3 4 5 7 8 9	22, 23, 33 Ag Misc.	11 W Dil Do	11 6 4 4 4 2 2 1				11 Ag 6B Ag,3B 4 Br 4B 2B 2B AgTb
C18 C50 Ag Misc..
(C24 Ag Misc	|....Do . . Do
\C34 B Misc
22 Br Misc .
(S22B 5J«	JM313W 42-16 . . Do
1A1665B 2ig
3223 B Misc.
B232 B ld-21	M201 W 42-13 . . Do
B23 AgTb ld-12
Total
34
		17c. MaleCCa.
No.	9W.	9 Int.	Int	Dil	RE	W	Remarks.
1 2	D37 W Dil	06 B BW	2 1				2B B
9 W Dil	Do			2
Total
3			2

138

INHERITANCE IN GUINEA-PIGS.

Taiile 78 — Continued.

17d. Female CCa.
	9 Int.	dHY.	Int	Dil	RE	W	Remarks.
1 3 3d •l 5 0 7 B	DlsAg 176-1 1 ) I'l \l' 1 T/i — 1	11 W Misc				2 2 1 2 1	Br. R Ag,2B 2 AgTb 2 AgTb AgTb, 2 B AgTb
.Do ....	1 1 3 2 o ml
i \„ 176-1	.Do
1 I >22fl \l' \7b-l	.Do
B33 A«Tb ld-18 B52 WTb la-3	M201 W 42-13 .Do
HI 10 \cTb la-1	. Do .			1 1
Hill \t:Tb la-1	.Do	3 1
Hi's \<'Tb la-1	.Do
Total	13			10

Summary of Cross 17.

No.	Intense (intense stock).	White (dilute stock).	Int	Dil	RE	W
17a 176 17c 17,/	cTVIntCC.... 9 9lntCC... dVInt CCa. •	W	6 34 3 13			2 10
Do Do
9 9 Int CCa • •	Do
Total	56			12

Table 79. ^

Cross 18. — Intense guinea-pigs, each of which had a dilute parent known to transmit albinism, mated with albinos or red-eyes to test whether the same intense animal can transmit both dilution and albinism.

Expectation: CCd X CraCra = CCra + CdCra (1 Int : 1 Dil).

CCa X CraCra = CCra + CaCra (1 Int : 1 RE or W).

18a. Male CCd by test.

No.

9 Red-eye.

d"Int.

Int	Dil	RE	W
3 1 3	3
3
4 3	3 8
3	3
2 3	1 2
1	7
2 25
30

Remarks.

515SAg(R) ArF,.

774SAg(R) ArF,..

515, 774SAg(R) ArF2.

D 10 R (Br) 35-1 . .

Do

Do

10

•".11 Sep (R) 281 SAg (R)

7 1 1 SAg (R^

•_• il BAg i: 113 -

7.V..SA ■ i:

ArF2. ArF2.

ArF2.

\.l \H

ArF,.

Do

D 30 R (Br) 36-1 . . .

Do

AA50S AgTb :W-2.. Do

M4308Ag(R) 1-. i Total, •* malt

Do Do

3Ag,SCrAg,2S3Y4Ag Ag

3 Ag,2 S2Y4Ag, S4Cr6

Ag

4 B.2 S2-Cr6, S4-Y4

3 Ag,S3Y4Ag, S4Y4Ag S6Cr6Ag, 2 S.Y^g 2 S4Cr5Ag, S6Cr6 Ag

3 Ag,2 S3Y4Ag, S6Y4 Ag

2 B,S3Cr6Ag

3 Ag,BiY4AgTb,SsCr6

Ag Ag.SaY^g, S,Y4Ag

4 S4CrsAg, S6Cr6

Ag 2B

TABLES.

139

Table 79 — Continued.

186. Female CCd by test.
No.	9 Int.	cf Red-eye or \V.	Int	Dil	RE	W	Remarks.
1 2 3	M292 B ii . . . . M353 B ei- ■ ■■	D 18 W 166-3 . . Do	2 1 2 5	3 1 1 5			AgTb, B,S6Cr6AgTb, 2Sep6 AgTb.SepB (Cross 20)
D6a R (Br) 35-1 . . Total, 3 females. . .	724 SAg (R) lea
18c. Male CCa by test.
No.	9 Red-eye or W.	dTnt.	Int	Dil	RE	W	Remarks.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15	271 SAg (R) ArF2.. 278 SAg (R) ArF2..	M224 BrAg 406-12 . Do	5 2 4 5		3 5 3 2	1 3 1 1 1 2 3	5Ag,2S4Ag(R),S7Ag (R) 2 Ag,5SAg(R) 4Ag,SsAg(R),SAg(R) S4AgTb (R) 2 Ag, 3 B,2 SAg (R) AgTb.SsAgTb (R) Sep8 (R) 2 AgTb,Sep4 (R) 4B,Sep6(R),2Sep(R) 2 Ag,SePl (R) Ag, B 2 Br, Ag 5 B, Ag 11 B 2 Ag, 8 B
716 SAg (R) ArF2	. .Do
233 SAg (R) ArF2.. 773 SAg (R) ArF2 .	M156 R (B) TV ... Do
236 Sep (R) ArF2.. 264 Sep (R) ArF2. .	AA433a AgTb 36-4 . . . Do	1 2 4 2 2		2 1 3 1
485 Sep (R) ArF2.. 235 Sep (R) ArF2.. 693 W ArF2 . .	D7 R (Br) 35-1 . . . D13 R (Br) 35-1 . . . Do
D42 W 166-1 . AA578 W 3c-18 . 3 9 9 W ArF2.. 4 99 W Misc. . 3 9 9 W ArF2.. Total, 8 males. .	Do
Do	3 6 11 10 57
M291 B & M339 B 40a-14 . M2B &		20	9 13 7 41

1 10

INHERITANCE IN GUINEA-PIGS.

Table 80. Cross {£.— Matioga <>f dilute from cross 18a or 43, with albino.

Expectation: C& X CaCa = CdCa + CrCa (1 Dil: 1 RE) (1-6). CdCa X CaCa = CdCa + CaCa (1 Dil : 1 W) (7-13).

10

11

12 13

14

9 Dil lor \\

nT-1 BtY«Ag L8a 5 D63S4Y4A4 13-2.. D121 W L8e 9.

[67 W

HI is w I \Y

D61 S»Y«A« D241 BeOiAg

D69 S4Cr>Ag M425 SYAg B772 W 266 W S781 W D69 S4Cr6Ag D20G Sr-Yi

22 ■'• L« 8. . 18c-6.

43-2 . .

ls(1 '.1

18a-5.

43-1 . .

ArF2..

ArF2..

ArFi..

18a-5.

18o-4.

d*W (or Dil).

BW36W BW

Do

D71 SjCrsAg lsa-5.

Do

D240 S2Y4AK 18a-9 . Do

BW:iii w BW. BW50W BW.

BW36 W BW.

D70S4Y4Ag 18a-5.

Do

Do

BW36W BW.. BW50W BW..

2 females CdCr .

2 males CdCr. .

3 females CdCa .

1 male CdCa . . .

2 females (?) . . .

Int

I)i

RE

W

Remarks.

S4-Y4,S3Ag(R),Sep,(R) Sep4, 2S3Ag(R),Sep4(R) SjCrjAg, S4-Cr6, S6Ag(R)

Sep, (R) SsAgtR), Sep8(R) S4Cr6Ag, Sep4 2 SBCr6Ag, S6-Cr6. S4Ag

(R)

SeY.Ag

2 S4CrBAg, Sep4

S4Cr6Ag

3 S4Cr5Ag SsY4Ag, S4-Cr6 3 S.Y.Ag Sepfi

Table 81.

Cross 20. — Matings of pure lea male 724 CrCr.

Kxpectation: CC X CrCr = CQ. (all Int) (1).

CCd X CrCr = CQ + CdCr (1 Int : 1 Dil) (2-3).

CdCd X CrCr = CdCr (all Dil) (4-5).

CdCa X CrCV - CdCr + CrCa (1 Dil : 1 RE) (6).

No.	9 Misc.	cf Red-eye lea.	Int	Dil	RE	W	Remarks.
1 2 3 4 r, 6	5 9 9 B Trior4-toe D6a R(Br) 35-1	724SAg(R)Ica... Do	9 2 1				9Ag 2 Ag, S2Y4Ag Ag 2S2Y4Ag 3 SaY4Ag B^r^Ag, SeAg(R)
1
D209R(Br) 36-3	Do .
AA6068jY»AgTb 40a-8. .	. .Do .	2 3 1	1
AA62 1 39-4	Do
SA61 Sep4 32-2	Do . .

Table 82. Cross 21. — Matings of pure lea male 575 CCr. 1 xpectation: CCr X CaCa - CCa + CrCa (1 Int : 1 RE).

No.	9U.	eflnt lea.	Int	Dil	RE	w	Remarks.
l	5 9 9 W BW . . .	575 An Ioa . .	9		4		4 Ag, 5 B, 2 S«Ag(R), 2 Sep4(R)

TABLES.

141

Table 83. Cross 22. — Matings of intense Fi lea (cross 21) with albinos. Expectation: CCa X CaCa = CCa + CaCa (1 Int : 1 W).

No.	9 W (or Int. Fi lea.	cflnt Fi lea (or W).	Int	Dil	RE	W	Remarks.
1 2 3 4 5 6 7	2 9 9 W BW 141 W ArF2....	13 Ag 21-1 . . . Do	4			3 4 4 2 2 4 6	4B 2B 2B Ag, 3B 4B
M79 W g1* D17W 166-3...	14 B 21-1 . . . Do	2
19 Ag 21-1 .... IllAg 21-1	cT W BW.... Do	2 4 4
17 B 21-1....	Do
Total
16			25

Table 84. Cross 23.— Matings of red-eye Fi lea (cross 21) with albinos of intense stock. Expectation: CrCa X CaCa = CrCa + CaCa (1 RE : 1 W).

No.	9 White.	d* Red-eye Fx lea.	Int	Dil	RE	W	Remarks.
1	8W BW....	I5Sep4(R) 21-1			3		3 Sep3(R)

Table 85. Cross 24- — F2 from red-eye Fi lea (cross 21). Expectation: CrCa X CrCa = CrCr + 2 CrCa + CaCa (3 RE : 1 W).

No.	9 Red-eye Fi lea.	d"Red-eye Fi lea.	Int	Dil	RE	W	Remarks.
1 2	12 S4Ag(R) 21-1 . . 16 S4Ag(R) 21-1	15 Sep4(R) 21-1 . .			5 12	5 1	(S2,S3,S6)Ag(R),S4(R),S5 (R) (B0,3S2,S3,S6)Ag(R);(B0l S2, 2 04, S5, 06) (R)
Do

Table 86.

Cross 26. — Matings of F2 lea (cross 24) with albinos.

Expectation: CrCr X CaCa = CrCa (all RE) (1-7).

CrCa X CaCa = CrCa + CaCa (1 RE : 1 W) (8-9).

No. 1 2 3 3a 4 5 6 7 8 9	9 White.	c? Red-eye F2 lea.	Int	Dil	RE	W	Remarks.
D86W 43-3.... M431 W 18c-4. .	126 BoWAg(R) 24-2 . .			5 6 5 4 5 2 1 2 3	2 1	S4Ag(R),S6Ag(R),S3 (R),2S6(R) 3S4Ag(R),2S6Ag(R), Sep4(R) 2 S5(R), 3 S«(R) S3, S4, Ss, S6 3 S6(R), 2 S6(R) S«Ag(R), S6(R) S6(R) S6AgTb(R),S6(R) 3 S4(R)
Do
D76, D78 W 18c-14 . . D76 W 18c-14 D77 W 18c-14. .	137 B0(R) 24-2 . .
. . Do
... Do
D125W 16c-^... BW48 W BW .	133 S2Ag(R) 24-1 . .
. . Do
D427 W 44-3	. .Do
D73 W 42-6 S755 W ArF9	134 Sep4(R) 24-1 .
. . Do

L42

INHERITANCE IN GUINEA-PIGS.

Table 87. Cross 26. — Matings of red-eye F, lea (cross 21) with dilute.

Expectation: CdCd X CrCa = CdCr + CdCa (aU Dil).

CdCtt X CrCa = CdCr X CdCa + CrCa + CaCa (2 Dil : 1 RE : 1 W).

No. l 4	9 Dil (or Red-eye l'i Eos.	cTRed-eyeFiIca (or Dil).	Int	Dil	RE	W	Remarks.
\ \ju-A A/H. 40a-6.. /AA244Sep4 39-15.. \M261 S(-p4 41-2... I2S«Ag(R) 21-1... I6S1WR) 21-1	15 Sep4(R) 21-1 . . . \ ... Do		3 8 1 2	4	5	BiCr.AgTb, Bi- Crs, Sept f5 Bi. S«, S6-Cr6, S8, S,(R),2S4(R), I S,(R) S3 S7Cr5Ag, Sr-Crj
M34 SeP6-Cr6 16c-l . . . . Do

Table 88.

Cross 27. — Matings of dilute from cross 26 with albinos.

Expectation: CdCr X CaCa = CdCa + CrCa (1 Dil : 1 RE) (1-6). CdCa X CaCa = CdCa + CaCa (1 Dil : 1 W) (8-12).

N...	9W (or Dil).	cfDil (or W).	Int	Dil	RE	W	Remarks.
1 2 3 4 5 6 7 8 9 10 11 12	3 9 9 W Misc.. . D221 W ,«t G30 W St ... .	DS9 Sep3 26-3 D196Bi 26-2 Do		4 1 1	6 3 3 2	1 3 2 2 1	2 S«. S8-Y4. Ss-Crs, 2 S,(R), Ss(R), 2 Sr(R). Ss(R) s, S5, S4(R), Ss(R), S7 (R) 3 S,(R) 3 S4, 2 S8(R) 54, S6-Cr6AgTb 55, 2 S5— \ 4 S5, Sc, S4, Se— Cr« 2 S4, 2 S4-Y4 S4
D115Bt-Crf 26-1... D197B, 26-2... D19SB, 26-2... BW4:; \\ BW.... 482 W ArF2... D42 W 166-1 . . BW56. 57 BW.. D1223rCrsAg 26-4... Dl95Sep4 26-2...	BW50 BW . .
\bW46 BW D113BiCr6AgTb 26-1 D123 Si-Cr5 26-4 D55 S5-Cr5 20-9, D114Sep3 26-1 BW50 BW . .		3 2 3 4 4
Do		1

Table 89. Cross 28. — Matings of pure Arequipa male 1007 CdCd with black guinea-pigs.

Expectation: CC x CdCd = CCd (all Int).

CCa X CdCd = CCd + CdCa (1 Int : 1 Dil).

No.

1 3

9 In'

29 9B Woe 4 9 9 B BW 1442 B BW

o" Dilute (Arequipa).

l1007SYAg

.... Do . Do . . .

Int	Dil	RE	W
4 10 3
2

Remarks.

3Ag,B 5 Ag, 5 B

2 Ag, B, SCrAg, Sep3

'1007 8YAg< dCdfron 1001 BRAgCCdand 1002 SCrAg CdCr pure Arequipa stock.

TABLES.

Table 90. Cross 29.— Matings of intense Fi (cross 28) with d"1007. Expectation: CCd X CdCd = CCd + CdCd (1 Int : 1 Dil).

143

No.	9 Intense ArFi.	d* Dilute ( Arequipa) .	Int	Dil	RE	W	Remarks.
1 2 3	SA4 Ag 28-3 SA8 Ag 28-2	4007 SYAg Do	3 2 3				2Ag,R Ag, R, Bi-Y3( Y3 2 Ag. B, BiYsAg, Y,
2 2
/SA4, SA8	l....Do
\SA10 B 28-2
/

!1007 SYAg CdCd from 1001 BRAg CCd and 1002 SCrAg CdCr pure Arequipa stock.

Table 91.

Cross 80.— Mating of dilute Fi (cross 28) with <? 1007. Expectation: CdCa X CdCd + CdCd + CdCa (all Dil).

No.	9 Dilute ArFi.	9 Intense (Arequipa).	Int	Dil	RE	W	Remarks.
1	SA3 Sep3 28-3	'1007 SYAg		2			BxYaAg. Y,

4007 SYAg CdCd from 1001 BRAg CCd and 1002 SCrAg CdCr pure Arequipa stock.

Table 92. Cross 81. — F2 from intense Fi Arequipa (cross 28). Expectation: CCd X CCd = CC + 2 CCd + CdCd (3 Int : 1 Dil).

No.	9 Intense ArFi.	(^Intense ArFi.	Int	DU	RE	W	Remarks.
1 2 3 4 5	(SA6B 28-2 \SA11B 28-2 SA13B 28-3 (SA8Ag 28-2 \SA10 B 28-2 SA8 Ag 28-2	JSA2 B 28-2 SA7 Ag 28-2 \ .Do	17 1 3 1 1	3			17 B, 2 B-Y3, B2-Y4 B Ag, 2B Ag, B2Y3Ag, Br-Y3 Ag, 2 BiYaAg
.Do	2 2
SA4 Ag 28-3	Do
Total . .
23	7

Table 93. Cross 32. — F2 from dilute Fi X intense Fi (cross 28). Expectation: CdCa X CCd = CCd + CCa + CdCd + CdCa (2 Int : 2 Dil).

