•*€>

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

■