Vol. 6, 1920 
GENETICS: S. WRIGHT 
321 
tions covering almost the entire range from solid color to solid white can 
be found in each line. Even after twenty generations of inbreeding, a 
guinea-pig with 20% of white in the coat may have a litter mate with as 
much as 90% of white. 
As a supplement to more direct attempts to determine the cause of these 
variations in the pattern, it seemed desirable to find the relative importance 
of heredity and environmental factors. 
The pattern of tricolors is really a combination of two patterns which 
are, in the main, inherited independently of each other, although obviously 
associated in development. These are the piebald pattern, consisting 
of colored spots on a white ground, and the tortoiseshell pattern consist- 
ing of black spots on a yellow ground. In a tricolor, a certain pattern of 
color and white is determined irrespective of the quality of the color. An- 
other group of factors determines which spots shall be black (or one of the 
modifications of black, such as agouti, sepia or brown) and which shall 
be yellow or red. 
A cross between a homozygous solid black and a tricolor of one of the 
inbred families produces black young with no yellow spots and only a 
very small amount of white, if any. On back-crossing with the tricolor 
stock, four kinds of young are produced in approximately equal num- 
bers, viz., solid black, black and yellow tortoiseshells, black and white 
piebalds, and black, yellow and white tricolors. As in the first generation, 
there is apt to be a small amount of white in the first two classes. The 
nearly solid black type has been back-crossed for seven generations with 
pure tricolor stock, producing young with as much as 255 / 256 of the blood 
of the latter stock, with the same results as in the first back-cross, four 
classes of young in equal numbers. It is clear that two independent 
Mendelian factors are involved. One of the factors is completely dominant 
over that for tortoiseshell, the other is nearly but not quite fully dominant 
over that for piebald. In the stocks considered in the present paper both 
of these dominant factors were absent, leaving all of the animals tricolor 
except for the occasional somatic bicolors and black-eyed whites to be ex- 
pected in a stock with a high average amount of white in the coat. The 
hereditary factors to be considered here, those by which one tricolor differs 
from another, are of the kind frequently known as modifiers. Crosses 
between inbred families at opposite extremes in amount of white have 
resulted in offspring which are almost exactly intermediate on the average. 
In the second generation, there is only a litlki more variability than in the 
first; much less than would be present if there were segregation of a single 
factor. 
The minor as well as the major variations of the piebald and tortoise- 
shell patterns appear to be largely independent of each other, although at 
least one factor has been discovered which affects both, viz., sex. The 
males of all stocks have on the average some 7% more color than the females 