No.	9 Dil ArFi.	cflnt ArFi.	Int	Dil	RE	W	Remarks.
1 2	SA3 SeDM 28-3	SA7 Ag 28-2 SA12 Ag 28-3	1 1	1 1			B, B2Y3Ag B, S4
Do

1 II

INHERITANCE IN GUINEA-PIGS.

Table 94. Cross 33. — Milting of dilute Fi Arequipa with albino.

l \j>. station: CdCa X CaCa =	CdCa+CaCa(l Dil: 1 W).
l	9 Dil ArF,.	c? White.	Int	Dil	RE	W	Remarks.
HAS Smh 28-3	75 W BW				1

Table 95. Cross 34. — Matings of intense Fi Arequipa with albinos. Expectation: CCd X CaCa = CCa + CdCa (1 Int : 1 Dil).

No. 1 2 3 4 5 6	9 Int ArF! (or W).	cfW(orlntArFi).	Int	Dil	RE	W	Remarks.
SA4Ag 28-3.... SA10, 11, 13 B 28-2,3. .	M313 W 42-16 Do	4 3 3 1 2	2 8 2 1 1 6			2Y4 3 B, R, 2 S3, 4 S3-O6. 2 Y« 2 Ag, B, S3Y4Ag, S3Cr6Ag 2 Ag, B, S* Ag, S3 Ag,B,S3Cr6Ag,2S6Y4Ag, 2 S6Cr5Ag, Sj-Crs
149 VI 22-2 161 W 24-1. . . .	SA26 Ag28-1.. . Do
349 W ArF2	Do
149, 349 W	Do
Total .	13	20

Table 96.

Cross 35. — Mating of cream of dilute selection stock with a red stock free from albinism or

dilution.

Expectation: CC X CdCa = CCd + CCa (aU Int).

No.	9 Intense.	d1 Dilute.	Int	DU-	RE	W	Remarks.
1	4 9 9R(Bi)Misc	00 Cr6(Br) Dil	12				11 R(Br), Y2(Br)

Table 97.

Cross 36. — Fi (cross 35) mated with father.

Expectation: CCd X CdCa = CCd + CCa + CdCd + CdCa (2 Int : 2 Dil) (3).

CCa X CdCa = CCd + CCa + CdCa + CaCa (2 Int : 1 Dil : 1 W) (1-2).

No. 1 a a	9 Intense Fi.	<? Dilute.	Int	Dil	RE	W	Remarks.
I'l R Br) 36 1.. . . D2 1: Br) 36 1 D6R(Br) 35-1 IM 36-1	00 Cr„(Br) Dil . . . Do	4	6 1 4 1		2 1 •	4 R(Br), 2 Y4(Br), 4 Cr6(Br) Cr6(Br) 3R(Br),Y2(Br), 3 Cr5(Br) 2 R(Br), Cr6(Br)
Do Do	3 2

TABLES.

145

Table 98. Cross 87. — Matings of dilute with dilute in the dilute-selection stock.

Expectation: CdCd X CdCd = CdCd (all Dil).

CdCd X CdCa = CdCd + CdCa (all Dil) (1).

CdCd X CdCa = CdCd + 2 CdCa + CaCa (3 Dil : 1 W) (2-11).

No.	9 Dilute.	c? Dilute.	Int	Dil	RE	W	Remarks.
1 2 3 4 5 6 7 8 9 10 11	D301 Y4 Dil ... . Do D300 Cr6 Dil	D292 Y4 Dil D298 Cr6 Dil Do		3 2 3 5 6			Y3) 2 CrB Y3, Cr6 Ys, 2 Cr5 Y4, 3 Cr6, Cr6 Y3, 4 Cr6, Cr6 Y3(Br),3Cr6(Br) Y3(Br),2Cr6Br Y3(Br), Cr5(Br), Cr6(Br) Y3(Br).3Cr6(Br) Y3(Br),Cr6(Br)
	1
D299 Cr6 Dil	. Do. .
D289Cr6 Dil.... D291 Cr6 Dil	D290 Cr6 Dil Do			1 2 1 2 2
(D260Y3(Br) Dil.... \D262, D263 Cr6(Br) DU. . . . D263 Cr6(Br) Dil	|d261 Cr6(Br) Dil Do		4 3 3 4 2
D262 Cr6(Br) Dil	Do . .
/D265 Cr6(Br) 37-7 . . \D266Cr6(Br) 37-7.. D265 37-7 . .	1 Do . . .
/ Do			1
Total (excluding 1)
32		10

Table 99. Cross 88. — Matings of dilute with albino in the dilute-selection stock.

Expectation: CdCd X CdCa = CdCa (all Dil).

CdCa X CaCa = CdCa + CaCa (1 Dil : 1 W).

No.	9 White (or Dilute).	c? Dilute (or W).	Int	Dil	RE	W	Remarks.
1 2 3 1 2 3 4 5 6	38a. D293 W Dil D294 W Dil	D292 Y4 Dil		4 2 2			4Cr6 2Crs 2 Cr6(Br) 2Cr5 Cr5 2Cr6
Do
D264 Y3(Br) 37-7 2 CdCd	11 W Dil .
	8
386. D293 W Dil D302 W Dil D303 W Dil D 75 W Dil Do /D276 W DU \D272 W Dil 5 CdCa ...	D290 Cr« Dil
	2 1 2
D298 CrB Dil
Do
D267 Crfi(Br) 37-7			4 1 2
D274 Cr6(Br) Dil		2 3		2 Cr6(Br) 3 Cr6(Br)
>D261 Cr6(Br) Dil
	10		7

L46

INHERITANCE IN GUINEA-PIGS.

Table 100.

—All matmgH of intense with intense which have given dilute young, except thcwe

given in cross 31.

ftmeeUtion: CCd X CCd = CC + 2 CCd + CdCd (3 Int: 1 Dil) (1-7).

CCd X CCft = CC + CCd + CCa + CdCa (3 Int : 1 Dil) (8-33). 3, 9, 11, 24, 26, 30, and 32 not wholly certain.

No

l

a

3 4

5

6

7 8 9

10

11 12

13

14 15

16 17

18 19

20 21 22 23

_'l

26 27 28 29 30 31 32 33

9 Int.

M35

ATM i AgTb

B68 AnTI> M25 AgLb

3 A

ld-15.

9-1.

AA>s An 39-4..

M25,M27aAg9-l. . . \B139Ag 39-23.

M177B lc-2..

M168B fV---

M169, M171B &...,

B68 AgTb

A443 B M90 Br M90 Br M114B A1117B A1566 AgTb

la-1

16-7

Do.

A529 BrAgTb ft.

\A_'02AgTb 406-8. M177B lc-2..

Ml 02 AgLb 3392 AgLb 3392, 3444 Ag 3392 Ag

A1310Ag M203 AgTb M82 A«

AA171 H Ml S3 B 20 B M7 B Al 120 !'. ". B

66-1 . Misc Misc Misc Misc

*v

2-19. 9-7. .

A

iV;..

Mi.-i-

rV

it-

d" Intense.

SALT, Ag 28-1,

A 78 1 AgTb &..

B 118 AgTb ld-6.

M91AgLb 8-4..

.... Do

> Do

Do Do Do

B 118 AgTb ld-6..

A469 AgTb \

M 189 AgTb 39-30.

Do

1357 B BAY .

A 1050 AgTb 3V . .

A A 15 AgTb Jj. .... Do

.... Do

AA235 AgTb 406-7 .

M2B A1539B .... Do ... . B5 AgTb A 1474 AgTb A1449 AgTb AA284 AgTb A 1161 AgTb

Do

AA299 AgTb A412 R(Br) M133 Ag A811 Br 12845 B

lrf-16. 39-18.

40-6.

8-4. . 4- toe.

Int

109

Dil

l47

RE

\Y

Remarks.

Ag, B, R, Si-Yj

4 AgTb, 3 R, BrYAg Tb, Y

5 AgTb, 3 BrAgTb, B, SsYaAgTh, S,

3 Ag, 3 AgTb, 2 SYAg Lb, 2 S,Y2AgLb

2 Ag, S3Y^gTb

Ag, 2 AgTb, B,YiAg Tb

Ag, 3 AgTb, SYAgTb

2 Ag, 2 AgTb, SCr6Ag

3 AgTb, 2 BrCrAgTb,

BrYAg

4 AgTb. 2 B, SjCrs AgTb

LBr

BrCrsAgTb, S5, Crj

3 AgTb, 3 B, S5Cr5

AgTb, S6 AgTb, B, S4Cr6AgTb 3 AgTb, BrAgTb, SCr

AgTb, S4 AgTb, B, SCrAgTb 3 AgTb, B, BrAgTb,

SCrAgTb, BrCrAg

Tb 2 AgTb, 2 SCrAgTb AgTb, 2 B, BrAgTb,

LBr-Cr5 2 Ag, 2 B, S6Cr5Ag SCrAg

2 Ag, B, SCrAg

3 Ag, S4Y4AB 3 Ag, SCrAg AgTb, B, Cr 3 AgTb, B, S AgTb, SCrAgTb AgTb, R, Cr

3 AgTb, SCrAgTb 2 AgTb, B, SCrAgTb 2 Ag, 2 B, SCrAg. S« Br, LBr 4B, S

ISxeeM of dilutes expected because the presence of at least one dilute young is used as <»

criterion f'<r admission to tin; talile.

TABLES.

147

Table 101.

Cross 40. — All matings of intense with dilute which have given dilute or albino young, except those of crosses 28, 29, 32 and 36.

Expectation: CCd X CdCd = CCd + CdCd (1 Int : 1 Dil), 1.

CCa X CdCd = CCd + CdCa (Int : 1 Dil), 2-5 (5?). CCd X CdCa = CCd + CCa + CdCd+CdCa (1 Int : 1 Dil), 6-19 (9, 10, 12, 14, 15, 16, 17?).

No.	9 Int (or Dil) .	cfDil (or Int).	Int	Dil	RE	W	Remarks.
	40a.
1	B139 Ag	39-23 . .	M328 B2-Y4	42-17.	2	3			2 AgTb, 3 S2Y2 Ag
2	AA606 S2Y*AgTh	40a-8 . .	AA573 BrAgTb	40a-7.	1	1			AgTb, S,Cr6Ag Tb
3	4 9 9 B	Misc. . .	AA241 SYAgTb	40a- 6.	5	3			AgTb, 4 B, S„ Cr&AgTb. Se- Cr6, LBr
4	169 B	4-toe . . .	A656 Br-Y	1 s	1	3			B, 3 S-Cr
5	AA497 SYAgTb	39-7.. .	AA284 AgTb	39-18.	2	1			AgTb, R, LBr
6	AA206 AgTb	39-2. ..	AA177 S6Cr6AgTb	41-4. .	3	4			3 AgTb, SYAg Tb, 2 S3Y3Ag Tb, SgCrsAg Tb
7	AA213 AgTb	39-15. .	AA253 S5Cr6AgTb 4	4	5			3 AgTb, BrAg
									Tb, S2Y2AgTb,
									BrYaAgTb, S6
									OfAgTb, S6
									CrfAgTb, Br
									Cr6AgTb
8	AA217 AgTb	iOb-S . .	. . Do		7	4			7 AgTb, SjYiApc
							Tb, 2 SsYgAg
									Tb, SCrAgTb
9	AA613 AgTb	40a-7	.Do			3			SaYsAgTb, SjCrj
							AgTb, S4Cr6
									AgTb
10	3a AgLb	Misc. . .	M34 Sep6-Cr6	16c-l .		2			2 Ss-Cr6
11	M261 Sep*	41-2...	AA299 AgTb	40a-6 .		1			S3
12	M282 Ag	15-12..	M116Sep5-Y4	42-11.	3	1			2 AgLb, B, S6- Cr6
13	30 Br	Misc. . .	M34 Seps-Cr6	16c-l .	1	2			R, B2, S5
14 15 16 17 18	MQQ R	42-13?	Do		2	1			2B,Ss-Cr6
Ml 01 R MQQ B	42-13?	Do .	3	1			B, 2 R, S7
M99B M155B M82 Ag	42 13?	A674 Sep6	i		1			S7-Cr7
A			3 3			2 S,t S7 2 AgTb, SY3Ag
9-7 ... .	M116Sep6-Y4	42-11.	2
									Tb, BrYsAg
									Tb, SCr6AgTb
19	SA13B	28-3. ..	M306 S;-Cr7	42-15.	36	2 441			S5— Crs, S7— Cr?

Excess of dilutes expected.

1 is

INHERITANCE IN GUINEA-PIGS.

Table 101 — Continued. ( ross 40 — Continued.

No 1 3 4 6 6 7 9 10 11 12 13	911.1 Coi mi).	•Dil inr hit).	Int	Dil	RE	w	Remarks.
406.1 U660B ?J« A1523AgTb & B132 AgTb Id 3... M442BrCr»AgTb 39-12.. Mil <r6 12 12 M181 BrCr»AgTb 41-6. . . AA203BrCrAgTb 39-17. . L1273 SCrAsTb «*»	AA253 S(CrtAgTb 406-8. AA199 SCrAgTb 39-15. M293Yi 42-14. AA573 BrAgTb 40a-7. AA197AgTb 2-10.. AA573 BrAgTb 40a-7 . A A 16 AgTb A--- . . Do	2 1 2 3 3 5	2 1 1 1 1 2 2 3 1		2 1 3 1 1 1 1 2 122	2 SCrAgTb 2 AgTb, SCrAg Tb AgTb, S6Cr«Ag Tb BrCr5AgTb 2 AgTb, SCrAg Tb 3 BrAgTb, Br Cr6AgTb, Br Cr6AgTb 2 AgTb, SCrAg Tb. LBr AgTb, 2 B, SCr AgTb, S5Crs AgTb, S8-Y4 4 AgTb, B SCr AgTb Ag, BrAg 4 B. Ss-Crs
AA 176 AgTb 41-4... V \17."> \irTb 41-4	AA177S6Cr6AgTb 41-4.. .Do
AA671 AgTb 40a-7.. Mil! BrC'r«Ag 44-1... Sep ArF2... Total	AA253 S6Cr,AgTb 406-8. Al 170 AgTb sV-..- M156R iV
2 4 24	1 152

•Expectation: CCa X CdCa = CCd + CCa + CdCa+ CaCa (2 int : 1 dil : 1W). 'Excess of dilutes and albinos expected.

Table 102.

Cross 41. — All mating8 of intense with albino which have given dilute young, except those

given in crosses 18 and 34.

Expectation: CCd X CaCa = CCa + CdCa (1 Int : 1 Dil).

No.	9 Int (or \Y).	c?W (or Int).	Int	Dil	RE	W	Remarks.
1 2 3 4 5 6 7	All u. AgTb ,V . . . M102 AgLb 8&-1. . B13' 39-23. A1227W «V • . ■ Al.JD'.tW j'j.... LA28W J*....	A504 W ,'fl A462 W ,'« . . . 20 w b\y . A781 AgTb £t \l.--i:> AuTb .,'... . Do	1 1 2 4 4 1 13	2 1 1 1 1 4 1 ll1			S, S-Y T> Q AgTb, S3Cr8AgTb 2 AgTb, S6Cr6 AgTb BrAgTb, B. R, R(Br), Cr(Br) Ag, BrAg.AgTb. R, SCrAgTb, BrCr6AgTb, 2 Cr B, S-Cr
131 W 4- Total	A412R(Br) fc. • ■

'Excess "f dilutes expected though not found.

TABLES.

149

Table 103. Cross 42. — All matings of dilute with dilute, except those of crosses 30 and 37. Expectation: CdCd X CdCa = CdCd + CdCa (all Dil) (M328, AA242, M394, CdCd). CdCa X CdCa - CdCd + 2 CdCa + CaCa (3 Dil : 1 W).

No,

3

4

5

5a

6

8

9

10

11 12 13

14 15 16 17 18 19 20

21

22 23 24

25

26

27

9 Dilute.

B141 S4Y4Ag 39-23.

AA242 S3Y3AgTb 40a-6.

AA244 Sep4 39-15 .

D44 Sep3 16a-3 .

D43 Sep3 16a-3 .

D45 Sep3-Cr6 16a-3 .

D26 S5Cr5AgTb 16a-4 .

/DllOSepB 16c-3.

\D107 Sep4 166-1 .

D215Sep3 16a-2.

M181 BrCr6AgTb 41-6. .

3520 Y4(Br) Dil . . .

3417 Cr6 Dil . .

/3417 Cr6 Dil . .

\3462 Cr6(Br) Dil . .

06 LBr-Cr Misc.

M127Cr5 42-13.

M44Cr5 42-12.

M164Cr7 44-6..

M126Cr6(Br) 42-13.

M164Cr7 44-6..

M44 Cr6 42-12 .

M296 SCrAgTb 39-27.

M310 Sep7-Cr7 40a-16.

M336Sep6 40a-17.

M336 Sep6

M394Sep4 42-22..

/M336, Sep6 40a-17 .

\M394Sep4 42-22..

M393 Sep5 42-22 . .

D278 Cr6(Br) 386-5 . .

<? Dilute.

M328 B2-Y4

42-17.

B117S4Cr5AgTb 39-14.

....Do

Do

D94 S3Y4Ag 166-9 .

....Do

D33S6Cr5AgTb 166-4.

D106 Sep4 166-1 .

Do

AA253 SBCr6AgTb 406-8 . A674 Sep6 I

....Do. ....Do. . . . . Do .

....Do

.... Do

....Do

.... Do

M306 Sep7-Cr7

....Do

M116Sep6-Y4

M335 Sep6 Do

M34 Sep6-Cr5 Do

42-15.

42-11.

40a-17

16c-l

Do.

...Do

D138 Cr5(Br)

Int

19!

36-1.

Total

CdCd X CdCa. CdCa X CdCa.

19.

Dil

RE

60 10 50

W

Remarks.

19

0

19

3 S2Y3AgTb, S«

Cr6Ag S6Cr6AgTb S6Cr6AgTb S4Cr6AgTb S3Cr6Ag, S3 Bt, S3Cr6Ag

5 S3

Bi, S3

SY4AgTb

S3Y3Ag,S6Cr6

Ag, BrCr6Ag,

Y4(Br),2Cr6

(Br) S5-Y4, Y2, Cr6 S-Cr, LBr, Y,

Cr5 2 B, S-Cr, S7-

Cr7,Cr6(Br),

Cr5 Y4 S7-Cr7

B2-Y4 St-Ctb, Cr5 S6-Cr6, Cr7, Cr6 SY2AgTb. 2 S

Y3AgTb Ss

be— Crg, 05, S4 S4-Y3

S4— Y4, Y3, CTq

So, Ss, S4, S3 Cr6(Br)

Recorded from a mixed pen before the study of dilution was begun, probably an error.

Table 104. Cross 43. — All matings of dilute with red-eye, except those of crosses 20 and 26. Expectation: CdCd X CrCa = CdCr + CdCa (all Dil) (1-2).

CdCa X CrCa = CdCr + CdCa + CrCa + CaCa (2 Dil : 1 RE: 1W) (3-4).

No.	9 Red-eye.	d" Dilute.	Int	Dil	RE	W	Remarks.
1 2 3 4	241SAg(R) ArF2.. 271SAg(R) ArF2.. 236Sep(R) ArF2.. D194Sep4(R) 26-2..	M328 B2-Y4 42-17 M333Y2 42-11.. M331 BrCr6Ag 42-10. .		3 4		1 2	3 SCrAg 2 S6Y4Ag, 2 S4Y4Ag S2Cr6AgTb, S6Cr5 AgTb, SjAgTb (R)
AA670 SeOsAgTb 40o-7. .		2	1

L50

IMIKUITANCE IN GUINEA-PIGS.

Table 105.

M j^ — All niatings of dilute with albino, except those of crosses 16, 19, 27, 33, and 38.

Expectation: CdCd X CaCa = CdCa (all Dil) (1).

CdCa X CaCa = C(,Ca X CaCa (1 Dil : 1 W) (2-6).

No. 1 2 3 4 5 6	9 Dil (or W).	tfW (or Dil).	Int	Dil	RE	W	Remarks.
8256BrYAsLb Misc.. A5(t."> Bep i'e • • • ■ 3 9 9 W Misc.. B176B-CI ArF*.. 8263 S( ThXe. ArF2.	A678 W A A868 W jV		9 1 5 2 1 2		1 3	7 SCr&Ag, 2 Br-Cr« S S4, S5Cr6AgTb, Br Cr«AgTh. 2 S, 2 Ss-Y* S4Cr6Ag S7Cr7Ag, Cr7
B117S4Cr6AKTl) 39-14 11 W Dil ... .
. . Do . .
07 W Misc..	A674 Sep« J

Table 106.

Cross 45. — Rough A (4-toe) X rough A (4-toe). Rrss X Rrss = 3 Rss + rrss (3 A : 1 Sm).

No.	9 Rough A.	c? Rough A.	A	B	C	D	E	Sm
1 2 3 4 5	3769 4-toe 96 4-toe	3609 4-toe Do	1
				2 1
/3769 4-toe \3770 4-toe 3769 4-toe 3770 4-toe	| .... Do 3987 4-toe Do . .	5 2 2
Total
10					3

Table 107.

Cross 46. — Rough A (tricolor) X rough A (tricolor) ; one or both of parents of each,

rough C or D. Rrss X Rrss = 3 Rss + rrss (3 A : 1 Sm) (1-8). or Rrss X RRss = Rss (all A) (9-12?).

No.	9 Rough A.	cf Rough A.	A	B	C	D	E	Sm
1 2 3 4 5 6 7 8 9 Id 11 12	4018 Tri 3941 Tri 3943 Tri	3775 Tri						1
3940 Tri Do	3 1 7 2 2 1 1 2 3 2 3
fR66 54-17 3943 Tri R65 54-17	}....Do Do					1 2 2
RlTi 47 :'.....	Do
R278 52 1 1 R357 Bed ">'_' 8	R248 52-10 Do	1 1
			1
R65 54-17 R171 47-3	R197 52-14 Do
Rl'U 64-1	Do
EU96 52 ll	Do
Total 1 to s
17 10	2				<7 4
Total 9 to 12

TABLES.

151

Table 108. Cross 47. —Rough A X rough C (tri); all mothers of tricolor stock except R175-4-toe. Rrss X RrSs = 3 Rss + 3 RSs + 2rr (3 A : 3 C : 2 Sm).

No.	9 Rough A.	d" Rough C.	A	B	C	D	E	Sm
1 2 3 4 5 6 7	R21 46-2 R23 46-2	R52 56-1 Do	1 1 1 1					2 1 1 2
R42 50-1	Do		1
R21 46-2 R23 46-2	R99 56-1 Do
	1 2 1	1
R42 50-1	Do	3 3	1		4
R175 49-1	Do
Total . . .
10	1	5	1		10

Table 109.

Cross 48.— Rough A (Tri) X rough E (Tri). Rrss X RRSS = RSs (all C).

No.	9 Rough A.	cf Rough E.	A	B	C	D	E	Sm
1	R42 50-1	4003 Tri			2	1

Table 110.

Cross 49. — Rough A (4-toe) X smooth (4-toe). Rrss X rrss = Rrss + rrss (1 A : 1 Sm).

No.	9 Smooth.	cf Rough A.	A	B	C	D	E	Sm
1 2	49 9 4-toe 5 9 9 4-toe Total	3922 4-toe.... 3609 4-toe....	18 10					13 19
1
28	1				32

Table 111.

Cross 50. — Rough A, B (tri) X smooth (4-toe etc.). Rough A, B with one or both of

parents paitial rough.

Expectation as in cross 49.

No.	9 Smooth (or rough B).	o" Rough A (or smooth).	A	B	C	D	E	Sm
1 2 3 4	79 9 Sm 4-toe R121 Sm 50-1.	3775 A Tri .Do .	13					14 3 3
R62 Sm 50-1 !R163 B 52-13	R22 A 46-2 99 Sm 4-toe	3 3
Total	19					20

*R163 may be RR.

I 52

INHERITANCE IN GUINEA-PIGS.

Table 112. I St. -Rough A X smooth (tri); smooth with one or both parents partial rough. Rm x n6S = Rifla + rrfla (1 C: 1 Sm).

Krss X rrSs = Rrss + RrSs + 2 rr (1 A : 1 C : 2 Sm). Krss X rrss = Rrss 4- rrss (,1 A : 1 Sm).

	(Smooth.	cf Rough A.	A	B	C	D	E	Sm
1 ■_> a t 6 G	R13 52-1 R123 ."-1-3 IU'l 54—3	R22 46-2			1 2	3		4 4 2
R76 4-toe
. . Do	1 2 2 1
R133 47-2	. . Do
R124 R133	.Do					8 1
R142 54-3	. . Do
Total ,	6		3	3		19

Table 113. Cross 52— Rough1 C, D (tri) X rough C (tri). RrSs X RrSs = 3 Rss 4- 6 RSs + 3 RSS + 4 rr (3 A : 6 C : 3 E : 4 Sm).

No.

l

2

3

4

5

6

7

6

9

10

11

12

13

11

15

9 Rough C. D.

3013

3246

Rll

R54

3245

3939

3809

3246

3724 D

3724, 3246

R51

R57

R61, R57

R98

R103

Tri . Tri . 52-1 52-1 Tri . Tri . Tri . Tri . Tri . Tri . 56-1. 56-1. 56-1 56-1 56-1

cfRough C.

3780 Tri . .

Do...

Do . . .

Do . . .

4019 Tri . .

R58 52-5.

Do...

Do . . .

Do . . .

Do . . .

R56 56-1.

Do . . .

Do . . .

Do . . .

Do . . .

Total.

18

B

19

E

12

Sm

17

Remarks.

Red-A

Red-E, Red-Sm

Red-A

Rcd-C

>A11 rough C except 3724.

Table 114.

Cross 63.— Rough C, D (tri) X rough E (tri).

RrSs X RRSS = RSs + RSS (1 C: 1 E). or RRSsX RrSS = RSs + RSS (1 C : 1 E).

	or RrSa X RrSS =	■ 3 RSs + 3 RSS + 2	rr (3C:3E:2Sm)	•
	9 Rough C, D.	cf Rough E.	A	B	C	D	E	Sm
1 a i	R6D 54-15 R88C 52-1 R286C 62-6 R222 C 52 12. . . .	4003 Tri				1	2 2
R200 Hod 52-7
R280 52-14			3 1
Do				2	1
I •■• -il
		4	1	6	1

TABLES.

153

Table 115. Cross 54- — Rough C, D (tri) X smooth (4-toe, etc.). RrSs X rrss = Rrss + RrSs + 2 rr (1 A : 1 C : 2 Sra).

No.	9 Smooth (or rough C,D).	cfRough CD (or smooth).	A	B	C	D	E	Sm
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17	3 9 9 Sm 4-toe R62 Sm 50-1	3780 C Tri Do	1		1			8 2 17 14 6 9
2
5 9 9 Sm 4-toe	R12 C 52-1 R26 D 54-15 R52 C 56-1 R102 C 56-1 R105 C 48-1 R106 C 48-1 . . Do	4 8 2 4 2 2		12 3 2 5
3 9 9 Sm 4-toe	5
4 9 9 Sm 4-toe . . ,
5 9 9 Sm 4-toe	3
B31 Sm ld-9
M253Sm la-10 M255 Sm la-10 .
	1 1
M380 Sm ^j . .	.Do .
131, M253, M255 Sm . .	.Do .	1				3 2 3 1 3 3 8
M384 Sep-Sm lb-9 . .	R99 C 56-1		2
R62 Sm 50-1	R112 C 52-11..
65 Sm 4-toe	Do	1 5 1 3		1 1
3246 C Tri	2967 Sm BB Do	2
3809 C Tri
3724 D Tri	. . Do . .			1	1
Total
34		29	13	1	79

Table 116. Cross 55 — Rough E (tri) X rough E (tri). RrSS X RrSS = 3 RSS + rrSS (3 E : 1 Sm).

No.	9 Rough E.	c? Rough E.	A	B	C	D	E	Sm
1 2	R221 52-12 1R201 Sm 52-7	R140 52-3 . .					2 2	3
.Do
				4	3

xSee note, cross 57.

Table 117.

Cross 56.— Rough E (tri) X smooth (4-toe).

RRSS X rrss = RrSs (all C) 1.

RrSS X rrss = RrSs + rrSs (1 C : 1 Sm) 2-5.

No.	9 Smooth (or rough E).	c? Rough E (or smooth) .	A	B	C	D	E	Sm
1 2 3 4 5	3 9 9 Sm 4-toe 1 9 Sm 4-toe R^l Sm la 1	4003 E Tri			11 1
R140 E 52-3					3 2 1 2
!R201 Sm 52-7 .... R921 Fi 5*> 12	iq W-Sm 4-toe			1
Tntnl (1}			11 2
Tntnl (9 ^					8

^ee note, cross 57.

l r,[

INHERITANCE IN GUINEA-PIGS.

Table 118.

(Wot* 57. — Smooth (tri) X smooth (4-toe, etc.); both parents of tricolor smooths were partial

roughs (cross 52).

rr X rr = rr (all Sm).

If, however, RSS, normally rough E, ie ever Sm: RrSS X rrss = RrSs + rrSs (1 C : 1 Sm).

No.	9 Smooth.	cfSmooth.	A	B	C	D	E	Sm
1 2 8 4 5 6 7 8	69 9	R131 52-4					14 2 6 2 3 3 2 1
RIBS 52-3 R13 52-1	99 4-toe
... Do
R164 52-13	. . Do
R 199 Red 52-7 R249 52-10	13 W 4-toe
... Do
R263 52-11.	.... Do
'R201 52-7. .	. . Do			1

'R201 was called rough E? at birth with the note that there seemed to be a trace of roughness on ODe hind toe. No roughness was apparent when adult and she was called Sm, but nevertheless was tested by mating with a 4-toe smooth. The result shows that she was, genetically at least, like a rough E.

Table 119. Cross 58. — Rough B, C (Lima) X rough B (Lima).

No.	9 Rough B.	c? Rough B.	A	B	C	D	E	Sm
1 2 3 4 5 6	L7 Lima . . . L97 60-6 L140 60-7	L5 Lima . . . L26 58-1 Do	2 1 1 3	2				1
1 2 1 1 7
(L97 60-6 .... \L81 Red 59-3 L99 Rough C 61-1 . .	|L98 60-6 .... Do				2 4 7
2 2	1 1
LSI, L97, L99 (above) .	Do	1 8
Total

Table 120. Cross 69. — Rough A (Lima) X smooth (Lima).

No.	9 Smooth (or rough A).	cf Rough A (or smooth) .	A	B	C	D	E	Sm	Remarks.
1 2 3 4 5 0 7 0 10	L18 8n 62-2.. I.ltSm 60-1	L9 A 58-1 . . Do	2 7 2 4					1 2 5 3 1 2 1 6 2	Red-B 2 Red-A, 2 Red-Sm Red-A Red-B 2 Sep(p)-Sm 2 Sep(p)-A
1 1 1
I-'lSro 60-2	Do .	3
126 Ba 58-1.	Do . .
L37 Sm 62-:<
L678n 59 :•!	Do	1 1 4 1 2 24	2 2 1
I.-'i. L67 8m (above) I -'_' A 60-2.. I >-.' A 50-3	Do
LI Sm Lima . Do
Ll00Sep(p)-A 61-1. • il	L82 Sep(p)-Sm 59-8 . .
8	3			23

TABLES.

155

Table 121. Cross 60. — Rough B (Lima) X smooth (Lima).

No.	9 Smooth.	d" Rough B.	A	B	C	D	E	Sm	Remarks.
1 2 3 4 5 6 7 8 9 10	L4 Lima . L6 Sep(p) Lima .	L5 Lima . Do	3 2	2				2 1 2 1 3 8 3 1 1 8 30	Red-A. 2 Red-B, Red-Sm Red-Sm
L34 62-2 . . L37 62-3 . . L43 62-1 . .	L26 58-1 Do	2
Do		1 2 2
L34, L43 (above) L41, L43 62-1..	Do
Do
L132 60-3 . .	Do	1 1 1 8
L75 60-4 . .	..Do .
(L14 60-1 . . \ L24 60-2 . . (L25 58-1 . . Total	) I-L131 60-3.. J
9

Table 122. Cross 61. — Rough C (Lima) X smooth (Lima).

No.	9 Rough C.	d" Smooth.	A	B	C	D	E	Sm	Remarks.
1	L56 59-3	LI Lima. . . .	2	1	1			2	Sep(p)-A

Table 123 Cross 62. — Smooth (Lima) X smooth (Lima). Offspring all smooth.

No.

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

9 Smooth.

L3B L2 Red L8 Red L2, L8 Red LI 2 Red

Do ... . L35 Red L17 Red L8 Red L8, L17 Red L64 Red L6 Sep(p) L19 Sep(p) L6 Sep(p) L19 Sep(p)

Do ... . L58 Red(p) L121 Red(p)

Do ... . L120 Red(p)

Lima Lima Lima Lima 62-2.

62-3... 62-3 . . . Lima . . (above) 62-13. . Lima . . 62-3 . . . Lima . . 62-3 . . .

62-3 . . 62-11

62-11

cfSmooth.

62-2. 62-1.

LI B Lima

Do

Do

Do

Lll Red

L40 Red

Do . . .

Do .. .

Do . . .

Do . . .

L59 Red

LI B Do...

L18 Sep(p) Do . . .

L82 Sep(p)

L59 (Red) Do

L149 Red(p) 62-11. Do

62-3. Lima

62-3.

59-8. 62-3.

B

Red

Sep(p)

Red(p)

156

INHERITANCE IN GUINEA-PIGS.

Table 124. Cms* 68. — Rough A, B (Lima) X smooth (4-toe, etc.).

No.	9 Rough A (or smooth).	cTSmooth (or rough A).	A	B	C	D	E	Sm
i a 3 i 5	L107 A 59-9.... I 10S \ 59-7	13W-Sm 4-toe.... . .Do	2 1 1					1
I 110 H 59-7	I),,					1 1 1 4
DIM) \Y Sm 18c-14. . I ) ' Hi \Y Sin 17d-7	L98 B 60-6 .	2 2 4
Do
Total	4

Table 125. Cross 64. — Rough C (low grade due to C. rufescens) X smooth (4-toe, etc.). RrSs X rrss = Rrse + RrSs + 2 rr (1 A : 1 C : 2 Sm).

x...	9 Rough C.	cfSmooth.	A	B	C	D	E	Sm	Remarks.
1 2 3	A606 Ag J A1687 64-1 . . A168SAg 65-1.. Total	166 4-toe...			2				2B-C B-Sm BrAgTb-A, 2 AgTb-A, 4 AgSrn
99 4-toe . .					1 4 5
AA83 g\	3 3
	2

Table 126.

Cross 65. — Smooth (some C. rufescens blood) X rough A.

rrSs X Rrss = Rrss + RrSs + 2rr (1 A : 1 C : 2 Sm). rrss X Rrss = Rrss X rrss (1 A : 1 Sm).

No.	9 Smooth.	cf Rough A.	A	B	C	D	E	Sm
1 2 3 4 5 6 7 8	A702 AgTb & V.05 J	2597 Ag stock Do	1	1	1 1
Mill' i	Do. . .	o 2 2 7 2 2
As42 J	Do. .					3
A913A£Tb fa	Do
0 9 9 AgTb tL-^J*.	Do					12 4 5
B2 la-5 /B240 la-5.... K61, K62 7- 6	R88 52-1 . . \ Do
/

TABLES.

157

Table 127. Cross 66. — Rough A X smooth; both parents with a little C. mfescens blood. Rrss X rrss = Rrss + rrss (1 A : 1 Sm).

No.	9 Rough A.	c? Smooth.	A	B	C	D	E	Sm	Remarks.
1 2 3	A1690 Ag 65-5 . . Do	AA83 B-Sm & . . . M91 Ag-Sm 8-4 . . AA83 B-Sni & . . .	4 2 8	1				4 4 4	3 AgTb-A, Red-A, Ag Tb-B, AgLb-Sm, Ag Tb-Sm, 2 Red-Sm 2 AgLb-A, 2 AgLb-Sm, 2 AgTb-Sm 4 AgLb-A, 4 B-A, 2 Ag Lb-Sm, 2 B-Sm
A1691 Ag 65-5. .
Total . .	14	1				12

Table 128. Cross 67.— Rough A (4-toe, tri) X smooth (pure lea).

Rrss X rrSS = RrSs + rrSs (1 C : 1 Sm). Young all light-bellied agouti.

No.	9 Rough A.	cf Smooth.	A	B	C 3 2	D	E	Sm
1 2 3 4	/R215 49-1 \R252 49-2 .... R236 50-3	>724 SAa(R) lea
.Do					1 1 2 4
R205 46-4
R213 49-1
				5

Table 129. Cross 68.— Rough A X smooth (pure C. cutleri).

Rrss X rrSS = RrSs + rrSs (1 C : 1 Sm). Young all light-bellied agouti.

No.

9 Rough A.

3986 3988

3986, 3988 A1691 Ag AA567 Ag

/AA568

\R80

Total .

4-toe . . 4-toe . . 4-toe . . 65-5. . 66-3 . . 66-3.. 54-15.

cfSmooth.

C128 Ag. .... Do .

Do .

Do .

Do .

Do

B

I)

E

Sm

15

158

INHERITANCE IN GUINEA-PIGS.

Table 130. -Rough C (tricolor) X smooth (pure C. cutlcri). RrSs X nSS = RrSs + RrSS + 2 rr (1 C : 1 E : 2 Sm). Young all light-bellied agouti.

No 1 2 3 4 5 6 7 8 9	9 Black rough < '. D.	cf Smooth.	A	B	C	D	E	Sm 2 1 2 1 1 2 3 12
3245 Tri R". » 5° 1	C128Ag
.Do				1	1 1 2 2 1
1? 1 V 54—4	Do .
•;■' i ri Rl l()	.Do
R 1 10 51 1	Do .
R 1 70 47-3	.Do			1
H 1 1 52-1	Do
R101 D 52-5	.Do .
R154 D 54-4	Do					2 9
			1	1

Table 131. Cross 70.— Rough A (guinea-pig) X rough C, D (|, \ cutleri).

I X RrSs = Rss + RSs + 2rr (3 A : 3 C : 2 Sm) All | cutlcri Rough C, D, except K58, \ blood. R116 and R137 may be RRss.

	9 Rough A (or C, D).	cf Rough D(orA).	A	B	C	D	E	Sm	Remarks.
1 2 3 4 5 6 7 8 9 10	R 116 B-A 46-4.. R 117 B-A 46-4	K54 Ag-D 68-1 . . Do	2		1				Ag-A, B-A, Ag-C B-D, B-Sm Ag-A, 2 B-A, Ag-E 2 B-C, B-Sm 2 Ag-B B-A, Ag-B, Ag-E B-A, B-D Ag-A. 2 Ag-C, B-Sm B-C, B-Sm
1		1
R137B-\ 47-1	Do .			2
A. \608 B-A 66-3	Do	O			1	1
K12Ag-D 68-3.. K14 Ag-D 68-3	R31 B-A 45-3..		2
.Do		2 1
Do	3609 B-A 4-toe . R76B-A 45-4.. Do	1 1 1	2 1 8	1 2	1 2	1 1 4
Do
K12 Ag-D 68-3
£58 B-C 70-5..	. . Do
Total ....		8	3

Table 132.

Cross 71. — Rough A (guinea-pig) X smooth (|, * cutleri).

Rrffl X rrSs = Rrss + RrSs + 2rr (1 A : 1 C : 2 Sm). All \ cutlcri except K79, \ cutleri.

No.	9 Smooth.	d1 Rough A.	A	B	C	D	E	Sm	Remarks.
1 8 I t. 7 8	K7Ag 77-1.. Klfi Ag 68 ■{	R31 B 45-3. . Do	3 3	2	1 2 o			7 9 1 3 1 1	Ag-A, 2 B-A, 2 B-B, B-C, 6 Ag-Sm, B-Sm Ag-A, 2 B-A, Ag-C. B-C, 3 Ag-Sm, 6 B-Sm 2 Ag-C, B-Sm 2 Ag-A, 2 Ag-Sm, B-Sm B-A, B-B, Ag-D, 2 B-C B-Sm Ag-C, Ag-Sm B-A, B-C
K66 \.- 88-1	Do . . .
K7, K.v> (above) K68 \ :' 77 1	Do Do	2 1
1	2	1
Kilo \, BS 8	Do
K81 \j 89 l K7'.i B 7^ 1 Total. . . .	3609 H 4-toe			1 1 9
:•,'.»•_"_• B 4-toe .	1 10	3
1		22

TABLES.

Table 133.

Cross 72.— Smooth (guinea-pig) X rough C, D ($, \ cutleri).

rrss X RrSs = Rrss + RrSs + 2rr (1 A : 1 C : 2 Sm). K71a, K92 may be RR.

159

No.	9 Rough C, D(orSm).	d1 Smooth (or rough C, D).	A 1 3	B	C	D 2	E	Sm	Remarks.
1 2 3 4 5 6 7 8 9 10 11 12 13 14	K12 Ag-D 68-3 . . . K 14 Ag-D 68-3...	00Cr(Br)Sm Dil. ... Do		6 3	Ag-A, 2 B-D 2 Ag-A, B-A, 3 Ag-Sm, 3 B- Sm /B-C. 2 Ag-Sm, 1 B-Sm Ag-A, 2 Ag-C Ag-A, Sep-Sm 2 B-Sm,Sep-Sm Red-A, W-D, B-Sm B-A, Ag-Sm Ag-A, Ag-Sm Ag-A, B-A, Ag- C, 2 B-Sm Ag-Sm Ag-C, B-D /B-A, 3 AgTb- \ C+D,B-Sm B-Sm
/K12 Ag-D 68-3 \A71a Ag-C 70-1 . . . K92Ag-C 70-9...	1 Do		1 2
/ Do
K157Ag-C 72-12.. K147 AgTb-D 72-13 . .	13Cr(Br)SmDil.... Do			1 3 1 1 1 2 1 1 1 21
K142 B-D 72-1 . . . D40Cr(Br)Sm 36-1... R173 B-Sm 49-1 . . .	00Cr(Br)SmDil... . K60 Ag-C 71-2 . . Do	2			1
20 B-Sm 4-toe . . AA533 Ag-Sm 66-2 . . . AA586 Ag-Sm 106-10 .	K54 Ag-D 68-1 . . K56 B-C 70-5 . .		1
Do			1 1	1 2
/M382AgTb-Sm 16-9. . . \B239 AgTb-Sm la-5. . . B239 AgTb-Sm la-5 . . .	Do	1
Dn
Total	12		6	6

Table 134.

Cross 78. — Smooth (4-toe) X rough A, B (| cutleri). rrss X Rrss = Rrss 4- rrss (1 A : 1 Sm).

No.	9 Rough A, B.	d1 Smooth.	A	B	C	D	E	Sm	Remarks.
1 2	K95 B-B 71-1 /K101 B-A 70-8 \K106B-A 71-2 Total . .	99 B 4-toe						1 1 2	B-Sm 2 B-A, B-Sm
|l3 W 4-toe	2 2

Table 135.

Cross 74- — Smooth (f, f cutleri) X rough A (? cutleri).

rrSs X Rrss = Rrss + RrSs + 2rr (1 A : 1 C : 2 Sm). rrss X Rrss = Rrss + rrss (1 A : 1 Sm).

No.	9 Smooth.	cf Rough A.	A	B	C	D	E	Sm	Remarks.
1 2	K68 Ag 77-1 K42 B 78-2	K59 B 71-2			1				B-C B-Sm
.Do					1 1
Total				1

160

INHERITANCE IN GUINEA-PIGS.

Table 136. Cross 75. — Rough A, B (J culleri) X rough A (J cutleri). Rres X Rres = 3 Rss + rres (3 A : 1 Sm).

No.	9 Rough A, B.	d* Rough A.	A	B	C	D	E	Sm	Remarks.
1 2	E60Ag-B 70-6 K53 B 71-1 .	K59B 71-2 Do	1 2 3						B-A 2B-A
Total

Table 137.

Cross 76. — Rough C (f cutleri) X rough C (£ cutleri). RSs X RSs = Rss + 2 RSs + RSS (1 A : 2 C : 1 E : ? Sm ?).

No.	9 Rough C.	cT Rough C.	A	B	C	D	E	Sm	Remarks.
1	K114B 74-1....	K93 Ag 70-9 . .			1	3	1		/Ag-C, Ag-D, 2 B-D. 1 Ag-E

Table 138.

Cross 77. — Black (BB) X agouti (pure cutleri). Parents and offspring all smooth.

No.	9 Black.	cf Agouti.	AgLb	Black	W
1	2 B 9 9 BB	C128Ag pureC...	6

Table 139.

Cross 78.— Black (BW) X agoutf (§, \, \ cutleri).

Expectation : 1 Ag : 1 black : some white. Parents and offspring all smooth.

No. 1 2 3 4 5 6 7 8 9 10	9 Black (or agouti).	I'd1 Agouti (or black).	AgLb	Black	W
G B 9 9 BW 2 B 9 9 BW 3B 9 9 BW K20Ag(}) 78-1.... K22A*(« 78-1....	C67 Ag K4 Ag (i) K-M Ag(I) 39 B BW . .	(J) 78-1 .... 78-1 ....	12 6 9 3 1 3	12 9 5	6
Do . .	3 2 4 1 5	1 3
K29Ag(*) 78-1....	Do . .
K103Ag(i) 78-4....	Do . .
K66Ag(|) 78-2....	Do . . .
E109Ag (I) 78-6....	Do . .	1
3 H 9 9 BW Total	K104Ag(i)	78-4 . .
35	41	10

PART III

FURTHER STUDIES OF PIEBALD RATS AND SELECTION, WITH OBSERVATIONS ON GAMETIC COUPLING

By W. E. CASTLE

THE PROGENY OF HOODED RATS TWICE CROSSED

WITH WILD RATS.

In 1914 Castle and Phillips published a report on breeding experi- ments with hooded rats, in which it was shown that the hooded color pattern — itself a Mendelian recessive character in crosses with the entirely colored (or "self") coat of wild rats — is subject to quantitative variation, and that different quantitative conditions of the hooded pattern are heritable. (Compare fig. 36, plate 7.) It was also shown that by repeated selection of the more extreme variations in the hooded pattern (either plus or minus) it is possible gradually to modify the racial mean, mode, and range as regards these fluctuations, without eliminat- ing further fluctuation or greatly reducing its amount. We concluded that the unit character, hooded color pattern, is a quantitatively vary- ing one, but were at that time unable to decide whether the observed variability was due simply and exclusively to variation in a single Mendelian unit factor or partly to independent and subsidiary modify- ing Mendelian factors.

Since publication of the above I have been engaged in further experi- ments designed to show which of the alternative explanations is the correct one, and these are now sufficiently advanced to indicate definite conclusions. Previous experiments had shown that when a race of hooded rats, whose character has been modified by selection (either plus or minus), is crossed with wild rats, the extracted hooded animals obtained in F2 as recessives show regression toward the mean condition of the recessive race before selection began. This result suggested that the regression observed might be due to removal by the cross of modifying factors, which selection had accumulated in the hooded race. If this view was correct, it was thought that further crossing of the extracted hooded animals with the same wild race should result in further regression, and that if this further regression was not observed a different explanation must be sought for the regression already noted.

The entire experiment has accordingly been repeated from the beginning, with the same result as regards regression in the first F, generation, but with no regression of the same sort in a second F2 con- taining twice-extracted hooded animals. So far from observing further regression as a result of the second cross with wild rats, we have unmis- takable evidence that the movement of the mean, mode, and range of the hooded character has been in the reverse direction. So the hypothe- sis of modifying factors to account for the regression and for the pro- gressive changes observed under selection becomes untenable.

In repeating the experiment of crossing hooded rats of our selected races with wild rats, great care has been taken to employ as parents individuals of the greatest racial purity and to inbreed the offspring

163

164

INHERITANCE IN RATS.

brother with sister, thus precluding the possibility of introducing modifying factors from other sources. In making the second set of crosses, the extracted individual has, wherever possible, been crossed with its own wild grandparent. In the few cases in which this was impossible, wild animals of the same stock have been used. This stock consisted of a colony of wild rats which invaded the basement of the Bussey Institution apparently from a nearby stable. Owing to faulty construction of the building they were able to breed in spots inaccessible to us, and it took many months of continuous and persistent trapping to secure their extermination. During this period we trapped a hun- dred or more of them, all typical Norway rats, colored all over, without even the white spot occasionally seen on the chest of wild rats. Two genei ations of rats from this wild stock have been reared in the labora- tory, and all have this same self-colored condition.

The hooded animals used in the experiments to be reported on in this connection consisted of 4 individuals of the plus-selected series, a male and 3 females, as follows:

Table 140.

Individual.	Grade.1	Generation.
9 5513.... c?6348 9 6600 .... 9 6955	+4| +4 +4* +4	10 10 12 12

'See figure 35, plate 7, for significance of the grades.

Each of these animals was mated with a single wild mate, and their children were weaned directly into breeding cages containing a male and two or three females (brother and sisters). In the case of two inatings, Fx males of the same parentage were at the time lacking and 1 1 1 ales from a different cross were used. The results of such matings are tabulated by themselves and serve a useful purpose as controls. The Fx animals all closely resembled their wild parents, but many of them had a white spot on the chest. They ranged from grade +5j to +6 (self).

The F2 animals are classified in table 141, where it appears that 73 of them were hooded and 219 non-hooded (i. e., like Fi), an exact 1 : 3 ratio. More than half of this F2 generation consists of the grand- children of 9 5513, produced by breeding her children brother with sister, those children all having been sired by the same wild rat. Her grandchildren include 41 hooded and 107 non-hooded young. The hooded young range in grade from +1| to +4, their mean grade being +3.05, a considerable regression from the grade of the grandmother, which was 4.26.

Hooded rats of the same grade and generation as the grandmother, when bred with each other, produced young of mean grade +3.84.

HOODED CROSSED WITH WILD. 165

(See table 10, Castle and Phillips.) The mean of the extracted hooded grandchildren in this case (being 3.05) shows a regression of 0.79 from that expected for the uncrossed hooded race. From the extracted hooded grandchildren of 9 5513, produced as just described by a cross with a wild male, 7 individuals, 2 males and 5 females, were selected for a second cross with the wild race. They ranged in grade from +2 to +3*. (See table 142.) They produced several litters of young of the same character as the first Fx young, all being similar to wild rats in appearance, except for the frequent occurrence of a white spot on the belly. These second Fx young were at weaning time mated, brother with sister, in breeding-pens, precisely as had been done with the first F/s. They produced 394 second F2 young, of which 98 were hooded and 296 non-hooded, a perfect 1 : 3 ratio. The hooded young varied in grade from +2 to +4, as shown in table 142, the data there being given for each family separately as well as for all combined in the totals. One family was very like another as regards the character of the hooded young, except that the higher-grade grandparents had grandchildren of slightly higher grade. Thus the average of all the 98 hooded young was +3.47, but the average of those descended from the 3 grandparents of lowest grade was less than this, while the average of those descended from the 3 grandparents of highest grade was greater. This is just what had been observed throughout the entire selection experiments. (See Castle and Phillips.)

If we weight each of the grandparents in table 142 in proportion to the number of its hooded grandchildren, then the mean grade of all the grandparents is +2.95. Since the mean grade of all the 41 first F2 hooded grandchildren, from which these 7 were chosen, was +3.05, it will be seen that these 7 are, so far as grade is concerned, fair repre- sentatives of the 41, being in fact of slightly lower mean grade. It is therefore all the more striking that their grandchildren, the second F2 hooded young (table 142) , are of higher grade. They regress in an oppo- site direction to that taken by the first F2 hooded young. Thus the original hooded ancestor ( 9 5513) was of grade 4.25. The grade of hooded young expected from such animals is 3.84. What she produced in F2, following a cross with the wild male, was young of mean grade 3.05^ Seven of these of mean grade 2.95 produced a second F2 contain- ing hooded young of mean grade 3.47. This is a reversed regression of 0.52 on the grade of their actual hooded grandparents, or of 0.42 on the group from which their grandparents were chosen. Their mean lies about midway1 between that which would have been expected from the original hooded female (5513) had no crossing with wild rats occurred and that which was observed in the first F2.

iln The Scientific Monthly (Jan. 1916) I have stated that a second cross showed "a return to about what the selected race would have been had no crossing at all occurred This is obviously inaccurate and should be corrected. It rests on a comparison with the combined average of both the older and the more recent experiments.

],,i; INHERITANCE IN RATS.

Obviously these facte do not harmonize with the assumption that the repression observed in the first F2 was due to loss of modifying fac- tors accumulated during the ten preceding generations of selection; for no further loss occurs in the second F2. On the other hand, a part ial recovery is made of what was lost in the first F2. This suggests the idea that that loss may have been due to physiological causes non- genet ic in character, such as produce increased size in racial crosses; for among guinea-pigs (as among certain plants) it has been found that Fx has :m increased size due to vigor produced by crossing and not due to heredity at all. This increased size persists partially in F2, but for the most pari is not in evidence beyond Fx. I would not suggest that the present case is parallel with this, but it seems quite possible that similar non-genetic agencies are concerned in the striking regression of the first 1 and the subsequent reversed regression in the second F2.

Whatever its correct explanation may be, the fact of the reversed regression in a second F2 is very clear, as other cases than those already iliscussed will show.

A hooded rat of grade +4 and generation 10, c?6348, had by a wild female several young of the character already described for the young of 9 5513. These, mated brother with sister, produced a first F2 (table 141) of 90 rats, 22 of which were hooded, 68 being non-hooded, again a good 1 : 3 ratio. The hooded young ranged from +2 to +4 in grade, their mean being 3.28. Of the 22 hooded individuals, 1 male and 7 females were mated with wild rats to obtain a second Fi, and the second Fx animals were then mated brother with sister to obtain the desired second F2. The character of this is shown family by family in table 143. It contained 497 individuals, of which 121 were hooded and 376 non-hooded, a ratio of 1 : 3.1. The weighted mean of the 8 selected grandparents is 2.93, which is 0.35 below the mean of the 22 first F2 hooded animals which they represent. The mean of the second I hooded young is 3.22, which indicates a reversed regression of 0.29 on the grade of the grandparents, but shows no significant difference from the mean of the grandparental group (3.28).

All except one of the 8 families classified in table 143 show unmis- takably the reversed regression. This exceptional family consists of the grandchildren of 9 9747. They have a mean grade of 2.90, sub- stantially the same as that of the entire group of grandparents but con- siderably lower than that of their own hooded grandmother. Appa- rently she did not come up genetically to her phenotypic grade. This the other grandparents of the group did. For those of lowest grade

2|) produced lower-grade hooded grandchildren than did the grand- parent > of highest grade (3-|, 4), as was found to be the case also in table L42.

We may next trace the inheritance of the hooded character through a third but smaller family produced by two successive crosses with wild

HOODED CROSSED WITH WILD. 167

rats, the hooded character in this case being derived from 9G955, grade +4, generation 12. The character of her first F2 descendants is shown in table 141. They consist of 5 hooded and 27 non-hooded individuals. The mean grade of the hooded young is 3.51, but the number of these young is too small to make this mean of much signifi- cance. One of the hooded young (d"9660,+3f) was mated with a wild female to secure a second F: generation and from this in due course was produced the second F2 generation (table 144) . It consisted of 21 hooded and 44 non-hooded young. The hooded young showed the usual range (2 to 4). Their mean grade was 3.50, substantially identical with that of the first F2 animals, but 0.25 below that of the actual hooded grand- parent. This family history is less satisfactory than the two already discussed because of the smaller numbers which it includes. It con- tains nothing contradictory to the interpretation already given, though reversed regression is not in this case in evidence.

In two cases Fi females could not be mated with brothers and so mates were taken from other families. Thus "mixed Fi matings" were made between children of 5513 and 6600 and children of 5513 and 6955. (See table 141.) The former mating produced 3 hooded and 12 non-hooded " first" F2 young; the latter produced 2 hooded and 5 non-hooded " first" F2 young. The grade of the hooded young pro- duced by these mixed matings was not different from that of brother- sister matings, so far as the small numbers permit one to judge. One of these mixed matings was carried into a second F2 generation. The first F2 hooded ^9711, +3|, was mated with a wild female, and the young were bred, brother with sister, producing 16 hooded and 33 non- hooded young. (See table 144.) The mean grade of the 16 hooded young was 3.28, nearly the same as that of the first F2 hooded grand- parent. No additional regression through loss of modifiers (or other agency) is here in evidence. The result is the same as that observed in families wholly unmixed. The attention of my pure-line critics, who think that in our mass-selection experiments insufficient attention has been given to individual pedigrees, is particularly directed to the foregoing case.

Having now discussed each family history separately, we may com- bine all the second F2 families in one table, in order to get a clearer impression of the results as a whole. (See table 145.) The second F2 generation thus combined includes 256 hooded and 749 non-hooded individuals, a ratio of 1 : 2.9, an unmistakable mono-hybrid Mendelian ratio. The mean grade of the hooded individuals is 3.34. The weighted mean grade of their hooded grandparents was 3.02, which indicates a reversed regression of 0.32 for the entire second F2 group of hooded animals.

Classified according to the grade of the (first F2) grandparent, they show a correlation between grade of grandparent and grade of grand-

108 INHERITANCE IN RATS.

child. The 1«>\\ er-grade grandparent has lower-grade hooded grand- children, and the higher-grade grandparent has higher-grade hooded grandchildren. This shows that the variation in grade is (in part at least) </< not ii pic. As the experiment yields no evidence that the varia- tion in the hooded character is due to independent modifying factors, there remains no alternative to the conclusion that the single genetic Mendelian factor concerned fluctuates in genetic value. Fluctuation accordingly is not exclusively phenotypic, as DeVries and Johannsen have thought, but may be genetic also. Hence racial changes may be effected through selection by the isolation of genetic fluctuations, as well as by the isolation of mutations. Moreover, genetic fluctuation makes possible progressive change in a particular direction, repeated selection attaining results which it would be quite hopeless to seek by any other means.

A SECOND REPORT ON MASS SELECTION OF THE HOODED PATTERN OF RATS.

The experiments in selection for the modification of the hooded pat- tern of rats, when reported on by Castle and Phillips in 1914, had been carried through 13 generations. Since then the experiments with the same selected races have been carried through 3 or 4 additional genera- tions, the results of which will now be described. Additional records have also been obtained for certain of the generations reported on by Castle and Phillips, which may now be combined with those previously pub- lished. Thus, revised data, based on larger totals, may be given for generations 12 and 13 of the plus-selection series and for generation 13 of the minus-selection series. These do not materially change the results previously obtained, but add to their trustworthiness. The additional generations of selection show a continued progressive movement of the racial character in the direction of the selection and indicate the exist- ence of no natural limit to the progress which selection can make in changing the hooded character.

For details concerning the earlier history of the experiments and the methods of grading the animals the reader is referred to the publica- tion of Castle and Phillips. The grading scale (exclusive of the newer and more extreme grades) is reproduced in figure 35, plate 7. Atten- t ion may be called to the fact that the entire selection series, both plus and minus, consist of animals descended from an original stock of less than a dozen individuals. These descendants number more than 33,000. In their ancestry, since the beginning of the selection experiment, not a -ingle cross out of the race has occurred. At the same time no effort has been made to avoid inbreeding. Brother and sister and cousin platings are frequent in our records. Under these circumstances it is inevitable that the selected races should have become much "inbred."

MASS SELECTION. 169

Our critics with a leaning toward the "pure-line" idea have insisted that nothing but brother-sister matings should have been employed in our experiments. We have several times endeavored to carry forward certain high-grade families on this basis, but have been unable to secure large enough numbers of offspring to make this possible; but we have in several cases produced families of considerable size, descended exclu- sively from a single pair of ancestors — notably in the case of our pure "mutant" race and in a race descended from one hooded and one wild rat, which race was continued through 8 filial generations. (See p. 21, Castle and Phillips.) It would have been impossible, in these and other races, to make as rapid progress as we secured through selection in our two principal races, for when only brother-sister matings are permitted, it often happens that a mate of proper grade can not be secured for an individual among its own brothers and sisters, though such a mate may be found among its cousins or more remote relatives. It being our first object to test the effectiveness of selection, we have made selec- tion of any individual within the group (series or family with which we were dealing) regardless of relationship, making the selection as rigid as the maintenance of a stock of considerable size would permit. More than once we have crossed the danger-line in advancing the standard of selection to such an extent that only small numbers of parents came up to it ; more than once we have had to relax our standard temporarily in order to keep the race alive.

That the long-continued inbreeding of our selected races has affected their vigor and fecundity is unquestionable. It is shown by the fact that the plus and minus races, which had a common origin many generations ago and have ever since been bred in the same room and under identical conditions, if crossed with each other, produce offspring of much greater vigor and fecundity than either parent strain. In this our observations on the effects of inbreeding are entirely in harmony with those of Darwin, Bos, Weismann, and of breeders of farm animals quite generally. Miss King is credited with the view that inbreeding of rats may increase their size, vigor, and fecundity, but this is cer- tainly contrary to common experience with these and other animals. It is probably true that under inbreeding it is possible, in exceptional cases, to isolate a strain relatively immune to ill effects from inbreeding (like Darwin's "Hero" morning-glory) or so inherently vigorous that it succeeds in spite of inbreeding. But it is very doubtful whether inbreeding of itself affects vigor other than disadvantageously. It is a sufficient test to cross-breed an inbred strain, in order to ascertain whether the inbreeding has increased or impaired its vigor.

170 INHERITANCE IN RATS.

PLUS AND MINUS-SELECTION SERIES.

The plus-selection experiment, when described by Castle and Phil- lips, had been carried through 13 generations, but the last 2 genera- tions were incomplete. The number of offspring included in genera- i ion 12 (table 140) has now been raised from 590 to 682 and the number of offspring included in generation 13 (table 147) has been raised from 194 to 52!). The mean grade of the parents for generation 12 has advanced from 4.09 to 4.10; that of the offspring has fallen from 3.94 to 3.93. Neither of these changes is of significant size. The correla- tion is now found to be 0.168 instead of 0.161.

In generation 13 (table 147) the changes are greater, as might be expected from the greater change in the number of observations. The mean of the parents is now 4.13 (formerly 4.22) ; that of the offspring is 3.94 (instead of 3.88). The correlation is 0.117, as compared with 0.132, the value previously obtained.

feneration 14 (table 148) includes 1,359 offspring of mean grade 4.01 . They are descended mostly from parents of grade +4 or higher, mean 4.14.

( feneration 15 (table 149) includes 3,690 individuals, more than have been produced in any other generation of the series. The mean grade of the parents was in this generation advanced about a quarter grade to 4.38; that of the offspring advanced a little, to 4.07.

Generation 16 (table 150) was also large, including 1,690 offspring. The grade of the parents was again advanced a little to 4.45; that of the offspring followed a similar amount, to 4.13.

In the three generations (14 to 16) which have been added since the last report, the grade of the selected parents has been advanced by 0.32, from 4.13 to 4.45; that of the offspring has advanced 0.19, from 3.94 to 4.13 (the mean grade of the parents three generations earlier).

The upper limit of variation of the offspring has meanwhile advanced from 5.25 to 5.87, the highest grade being found in a rat black all over except for a few white hairs on the chest. This rat has produced a few offspring of almost as high grade, though the most of his young are of much lower grade.

In the minus-selection series, generation 13, in our previous report, contained 571 offspring. This number has now been raised to 1,006 (table 151), the mean grade of both parents and offspring being prac- t ica lly unchanged by the additional young recorded. The parents are of mean grade -2.49, the offspring of mean grade —2.40.

In the next generation (14) the offspring number 717, their mean grade being -2.48, that of the selected parents being -2.64. (See table 152.)

Generation 15 includes 1,438 young of mean grade -2.54. The mean made of the parents is -2.65. (See table 153.)

MASS SELECTION. 171

Generation 16 is the largest in the minus-selection series. It includes 1,980 young of mean grade —2.63. The grade of the parents is —2.79. (See table 154.)

Generation 17 (table 155) includes 868 young of mean grade —2.70. The grade of their parents is —2.86.

Four generations of selection have thus been added to the minus series as it stood at the last report. The mean grade of the parents has been advanced from —2.49 to —2.86; that of the offspring from -2.40 to -2.70, the former is an advance of 0.37, the latter of 0.30. In the plus series the corresponding changes for one less generation of selection (three), were 0.32 and 0.19, respectively. In both series a change in the mean of the offspring attends that in the parents, coin- ciding with it in character but not quite equaling it in amount.

The lagging behind of the offspring, as compared with their selected parents, gives a good illustration of regression, the phenomenon made familiar by Galton's researches, but explained away by Johannsen as due to a sorting-out action of selection on mixed races. The extent to which in these experiments the offspring lag behind their parents or " regress on their parents" is indicated in each table in the column headed " regression." Tables 146 and 150 illustrate particularly well how the offspring regress toward the general average of the race for the time being. The offspring of parents substantially the same grade as the general average of the race show no regression; the offspring of parents below this average show regression upward (indicated in the tables by the minus sign) ; the offspring of parents above the racial average show regression downward, the amount of the regression increas- ing with the aberrant character of the parents.

If one examines either selection series as a whole (compare Castle and Phillips), he will notice that the point (toward which regression occurs) changes with the progress of the selection. At the beginning of the plus-selection series regression was toward a grade of about + 1.75 (see table 1, Castle and Phillips); after about 15 generations of plus selection it has advanced to +4.00. (See tables 148 to 150.) At the beginning of the minus-selection series, regression occurred toward a grade of 0 to —1 (Castle and Phillips, tables 16 and 17); in generation 17 (table 155) regression is apparently toward grade —2.62. These grades toward which regression occurs represent points of racial equilibrium or stability at which the race would tend to remain in the absence of further selection, but these points of equilibrium are capable of being moved either up or down the scale of grades at the will of the breeder, provided he has patience and persistency and will select repeatedly.

Regression indicates that there is not complete agreement between the somatic and the genetic character of the parents selected. But the steady movement (in the direction of the selection) of the point of

172 INHERITANCE IN RATS.

equilibrium toward which regression occurs serves to show that geno- typic as well as phenotypic fluctuations occur in the material on which selection is brought to bear. DeVries and Johannsen have damned the word fluctuation by ascribing to it purely phenotypic sig- oificance. Is it not worth while to rescue the term from its present odious position, since it is clear that variation having a genetic basis may in every way resemble somatic fluctuation, except in its behavior under selection? Fluctuation may conceivably be either somatic or genetic or both. No one, in advance of actual experiment, can tell what its nature is in a particular case. In the case under discussion the fluctuation is obviously partly somatic and partly genetic. The somatic fluctuation occasions regression, the genetic fluctuation per- mits a change (under selection) of the point toward which regression occurs — that is, in the general average of the race.

Tables 156 and 157 show (generation by generation) the progress made by selection in modifying the racial character. It will be observed that as the mean advances in the direction of the selection both the upper and the lower limits of variation move in the same direction. The amount of the variation as measured by the standard deviation is less in the last half of the experiment than in the first half. It is also steadier, owing in part doubtless to the fact that the numbers are larger, and in part to a more stable genetic character of the selected races. But the genetic variability is plainly still large enough to per- mit further racial modification and there is no indication that it will cease until the hooded character has been completely selected out of existence, producing at one extreme of the series all-black rats, and at the other end of the series black-eyed white rats.

FURTHER OBSERVATIONS ON THE "MUTANT" SERIES.

Castle and Phillips described, under the name " mutants," 2 rats of the plus-selection series of very high grade. They proved to be heterozygotes between the average condition of the plus-selected race at that time, about +3.75, and a new condition, not previously known in our hooded races, but resembling that seen in "Irish" rats, which are black all over except for a white spot on the belly and would be classed on our grading scale as about +5|. In later generations we secured animals homozygous for the darker condition just described (that of Irish rats). The homozygous "mutant" race proved to be very stable in color-pattern, varying only from b\ to 5f, with a majority of ani- mals graded 5|. Attempts to alter the modal condition of the race by selection have thus far proved futile because of our inability to increase the race sufficiently to afford a basis for selection. Its inbred- ness and its feebleness are perhaps causally related.

The suggestion was made that the change from our plus-selected race, which had occurred in the mutant stock, might be due to some supplementary modifying factor, not to a change in the hooded factor itself. If so, a cross with a race lacking the hooded factor or its "modi- fiers" might serve to demonstrate their distinctness by separating one from the other. A wild race seemed best suited for a test of this hypothesis, since it would be free from suspicion on the possible ground of harboring either the hooded pattern or its supposed modifier, which had converted the hooded pattern into the mutant. It was to be expected, if the hypothesis were correct, that the mutant character was hooded plus modifier; that then a cross with wild should produce in F2 hooded young (lacking the modifier) as well as mutants and selfs. But if the mutant race had arisen through a change in the hooded factor itself, then the cross should produce only mutants and selfs, without hooded young in F2. Crosses have now been made on a sufficient scale to show beyond question the correctness of the latter alternative, which is entirely in harmony also with the results described in the preceding parts of this paper.

Six homozygous "mutant" females of grade +5| were mated with wild males of the same race described in Part I. They produced 46 young, all gray like wild rats and of grades as follows:

Grade 5£ 5f 5| 6

No 1 15 7 23

Exactly half of the 46 Fx rats bore no white spot, i. e., were of grade +6. Seven more bore only a few white hairs (grade 5|). The remain- der were very similar to the mutant parent in grade.

Several matings were made of the Fj rats, brother with sister, which produced 212 F2 young. About a quarter of these were black (non- 173

171 INHERITANCE IN RATS.

agouti), the rest being gray (agouti). Both sorts included about equal numbers of individuals with and without white spots. No difference was observed in this respect between the progeny of spotted and of unspotted parents. Table 158 shows the F2 young grouped family by family according to grade. Three of the four families are descended from a single mutant grandparent; the fourth family is descended from two different mutant grandparents which were bred simultane- ously to the same wild male in the same cage. The 10 F2 young of this family may have been produced either by full brother and sister, or by half-brother and half-sister; it is uncertain which. All other 1 young were produced by brother-sister matings.

It will be observed that the F2 young (table 158) which are white- spotted are in no case hooded. Their range of variation does not fall beyond that of the uncrossed mutant race. It is certain, therefore, that the "mutant " condition is not hooded plus an independent Mendel- ian modifier. It is a changed form of white-spotting, alternative to the form of spotting found in the race from which it was derived (the plus- selection series, generation 10). It is, without much doubt, also alter- native to the self condition of wild rats, though fluctuation in grade ob- scures the segregation, which may, very likely, be imperfect. This serves to confirm the general conclusion that throughout the entire series of experiments with the hooded pattern of rats we are dealing with quan- titative variations in one and the same genetic factor.

GAMETIC COUPLING IN YELLOW RATS.

Two yellow-coated varieties of the Norway rat (Mus norvegicus) made their appearance as sports or mutations in England a few years since (Castle, 1914) and are now recognized as distinct varieties by fanciers. Both are similar in appearance except for the eye color. In one variety the eye is pink, showing under gross inspection only the color of the blood in the retina. In the other variety the eye is a reddish-black, owing to the combined effect of the red-colored blood and the black-pigmented retina. Since the retinal pigment is much less in this variety than in rats with gray or black coats, the eye is redder. It will be convenient to distinguish the dark-eyed yellow variety as red-eyed, reserving the name black-eyed for gray or black rats.

In the coats of both the pink-eyed and the red-eyed varieties of yellow rats black pigment is very feebly developed. It is in fact of a pale cream color. But the true yellow pigment seen on the tips of the hairs of gray rats is retained in full intensity in the yellow varieties. For this reason agouti varieties of yellow rats are much brighter-colored than non-agouti varieties. A non-agouti yellow variety has fur cream- colored throughout its length ; the corresponding agouti variety has fur of this same cream color at its base, where the fur of gray rats is black- pigmented, but the hair- tips are of a bright yellow color of exactly the same shade as the hair- tips of gray rats. Hence it is clear that in these yellow varieties of rats a genetic factor for black pigmentation has been affected without any apparent change in the genetic apparatus for producing ordinary yellow pigment.

This is quite different from the genesis of yellow coat in most ro- dents— for example, in guinea-pigs and rabbits — in which black pig- ment is not apparently changed in character but merely in distribution, being "restricted" chiefly to the eye. In the yellow varieties of rats black pigment seems to be affected in the same way as in the pink-eyed variety of guinea-pigs and mice, viz, to be greatly weakened without affecting in the least the development of yellow pigment. The genetic behavior as well as the appearance of the pink-eyed yellow variation in rats is in every way parallel with the behavior of the variations known by the same name in mice and guinea-pigs. But red-eyed yellow in rats is a genetically distinct variation, as we shall presently see. In no other mammal does there occur a parallel variation, so far as I know. Both red-eyed yellow and pink-eyed yellow were found to be recessive Mendelian variations in crosses with black-eyed rats. From a cross between black-eyed and red-eyed an F2 generation oL609 rats was raised, of which 452 were black-eyed and 157 red-eyed; expected, 457 : 152. From a cross between black-eyed and pink-eyed rats, cer- tain F2 females were back-crossed with a pure pink-eyed male. They produced 46 black-eyed and 39 pink-eyed ; expected, 42 of each.

175

17f, INHERITANCE IN RATS.

The pink-eyed yellow and red-eyed yellow of rats are complementary variations; for when the two varieties are crossed with each other they produce r\ offspring which are either gray or black pigmented, according as their yellow parents did or did not transmit the agouti factor. These F! reversionary grays or blacks are paler in pigmenta- tion than ordinary gray or black rats, indicating that neither character in a heterozygous form is the full complement of the other. But it is evident that in homozygous form each is the full complement of the other, since in F2 and later generations grays and blacks of full intensity

are obtained.

The F] black-eyed animals (blacks or grays) obtained by crossing pink-eyed yellows with red-eyed yellows, if mated with each other, pro- duce an F2 generation containing (1) black-eyed young (black or gray), (2) red-eyed yellow young, and (3) pink-eyed yellow young. We have obtained thus far 324 such F2 young, of which 162 were of class (1), 90 of class (2), and 72 of class (3).

If, as suggested, red-eyed yellow and pink-eyed yellow are due to mutually independent Mendelian factors, then F2 should contain four classes instead of the apparent three; wTherefore it seemed probable that one of the three classes was really composite and that the three should be as 9 : 3 : 4. On this basis the F2 expectation would be 182 : 61 : 81 instead of the observed 162 : 72 : 90. Hence there appear to be fewer black-eyed young than are expected. Further, when we came to test the other F2 classes to discover which of them was com- posite, we found very few individuals wThich would fall in the hypotheti- cal fourth class transmitting both pink-eyed and red-eyed yellow in the same gamete. Instead of 1 in 16 as expected, we have been able to discover a much smaller number of double recessives. Both a defi- ciency in double recessives and a deficiency in double dominants (the black-eyed class), which have been observed among the F2 rats, would be expected if pink-eyed yellow and red-eyed yellow are due to "linked genes," i. e., to factors located near each other in the germ-plasm. For in the cross under consideration each form of yellow enters the Fi zygote in a different gamete. Hence, in the gametes arising from such zygotes we should expect the two forms of yellow to show mutual repulsion. If they did so, then the gametes formed by F: zygotes, of the four possible combinations, RP, Rp, rP, and rp, would not be equally numerous, but Rp and rP should be more numerous than RP and rp. That this is true is indicated by the facts presently to be stated. To test the gametic composition of the F2 yellows, those which wore red-eyed were mated with pink-eyed yellows of pure race, and those which wore pink eyed were mated with red-eyed yellows of pure race. For it was known that, since red-eyed yellow is a recessive variation, every red-eyed F2 yellow must be homozygous for red-eye, but conceivably it might bo either heterozygous for pink-eye or might lack it altogether. A cross with the pure pink-eyed yellow race would

GAMETIC COUPLING IN YELLOW RATS.

177

decide between these possibilities. Further, it was clear that every pink-eyed F2 yellow must be homozygous for that character, which is also recessive, but might be either homozygous or heterozygous for red-eye without affecting its appearance, or might even lack the gene for red-eye altogether. A cross with pure red-eyed animals would suffice to show in each case which possibility was realized. In accord- ance with this reasoning the proposed tests have been made in the case of 45 red-eyed and 40 pink-eyed F2 yellows.

Of the 45 red-eyed yellows tested, 32 have given exclusively black- eyed young (blacks or grays), no test being considered adequate which did not produce 4 or more young; but 13 of the tested animals gave a mixture of black-eyed and of red-eyed young in approximately equal numbers. The former group, numbering 32, evidently lacked the gene for pink-eye, since they always produced atavists in crosses with pink- eyed yellows; the latter group, numbering 13, were evidently hetero- zygous for pink-eye, since only part of their young were atavistic.

Of the 40 pink-eyed F2 yellows which were tested, 27 produced only black-eyed young; these evidently lacked the gene for red-eye. Ten others produced both black-eyed and red-eyed young, being evidently heterozygous for red-eye. Three have produced only red-eyed young, which shows them to be homozygous for red-eye as well as for pink-eye. Hence they are the double recessives, expected to be one-sixteenth of all F2 rats if no linkage occurs, but less numerous if linkage occurs.

We are now in a position to estimate the strength of the linkage shown. If we designate by r the recessive gene for red-eye, and by p the recessive gene for pink-eye, then in the current Mendelian terminol- ogy the following F2 classes are to be expected in the frequencies shown, if no linkage occurs:

Black-eyed.	Red-eyed.	Pink-eyed.
1 RRPP... 2 RrPP . . . 2 RRPp . . .	1 rrPP 2 rrPp	1 RRpp 2 Rrpp 1 rrpp
4 RrPp
9	3	4

For the present we may pass by the black-eyed classes, since none of these were individually tested. The individual tests already described have shown the existence of the expected two classes of red-eyed and three classes of pink-eyed young, but in proportions very different from those given in the table. Among the red-eyed, instead of the expected 1 rrPP : 2 rrPp, we observe 32 : 13. Among the pink-eyed, where we expect 1 RRpp : 2 Rrpp: 1 rrpp, we observe 27 : 10 : 3. These are very different frequencies from those expected, and they strongly suggest linkage. How strong is the linkage? We may estimate it

178

INHERITANCE IN RATS.

i nun the actual proportions of the four possible kinds of gametes which the Fj parents produced. With no linkage these gametes should be of four sorts, all equally numerous, viz, RP + Rp + rP + rp- Linkage would tend to increase the proportion of the two middle classes (Rp and pP, the original combinations) at the expense of the extremes (RP and rp, the double dominant and double recessive classes). The latter may be called "cross-over" classes, the former " non-cross-over." In producing the 85 F2 yellow rats which were tested, twice that number of gametes were concerned, viz, 170. From the demonstrated genetic constitution of the tested animals, we can estimate how many cross- over and how many non-cross-over gametes entered into each.

Zygotes.	Cross-over gametes.	Non-cross-over gametes.
32rrPP 13 rrPp 27 RRpp 10 Rrpp 3 rrpp	13 10 6	64 13 54 10
85	29	141

The estimated proportion of cross-over to non-cross-over gametes is seen to be 29 : 141 or 1 : 4.8. In the terminology of Bateson and Pun- nett this would be a reduplication series lying between 1:4:4:1 and 1:5:5:1; in the terminology of Morgan, 17 per cent of the gametes formed by F: individuals are cross-over gametes.

We can test this linkage theory in another way. If linkage exists it should modify the proportions of the apparent classes in F2 as well as of the real classes, which we have just been considering. The apparent classes are three, viz, black-eyed, red-eyed, and pink-eyed, with observed frequencies of 162 : 90 : 72. If no linkage exists the expected frequencies are 182 : 61 : 81, which deviate considerably from the expected frequencies. But if linkage exists, it will lessen the discrep- ancies. Linkage of 17 per cent strength will change the expectations to 164 : 79 : 81 . This alteration shows agreement almost perfect in the case of the black-eyed class, a much reduced discrepancy in the case of the red-eyed class, and no change in the pink-eyed class — on the whole a much improved agreement between expected and observed frequencies.

Sturtevant has called attention to the fact that double recessives could occur among our F2 animals only as a result of cross-overs occur- ring simultaneously in the gametes of both parents, a fact which Wright and I considered too obvious to demand comment in our preliminary paper, but recognized in our calculation by counting two cross-over gametes for every double recessive zygote. Sturtevant has questioned the adequacy of our tests in the case of these doubly recessive indi- vidual- because apparently he had formed the idea, from studies made "ii insects, that crossing-over could occur only in the gametogenesis of

GAMETIC COUPLING IN YELLOW RATS. 179

one sex. I may say, therefore, that the classification of two animals as double recessives made in our prehminary paper was based on tests which had produced 14 and 9 yellow young respectively. The only possible alternative classification would have involved an expected 1 : 1 ratio of black-eyed to red-eyed young. The chances are overwhelm- ingly great against the observed results being departures due to ran- dom sampling from this expectation. The additional case of a double recessive reported in this paper is so classified on tests which, to the present time, have produced all together 28 yellow young. The num- ber of young produced in each of the other tests is indicated below. Tests taken to indicate that the parent was of the formula rrPP produced only dark-eyed young (gray or black-coated), as follows:

No. of young. ... 4 5 6 7 8 9 10 11 12 13 16 17 19 Cases 244522 13 123 1 2 = 32

Tests showing the parents to be of the formula rrPp gave the follow- ing numbers in 13 tests:

Dark-eyed : Pink-eyed. . .2: 6 4:2 5:4 1:5 3:3 5:3 4:6 1:5 5:5 3:4 3:2 1:5 6:3

Pink-eyed animals were classified as of formula RRpp on the basis of the following tests, which yielded only dark-eyed young:

No. of young.... 4 6 7 8 9 10 11 12 13 14 15 16 18 Cases 213222161311 2 = 27

Pink-eyed animals were shown to be of formula Rrpp by the follow- ing tests:

Dark-eyed: Red-eyed... 10: 7 8:4 1:4 7:8 6:7 5:6 3:1 4:6 4:2 8:2

Both red-eyed and pink-eyed yellow rats, when crossed with albinos, produce an F: generation consisting exclusively of black-eyed (black or gray) young. F2 from the red-eyed cross consisted of black-eyed, red-eyed, and albino young, and F2 from the pink-eyed cross consisted of black-eyed, pink-eyed, and albino young. If no linkage occurs the expectation in each case is 9:3:4, and we at first supposed that this was the ratio approximated. But a summary of all litters thus far obtained indicates a probable linkage between albinism and the two yellow variations.

Thus, red-eyed non-agouti yellows mated with albinos from our plus- selected hooded race produced 17 black F! young. These have given us 58 F2 young, of which 30 are black-eyed, 18 red-eyed, and 10 albinos. A 9 : 3 : 4 ratio would call for 32.5 : 11 : 14.5. It is evident, therefore, that we have too many red-eyed young and too few black-eyed and albinos. Linkage (in this case, repulsion) between red-eye and albin- ism would tend to increase the number of red-eyed and to decrease the number of black-eyed without changing materially the expecta- tion for albinos; hence, linkage seems probable. Linkage involving 1 cross-over to 3 non-cross-over gametes, or 25 per cent cross-over

ISO

INHERITANCE IN RATS.

gametes, would give an expectation of 29.9 black-eyed : 13.6 red-eyed : 14.5 albinos, which agrees much better with the observed numbers (30 : 18 : 10) than does the 9:3:4 distribution. But if red-eye is linked with albinism as well as with pink-eye, then albinism and pink- eye should be linked with each other. Apparently such is the case, for three F2 Utters from the cross pink-eye X albino include 12 black, 12 pink-eyed, and 3 albino young. A 9:3:4 ratio (expected if no linkage occurs) would call for 15 black, 5 pink-eyed, and 7 albinos. Linkage of 5 : 1 would call for 14 : 6 : 7, and perfect linkage would call for 14 : 7 : 7. It is evident that the observed numbers of blacks and albinos are too small on any of these hypotheses, but the existence of linkage would tend to diminish the number of blacks and albinos in proportion to the number of pink-eyed, which is the nature of the deviation observed. To determine definitely whether linkage really occurs between the yellow variations and albinism, and if so, what is its strength, further experiments are needed, which are now in progress. It will also be desirable to determine whether the linkage strength is the same in both sexes.

SUMMARY.

Two yellow variations in rats wThich have recently arisen as muta- tions show mutual repulsion in heredity. When crossed with each other they produce an Fx generation composed exclusively of rever- sionary dark-eyed individuals. The F2 young are of three apparent classes, dark-eyed, red-eyed, and pink-eyed. Their numerical propor- tions deviate somewhat from the typical 9:3:4 ratio. Further, the proportions of the several expected classes of red-eyed and pink-eyed young do not agree with those usually observed in an F2 Mendelian population. But in both cases the deviations are largely accounted for by the supposition that the genes of the respective yellow variations are "linked" (in this case showing repulsion) and that the proportion of "cross-over" gametes is about 17 per cent, or in other words, that non- cross-over gametes are about 4.8 times as numerous as cross-over gametes.

Note. — In the foregoing discussion it has been assumed that the ratio of cross-over to non-cross-over gametes is the same among gametes which take part in producing yellows as among those which take part in producing black- eyed individuals. Theoretically it should be slightly different, as the following table will show :

Ratio cross-over	Per cent	Per cent among	Per cent among
to non-rru---over	on '->-over	gametes pro-	gametes produc-
gametes.	gametes.	ducing yellows.	ing black-eyed.
1: 1	50	42.9	55.6
1: 2	33.3	29.4	36.8
1:3	25	22.6	27.3
1:4	20	18.4	21.6
1:4-4	18.5	17.1	19.8
L:fi	16.7	15.5	17.8
1:6	14.3	13.4	15.2

TABLES.

181

TABLES.

Table 141 shows the classification of extracted hooded first F2 young obtained from crossing hooded rats of the plus-selected series with wild rats.

			Table	141.
Hooded grandparents.	Grade of hooded grandchildren.	Total hooded.	Total non- hooded.	Means of hooded.
1*	1 ;1	2	2\	01 •"2	2|	3	31	3*	•J3	4
9 5513, +4J, gen. 10	1		3 1	2	1 1	7 2	8 4	6 3 1 1 11	5 4 1 1 1 12	7 6 3 16	1 1 2	41 22 5 3 2	107 68 27 12 5	3.05 3.28 3.51 3.17 3.37
cf6348, +4, gen. 10
9 6955, +4, gen. 12
9 5513, +41, and 9 6600, +41, gen. 12							2
9 5513, +41, and 9 6955, +4, gen. 12
Totals	]		4	2	2	9	14
73	219	3.17

Table 142 shows the classification of extracted hooded second F2 young obtained from crossing first F2 hooded rats (table 141) with wild rats. The hooded grandparents were themselves grandchildren of 95513, +4J, generation 10, on the side of both parents.

Table 142.

Hooded grand- parents.	Grade of hooded grandchildren.	Total hooded.	Total non- hooded.	Means of hooded.
2	21	o|	03 "4	3	31	o2	3f	4
9 9619, +2 c?9686, +2f... 99620, +2f... 99729, +2| d*9727, +3 9 9728, +3	1	1 1	2 1	1 2 1	1 1 3 1	1 1 2 1 2 5 1	1 2 2 4 11 6 2	1 3 4 7 8 7	1 4 1 5	2 5 13 10 30 22 16	8 28 24 22 104 68 42	3.37 3.40 3.06 3.62 3.47 3.55 3.70
9 9621, +31 Totals . . .
1	2	3	4	6	13	28	30	11	98	296	3.47

Table 143 shows the classification of extracted hooded second F2 young obtained from crossing first F2 hooded rats (table 141) with wild rats. The hooded grandparents were themselves grandchildren of (^6348, +4, generation 10, on the side of both parents.

Table 143.

Hooded grand- parents.	Grade of hooded grand- children.	Total hooded.	Total non- hooded.	Means of hooded.
1 2 ■■■4	2	21	24	2f	3	31	31	3f	1
0^9639, +2 9 9704, +2f . . . .	1	2 1 7	1 1	1 1	3 1 2 1 1	6 1 7 2 5 2	4 1 1 1 1	15 4 1 4 5 4 2	6 4 2 8 3 1	1 1 2 2 1	39 6 1 27 16 21 9 2	110 16 10 76 47 74 40 3	3.24 3.17 3.50 2.90 3.28 3.48 3.36 3.87
9 9765, +3 9 9747, +31 ... . 9 9703, +3|. . . .	1
9 9705, +3| 9 9748, +31 9 9796, +4 Totals
2	in	2	2	8	23	8	:;.-,	24	7	121	376	3.22

1S2

INHERITANCE IN RATS.

Table 1 11 shows the chissification of extracted hooded second Fs young obtained from crossing first Fj hooded rats with wild rats. The hooded grandparent, d"9660, +3$, was I grandson of v 8965, +4, generation 12, on the side of both parents. The hooded grand- parent, d"9711, + 3i, was a grandson, on the side of one parent, of 9 5513, +4J-, generation 10, and on the side of the other parent, of 96955, +4, generation 12. (See table 141.)

Table 144.

Hooded grand- parents.	Grade of hooded grandchildren.	Total hooded.	Total non- hooded.	Means of hooded.
2	-•;	i i	9J - « 1 2 3	3 1 2 3	31 2 4 6	n	35	4
cf9660, +3 J... c?9711. +31. ..	1		5 4 9	9	2 1	21 16	44 33	3.50 3.28
Totals	1
11	3	37	77	3.40

Table 145 is a combination of tables 142 to 144, in which the second F2 young are classi- fied according to the grade of their first F2 hooded grandparent.

Table 145.

Grade of hooded grand- parents.	Grade of hooded grand- children.	Total hooded.	Total non- hooded.	Means of hooded.
H	2	21	2]	2 i	3	33	3i	31	4
2 n 3 31 31 31 4	l l 2	2 2 7 1 12	1 1 1 1 4	2 1 2 1 6	3 1 2 4 4 1 15	6 3 4 9 9 1 32	5 4 7 6 3 2 27	16 12 18 10 11 5	6 8 IS 13 13 9 1	1 1 5 7 4 2 1	41 34 53 59 46 21 2	118 90 182 151 161 44 3	3.25 3.29 3.48 3.22 3.39 3.50 3.87
3.02	72	65	21	256	749	3.34

Table 146 shows the classification of generation 12, plus-selection series. enlargement of table 12 of Castle and Phillips.

Table 146.

This is an

Grade of parents.	Grade of offspring.	Totals.	Means.	Regres- sion.
21	2 J	2J	3	31	4	31 20	4 7	H	41	i;	5	5j
31					4					35	3.83	-.08
3J						3	23	21	6	2	2			57	3.94	-.06
4				1	2	12	62	66	L2	6	1	1	2	164	3.91	.09
H	o				3	26	106	91	30	6	3	1		267	3.87	.25
41				1		11	25	35	16	5	2			95	3.95	.30
4|						3	6	14	10	7	3	1	1	45	4.17	.20
41							1	•I	1			1		7	4.14	.36
4f				1			3	4						11	3.91	.71
41 4J 5
2			3	5	68	246	242	B2	26	1 12	4	3	1	4.75	.25
4.Ki	682	3.93	.17

TABLES.

183

Table 147 shows the classification of generation 13, plus-selection series. This is an enlarge- ment of table 13 of Castle and Phillips.

Table 147.

	Grade of offspring.
G'ade of parents.		Totals.	Means.	Regres- sion.
	21	3	ol	35	31	4	41	1 1 ' 2	41	5	->.;
3*			1	1	1							3	3.50	0
3* 3i
			1	3	11	3	1	1	1		21	4.08	-.33
31		1	2	4	9	7	4	1	3			31	3.90	-.03
4		1	7	33	60	59	25	17	2	1		205	3.90	.10
4*			1	11	32	33	13	5		1		96	3.92	.20
4i	1			7	23	33	13	1	2			80	3.93	.32
4|		1		2	8	7	6	1	3		i	29	4.03	.34
4*				1	15	17	10	5	1	1		50	4.05	.45
4|					1	2	1					4	4.00	.62
41					3	1		1				5	3.95	.80
41				1		2	1		1			5	4.05	.82
4.13	1	3	11	61	155	172	76	32	13	■l	i	529	3.94	.19

Table 148 6hows the classification of generation 14, plus-selection series.

Table 148.

	Grade of offspring.
Grade of		Totals.	Means.	Regres-
parents.														sion.
	n	2?	3	33	3*	32	4	41	H	41	5	•:>!	H
3*					2	6	3		1					12	3.83	-.33
3|					2	9	4	7	l		1			24	4.02	-.39
3!	i		2	1	11	32	45	18	5					115	3.90	-.15
31			1		28	52	63	28	8	3	1			184	3.97	- .10
4				2	4f	84	122	50	6	2				306	3.92	.08
4|					19	74	72	56	15	3	1	l		241	3.99	.13
41					6	25	48	42	8	1				130	4.04	.21
41				1	2	24	36	29	6	2				100	4.04	.33
4*					3	12	31	37	12	3	2	l		101	4.16	.34
41						11	13	15	11	5	2	l		58	4.23	.39
41						5	16	14	8	2				45	4.17	.58
41 5 51						1	7	14	6	3	1		l	33	4.33	.54
						1	4	1					6	4.25	.87
51			3					1	1		1	l 4		4	4.75	.50
4.14	i		4	113	335	461	315	89	24	9	i	1,359	4.01	.13

[84

INHERITANCE IN RATS.

Table 148 Axmt the Hassification of generation 15, plus-aelection series.

Table 149.

( trade of	( irade of offspring.	Totals.	Means.	Regres- sion.
2 J		3	31	31	3i	4	41	41	41	5	5J	51
3| 31 4 41 41 41 41 4| 4| 41 5 51 51	1		1 1	l 2 1 2 1 1 1	3 5 16 29 22 18 9 3 3	3 10 58 184 L83 207 99 37 25 13 1								7 30 156 670 721 969 506 237 168 146 43 18 19	3.57 3.80 3.91 4.00 4.02 4.06 4.10 4.11 4.22 4.37 4.27 4.37 4.36	.18 .07 .09 .12 .23 .31 .40 .51 .53 .50 .73 .75 .89
11 46 255 296 357 159 87 38 25 9 4 2	2 28 165 165 290 175 88 58 42 21 4 10	7 29 44 71 48 14 24 34 10 7 5	3 8 L'l 6 5 5 15 2 3 1	1 1 3 7 2 9 12 1	1 6 4	2 1 1
4.38	1		2	9	108	820	1,289	1,048	293	69	36	11	4	3,690	4.07	.31

Table 150 shows the classification of generation 16, plus-selection seiies.

Table 150.

Grade of parents.	Grade of offspring.	Totals.	Means.	Regres- sion.
31	31	3J	4	41	41	4f	o	bj	5*	bi	Os
41 41 4| 41 4| 4} 41 5 51	1	■i 4 g 7	26 34 149 25 12	64 64 316 82 50 12 23 8 1	37 40 271 69 61 16 26 25 8	8 5 58 18 26 19 8 15 6	1 9 5 11 10 o o 7	1 4 6 2 2 1	1 2 1 1 1 2 1	1 2 1		1	139 149 816 206 166 66 69 61 18	4.04 4.02 4.08 4.10 4.24 4.47 4.21 4.44 4.39	.08 .23 .29 .40 .38 .38 .66 .56 .73
		5
		1
4.45	1		252	620	553	163	46	16	9	4		1	1,690	4.13	.32

Table 151 shows the classification of generation 13, minus-selection series, enlargement of table 28 of Castle and Phillips.

Table 151.

This is an

	Grade of offspring (minus).
Grade of		Totals.	Means.	Regres-
parents.											sion.
	u	2	21	21	2|	3	31	31
-2J	7	i::	25		L2	1			116	2.25	0
	8	71	80	^7	56	19	5		329	2.39	0
-21	8	65	65	92	46	11	2	1	290	2.38	.12
-2|	3		33	58	n	4	4	1	170	2.50	.12
	1	5	•4	16	L2	1	1		46	2.56	.19
-21		i	8	10	7	2			34	2.42	.45
-3		4	1 216	6 297	5 182	3	2		21	2.59	.41
-2	27	221	47	1 i	2	1,006	2.40	.09

TABLES.

185

Table 152 shows the classification of generation 14, minus-selection series.

Table 152.

Grade of parents.	Grade of offspring (minus).	Totals.	Means.	Regres- sion.
1	H	li	14	2	2|	21	2f	3	31	31
-2i					2	2	14	8				26	2.54	-.29
-2|				2	8	20	22	11	2			65	2.40	-.03
-21	1			3	40	50	73	29	2	1		199	2.36	.14
-2|				2	23	32	59	44	10	1	1	172	2.44	.18
-2f					7	20	42	43	11	4		127	2.60	.15
-21					5	10	25	14	2	2		58	2.52	.35
-3					1	1	15	8	o o	5		33	2.69	.31
-31						6	3	7	4			20	2.56	.56
-3J								4	3			7	2.85	.40
-3|		—	—				O	4	3			10	2.75	.62
-2.64	1	7	86	141	256	172	10	13	1	717	2.48	.16

Table 153 shows the classification of generation 15, minus-selection series.

Table 153.

	Grade of offspring (minus).
Grade of parents.			Totals.	Means.	Regres- sion.
	1 a 1 4	2	21	21	2f	3	31	31
-21 -21		1 11	4 23	4 15	4 13				13 64	2.46 2.39	-.33 -.14
2
-21	1	15	24	47	29	3			119	2.45	-.07
-21	2	39	65	102	68	14			290	2.49	.01
_9£ z8	1	41	97	137	99	24	6	2	407	2.50	.12
"4		7	37	91	70	31	3	1	240	2.60	.15
_9i		4	18	62	70	22	7		183	2.64	.23
-3			1	17	16	13	4		51	2.76	.24
-3|			4	11	17	7	1	2	42	2.73	.34
-31				5	12	8	o	1	29	2.85	.60
-2.65	4	118	273	191	398	124	24	6	1,438	2.54	.11

Table 154 shows the classificatioi of generation 16, minus-selection series.

Table 154.

	Grade of offspring (minus).
Grade of		Totals.	Means.	Regres-
parents.																sion.
	1	u	u	1 3 1 4	2	21	21	•21	3	31	31	Q8 •J 4	4
-21					3	5	1							9	2.19	.06
-2|	1				5	4	3	1						14	2.16	.21
-21					4	27	61	56	3	1				152	2.55	-.05
-2f				1	16	56	188	148	36	5				450	2.58	.04
-2f				1	10	36	130	151	28	5	1			362	2.62	.13
-21				1	11	45	187	230	71	16	1		1	563	2.66	.21
-3					2	12	95	128	65	12	1	1		316	2.73	.27
-31						6	30	36	15	10	1			98	2.74	.38
-31								12	3	1				16	2.83	.42
-2.79	1			3	51	191	695	762	221	50	4	1	1	1,980	2.63	.16

I Mi

INHERITANCE IN RATS.

Tat le 155 &SM* the .-luesification of generation 17, minus-s* lection Feries.

Table 155.

< '. rade of parents.	Grade of offspring (minus).	Totals.	Means.	Regres- sion.
1	2 2 2 4	-1 in 12 34 3	2J 40 77 110 28 5 3	2| 49 81 145 42 18 16	3 11 2S 51 18 17 15 1	3i 1 1 19 7 6 6	34 1 1 1 1	32	4	41 1
-2| -2| -21 -3 -3| -3i -3f	113 202 364 98 48 41 2	-2.63 -2.65 -2.71 -2.75 -2.92 -2.92 -3.25	0 .10 .16 .25 .20 .33 .12
-2.86	1	8	59	263	351	141	to	4			1	868	-2.70	.16

Table 156 summarizes the results of the \ lus-selection of hooded rats continued through sixteen successive generations.

Table 150.

Genera- tion.	Mean grade of parents.	Mean grade of offspring.	Lowest grade of offspring.	Highest grade of offspring.	Standard deviation of offspring.	Correla- tion, parents- offspring.	Number of offspring.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16	2.51 2.52 2.73 3.09 3.33 3.52 3.56 3.75 3.78 3.88 3 98 4.10 4.13 4.14 4.38 4.45	2.05 1.92 2.51 2.73 2.90 3.11 3.20 3.48 3.54 3.73 3.78 3.92 3.94 4.01 4.07 4.13	+ 1.00 -1.00 + .75 + .75 + .75 + 1.50 + 1.50 + 1.75 + 1.75 +2.25 +2.75 +2.25 +2.75 +2.75 +2.50 +3.25	+3.00 +3.75 +4.00 +3.75 +4.25 +4.50 +4.75 +4.50 +4.50 +5.00 +5.00 +5.25 +5.25 +5.50 +5.50 +5.87	.54 .73 .53 .47 .50 .49 .55 .44 .35 .36 .29 .31 .34 .34 .29 .29	.29 .31 .33 .06 .16 .18 .21 .09 .21 .11 .23 .16 .13 .31 .30 .31	150 471 341 444 610 861 1,077 1,408 1 , 322 776 697 682 529 1.359 3,690 1,690
Total							16,107

TABLES.

187

Table 157 summarizes tbe results of the minus-selection of hooded rats continued through seventeen successive generations.

Table 157.

Genera- tion.	Mean grade of parents.	Mean grade of offspring.	Lowest grade of offspring.	Highest grade of offspring.	Standard deviation of offspiing.	Correla- tion, parents- offspring.	Number of offspring.
1	-1.46	-1.00	+ .25	-2.00	.51		55
2	-1.41	-1.07	+ .50	-2.00	.49	-.03	132
3	-1.56	-1.18	0	-2.00	.48	.20	195
4	-1.69	-1.28	+ .50	-2.25	.46	.02	329
5	-1.73	-1.41	0	-2.50	.50	.18	701
6	-1.86	-1.56	0	-2.50	.44	.16	1,252
7	-2.01	-1.73	0	-2.75	.35	.14	1,680
8	-2.05	-1.80	0	-2.75	.28	.09	1,726
9	-2.11	-1.92	- .50	-2.75	.28	.05	1,591
10	-2.18	-2.01	-1.00	-3.25	.24	.15	1,451
11	-2.30	-2.15	-1.00	-3.50	.35	.08	984
12	-2.44	-2.23	-1.00	-3.50	.37	.40	1,037
13	-2.48	-2.39	-1.75	-3.50	.34	.18	1,006
14	-2.64	-2.48	-1.00	-3.50	.30	.28	717
15	-2.65	-2.54	-1.75	-3.50	.29	.35	1,438
16	-2.79	-2.63	-1.00	-4.00	.27	.26	1,980
17	-2.86	-2.70	-1.75	-4.25	.28	.22	868
Total. ..							17,142

Table 158 shows the classification of the F2 young obtained by crossing homozygous "mutant" with wild rats.

Table 158.

Mutant grandparents.	Grade of off- spring.	Totals.
5	5i	:>l	5!	5|	6
9 0630, +5 i		3 2	12 22 \2 3	9 29 11 1	2 1	20 59 19 5	46 114 42 10
90698, +5* 90694, +5i	1
9 0630, +5i, or 0636, +5|		1
Total
1	6	49	50	3	103	212

BIBLIOGRAPHY.

Allen, G. M.

1914. Pattern development in mammals and birds. Amer. Nat., 48, pp. 385-412;

467-484; 550-566.

1903. The present state of knowledge of color-heredity in mice and rats. Proc.

Zool. Soc, vol. 2, pp. 71-78. 1913. Mendel's principles of heredity. Univ. Press, Cambridge, xiv + 413 pp.,

38 figs., 6 pis.

Hat R, E.

1911. Einfuhrung in die experimentelle Vererbungslehre. Berlin, Gebriider Born-

traeger, iv + 293 pp., 80 figs., 9 pis. Calkins, G. N., and Louise H. Gregory.

1913. Variations in the progeny of a single exconjugant of Paramecium caudatum.

Jour. Exi . Zool., 15, pp. 467-525.

Castle, W. E.

1905. Heredity of coat characters in guinea-pigs and rabbits. Carnegie Inst.

Wash. Pub. No. 23, 78 pp., 6 pis.

1906. The origin of a polydactylous race of guinea-pigs. Carnegie Inst. Wash.

Pub. No. 49, 29 pp.

1907. On a case of reversion induced by cross breeding and its fixation. Science,

n. s., 25, pp. 151-153. 1907a. Color varieties of the rabbit and of other rodents : Their origin and inheritance. Science, n. s., 26, pp. 287-291.

1908. A new color variety of the guinea-pig. Science, d. s., 28, pp. 250-252.

1912. Or the origin of an albino race of deer mouse. Science, n. s., 35, pp. 346-348. 1912a. On the origin of a pink-eyed guinea-pig with colored coat. Science, n. s.,

35, pp. 508-510. 1912b. On the inheritance of the tricolor coat in guinea-pigs and its relation to Galton's law of ancestral heredity. Amer. Nat., 46, pp. 437-440.

1913. Reversion in guinea-pigs and its explanation. Carnegie Inst. Wash. Pub.

No. 179, pp. 1-10.

1914. Size inheritance and the pure line theory. Zeit. f. ind. Abst. u. Vererbungs-

lehre, 12, pp. 225-237. 1914a. Some new varieties of rats and guinea-pigs and their relation to problems of color inheritance Amer. Nat., 48, pp. 65-73.

1915. Some experiments in mass selection. Amer. Nat., 49, pp. 713-727. and G. M. Allen.

1903. The heredity of albinism. Proc. Am. Soc. Arts and Sci., 1, 38, No. 21, p. 603.

and H. D. Fish.

1915. The black-and-tan rabbit and the significance of multiple allelomorphs. Amer. Nat., 49, pp. 88-96.

and A. Forbes.

1906. Heredity of hair-length in guinea-pigs and its bearing on the theory of pure gametes. Carnegie Inst. Wash. Pub. No. 49, pp. 1-14. and C. C. Little.

1909. The peculiar inheritance of pink eyes among colored mice. Science, n. s., 30,

pp. 313-314. and J. C. Phillips.

1914. Piebald rats and selection. Carnegie Inst. Wash. Pub. No. 195, 56 pp., 3 pis. and Walter, Mullenix, and Cobb.

1909. Studies of inheritance in rabbits. Carnegie Inst. Wash. Pub. No. 114, 68 pp., 4 pis.

and S. WbiGHT.

L915. Two color mutations of rats which show partial coupling. Science, n. is., 42, pp. 193-195. Cole, L. J.

191 1. Studies on inheritance in pigeons. 1. Hereditary relatione of the principal colors. Agri. Exp. Sta., Rhode Island State College, Bull. 158.

188

BIBLIOGRAPHY. 189

Cuenot, L.

1902. La loi de Mendel et I'h6redit6 de la pigmentation chez les souris. Arch. Zool.

Exp. et Gen. (3), 10, Notes et revue, p. 27.

1903. 2me note. Ibid. (4), 1, Notes et revue, pp. 33-41.

1904. 3me note. Ibid. (4), 2, Notes et revue, pp. 45-56.

1905. 4me note. Ibid. (4), 3, Notes et revue, pp. 123-132.

1907. 5me note. Ibid. (4), 5, Notes et revue, p. 1.

1908. 6me note. Ibid. (4), 6, Notes et revue, pp. 7-15. 1911. 7me note. Ibid. (4), 8, Notes et revue, pp. 40-56.

Darbishire, A. D.

1902. Note on the result of crossing Japanese waltzing mice with European albino races. Biometrika, 2, pp. 101, 165, 282; 3, p. 1. Davenport, C. B., and G. C. Davenport.

1907. Heredity of eye-color in man. Science, n. s., 26, p. 589.

1909. Heredity of hair-color in man. Amer. Nat., 43, p. 193.

1910. Heredity of skin-pigmentation in man. Amer. Nat., 44, p. 641. Detlefsen, J. A.

1914. Genetic studies on a cavy species cross. Carnegie Inst. Wash. Pub. No. 205, 134 pp., 10 pis. Dexter, J. S.

1914. The analysis of a case of continuous variation in Drosophila by a study of its

linkage relations. Amer. Nat., 48, pp. 712-758. Durham, F. M.

1904. On the presence of tyrosinases in the skins of some pigmented vertebrates.

Proc. Roy. Soc. London, 74, pp. 310-313. 1907. Note on melanins. Jour. Phys., 35, pp. 47-48.

1911. Further experiments on the inheritance of coat color in mice. Jour. Genet.,

1, pp. 159-178. East, E. M.

1910. A Mendelian interpretation of variation that is apparently continuous. Amer. Nat., 44, pp. 65-82.

and R. A. Emerson.

1913. The inheritance of quantitative characters in maize. Neb. Agri. Exp. Sta. Research Bull., No. 2, 120 pp., 21 figs. Goodale, H. D., and T. H. Morgan.

1913. Heredity of tricolor in guinea-pigs. Amer. Nat., 47, p. 321.

GORTNER, R. A.

1910. Spiegler's "white melanin" as related to dominant or recessive white. Amer.

Nat., 44, p. 497.

1911. Studies on melanin. IV. The origin of the pigment and color pattern in the

elytra of the Colorado potato beetle. Amer. Nat., 45, p. 743. Groth, B. H. A. (Joint author with Halstead and others.)

1913. Experiments with peppers. Report Bot. Dept. N. J. Agri. Col. Exp. Sta.,

pp. 549-588, illustr. Hagedoorn, A. L.

1912. On tricolor coat in dogs and guinea-pigs. Amer. Nat., 46, p. 682. Hoge, M. A.

1915. The influence of temperature on the development of a Mendelian character.

Jour. Exp. Zool., 18, pp. 241-290. Holmes, S. J., and H. M. Loomis.

1910. Eye color and hair color in man. Biol. Bull., 18, p. 50. Hooker, D.

1915. R61es of nucleus and cytoplasm in melanin elaboration. Anat. Rec, 9, p. 393. Jennings, H. S.

1909. Heredity and variation in the simplest organisms. Amer. Nat., 43, pp.

321-337.

1910. Experimental evidence on the effectiveness of selection. Amer. Nat., 44,

pp. 136-145. Keeble, F., and E. F. Armstrong.

1912. The role of oxydases in the formation of the anthocyan pigments of plants. Jour. Genet., 2, pp. 277-306.

190 INHERITANCE IN RATS.

I.ITTLE, C. C.

I'.Ul. The "dilute" forms of yellow mice. Science, n. a., 33, pp. 896-897.

l'.M l/». The influence of heredity and of environment in determining the coat colors

in mice. Science, n. s., 34, p. 563. 1 « 1 1 3 . Experimental studies of the inheritance of color in mice. Carnegie Inst.

w aah. Pub. No. 179, pp. 11-102, 5 pis. 1914. Dominant and recessive spotting in mice. Amer. Nat., 48, pp. 74-82. 1914a. Coat color in pointer dogs. Jour. Her., 5, pp. 244-248. M \cCurdy, H., and W. E. Castle.

1907. Selection and cross-breeding in relation to the inheritance of coat-pigments

and coat-patterns in rats and guinea-pigs. Carnegie Inst. Wash. Pub.

No. 70, 50 pp., 2 pis. M \< Dowell, E. C.

1914. Size inheritance in rabbits. Carnegie Inst. Wash. Pub. No. 196, 55 pp. Morgan, T. H.

1909. Breeding experiments with rats. Amer. Nat., 43, pp. 182-185.

1911. Moulting and change of color of coat in mice. Science, n. s., 34, p. 918.

1914. Heredity and sex. Columbia University Press. Nehhing, A.

1889. Ueber die Herkunft des haus-Meerschweinchen. Sitzungsb. der Naturf. Gesell. zu Berlin, pp. 1-4, 4 figs.

1893. Ueber Kreuzungen von Cavia aperia und Cavia cobaya. Sitzungsb. der

Naturf. Gesell. zu Berlin, No. 10, pp. 247-252.

1894. Kreuzungen von zahmen und wilden Meerschweinchen. Zool. Gart., 35,

pp. 1-6, 39-43, 74-78. Nilsson-Ehle, H.

1909. Kreuzungsuntersuchungen an Hafer und Weizen. Lunds Univer. Arsskrift. Onslow, H.

1915. A contribution to our knowledge of the chemistry of coat colors in animals

and of dominant and recessive whiteness. Proc. Roy. Soc, B-89, pp. 36-58. Pearson, K., E. Nettleship, and C. H. Usher.

1913. A monograph on albinism in man. Parts I and IV. Dulau and Co., London. Phillips, J. C.

1912. Size inheritance in ducks. Jour. Exp. Zool, 12, pp. 369-380.

1914. A further study of size inheritance in ducks with observations on the sex ratio

of hybrid birds. Ibid., 16, pp. 131-148.

PUNNETT, R. C.

1912. Inheritance of coat-color in rabbits. Jour, of Genet., 2, p. 221. Riddle, O.

1909. Our knowledge of melanin color formation and its bearing on the Mendelian description. Biol. Bull., 16, pp. 316-351. Ridgway, R.

1912. Color standards and nomenclature. Wash., iv 4- 44 pp., 53 pis. BoUiAB, I. B. J.

1909. Inheritance of color and of supernumerary mammae in guinea-pigs, with a note on the occurrence of a dwarf form. Reports to the Evol. Com. of the Roy. Soc. London, Report 5, pp. 51-79, 1 fig., 1 pi. Stirtevant, A. H.

1913. The Himalayan rabbit case, with some considerations on multiple allelomorphs. Thomas, O.

1901. On mammals obtained by Mr. Alphonse Roberts on the Rio Jordan, S. A., Minas Geraes. Ann. and Mag. Nat. Hist., London, 8, No. 49, pp. 526-539. Troiessart, E. L.

1904. Catalogus Mammalium. Berlin. Wactbboubb, ('.. EL

184s. A natural history of the mammalia. H. Bailliere, n, 500 pp., 21 pis.

W II.DKU, II H

1909. H ist ory of the human bodv. 573 pp.

W RIGHT, S.

1915 The albino ^i-ies of allelomorphs in guinea-pigs. Amer. Nat., 49, pp. 140-14S.

EXPLANATION OF PLATES.

PLATE 1.

Colored photographs of the fur of guinea-pigs, showing grades of dilution due to different combinations of the allelomorphs of albinism. The skin of each animal from which fur is shown was opened along the median ventral line and a compltte section across the middle of the body is shown in skins 1, 6, 9, 10, and 11. But in the other skins 2, 3, 4, 5, 7, and 8, only a section extending from the mid-dorsal to the mid-ventral line is shown. Figures 1 to 4 show agouti fur (AAEE) ; 5 to 8, non-agouti fur (aaEE) ; and 9 to 11 show fur in which the extension factor is wanting (ee). The uppermost row of skins are intense pigmented, all others are dilute of intensities, diminishing toward the bottom of the plate. Factors A and E are written as homozygous, though this ia not known in all cases.

Fig. 1. Black-red agouti C C AAEE

2. Dark sepia-yellow agouti Cd Cd AAEE

3. Dark sepia-cream agouti Cd Cr AAEE

4. Light sepia-cream agouti Cd Ca AAEE

5. Black C C aaEE

6. Dark sepiai Cd Cd aaEE

7. Dark sepia2 Cd Cr aaEE

8. Light sepia6 Cd Ca aaEE

9. Red C C aaee

10. Yellow4 Cd Cd aaee

1 1 . Creanio Cd Ca aaee (Cd Cr similar)

PLATE 2.

Photographs of guinea-pig skins, showing further grades of reduction in color due to albin- ism and its allelomorphs. Sections of skin extending entirely across the body are shown in all cases. The arrangement is similar to that of plate 1.

Fig. 12. Dark sepia-white agouti (red-eyed) Cr Cr AAEE

13. Light sepia-white agouti (red-eyed) Cr Ca AAEE

14. Albino (known to transmit agouti) Ca Ca AAEE

15. Dark sepia (red-eyed) Cr Cr aaEE

16. Light sepia (red-eyed) Cr Ca aaEE

17. Albino (transmits only non-agouti) Ca Ca aaEE

18. White (red-eyed) Cr Cr aaee (Cr Ca similar)

19. Albino (from yellow stock) Ca Ca aaee

PLATE 3.

Colored photographs of the skins of guinea-pigs.

Fig. 20. A half -grown guinea-pig of race C; color, pale cream. The eyes were bro wn-pigmented .

21. An Fi male hybrid whose mother was an albino of race B and whose father

was a pure cutleri. Compare figures 23 and 34.

22. An Fi male hybrid whose mother was a brown-eyed cream animal of race C

and whose father was a pure cutleri. Compare figures 20 and 23.

23. A pure cutleri male.

24. A pure cutleri female.

PLATE 4.

Colored photographs of the skins of F2 hybrids produced by crossing brown-eyed cream guinea-pigs with Cavia cutleri. Compare figures 20, 22, and 23. Fig. 25. Golden agouti.

26. Pale black (sepia).

27. Brown or chocolate.

28. Cinnamon.

29. Yellow.

30. Albino.

191

192 EXPLANATION OF PLATES.

PLATE 5.

Colored photographs of skins showing new color varieties of guinea-pigs. I i'.. 31. Silver cinnamon or red-eyed cinnamon.

32. Rcnl-and-pink-cyed black spotted with white.

33. Pink-eyed golden agouti spotted with red and with white, hence a "tri-

color." M. Albino with sooty fur and black pigmented extremities, similar to race B.

PLATE 6.

1 . mure of hybrid guinea-pigs and of their parent races, natural size, to show extent of variation. The longest and the shortest femur in each group of individuals is shown with 3 or 4 others of intermediate length placed between them. In the left half of the plate are shown the femurs of males, and in the right half the femurs of females. Top row, Cavia culleri. Second row, race B guinea-pigs. Third row, Fi hybrids pro- duced by the cross of C. cutleri cf X race B 9 • Fourth row, F2 hybrids from the same cross.

PLATE 7.

Fig. 35. A scale of grades used in describing the pattern of piebald rats. Rats like the pictures toward the left of the scale are known to fanciers as "hooded" ; the grade at the extreme right would be called "Irish" by fanciers.

36. Skins of a pair of rats and of their 9 young. One parent was an "Irish" rat, the

other "hooded." Four of the young are hooded, five are Irish. Hooded is re- cessive to Irish in crosses. The Irish parent in this case was a heterozygote. Note individual variation in each group of young.

37. A typical smooth-coated guinea-pig.

38. A rough-coated guinea-pig, well-rosetted, grade A.

39. A poorly-rosetted rough guinea-pig, grade C.

PLATE 1

Variations of intensity of coat pigments, due to albino allelomorphs, in agouti series (1-4), black series (5—8), and yellow series (9—11).

PLATE 2

Albinism and its non-yellow allelomorphs in agouti series (12 — 14), Mark series (15—17), and yellow series (18, 19).

PLATE 3

20

21

22

23

24

SmjpBB^^ ^^*"

Fig. 20, half-grown guinea-pig, race C. Figs. 23, 24, male and female Cavia cutleri, adult. Fig. 22, F\ hybrid, race C x Cavia cutleri, adult. Fig. 21, F1 hybrid, race B (Plate 5, Fig. 34) x Cavia cutleri, adult.

PLATE 4

25

26

27

F2 hybrids, race C x Cavia cutleri. Fig. 25, agouti; 26, black; 27, chocolate; 28, cinnamon; 29, yellow; 30, albino.

PLATE 5

31

32

34

Some new guinea-pig color varieties. Fig. 31, red-eyed cinnamon; 32, red-and-pink-eyed black spotted with white; 33, pink-eyed golden agouti spotted with red and with white; 34, albino with sooty fur and black pigmented extremities, similar to race B.

PLATE 6

Femurs of Cavia cutleri (male and female), of race B, and of their FL and F^ hybrids, showing complete range of variation in each. Natural size.

PLATE 7

— 1

+ 1

4-3

4-

36

35

37

38

39

Fig. 35, a scale of grades for piebald rats, n*"'*^^^^^^ young. Fig. 37, a smooth guinea-pig. Fig. 38, a well-rosettea rougn gu caps-s A. Fig. 39, a poorly rosetted rough guinea-pig, grade L-.

QH

431

C3925

Castle, William Ernest Studies of inheritance

BioMed.

PLEASE DO NOT REMOVE CARDS OR SLIPS FROM THIS POCKET

UNIVERSITY OF TORONTO LIBRARY

■