Vol. 6, 1920 
GENETICS: S. WRIGHT 
323 
family 39 during this period were a 11 descended from a single mating in 
the fourth generation of inbreeding, and with a few exceptions, from one 
in the eighth. Family 35 was descended wholly from one mating in the 
seventh generation and largely from a single one in the twelfth. Family 
13 was descended wholly from one mating in the second generation and 
largely from one in the seventh. 
It will be seen that each family varies almost through the entire range, 
but that the mean values are markedly different. Family 39 has only 
about 20% white on the average, family 13 at the opposite extreme has 
about 85%, while family 35 has its average near 50%. On attempting to 
/.O 
.9 
.S 
.7 
.6 
.5 
,4 
.3 
.2 
J 
it- 
/ 
/ 
/ / 
/ / 
// 
\ 
\ 
/ 
■A 
— -f- 
\ 
\ 
\ 
1 
1 
\ 
\ 
/ 
— \ 
\ 
-t 
1 
-h- 
\ 
\ 
-4- 
1 
\ 
...1 
0 5 /5 25 35 45 55 65 75 85 9S 100 
FIG. 2 
The solid line shows the average distance between quartile and median in dis- 
tribution curves from different inbred families in which the midpoint between 
quartile and median falls at different percentages of white. Based on 68 quartiles. 
The average quartile distance near 50 per cent white is taken as the unit. The bro- 
ken line shows the expected quartile distances according to the hypothesis discussed 
in the text. 
compare the variability of the families, a difficulty at once becomes ap- 
parent. The distribution curve of family 35 is practically symmetrical, 
but those of families 39 and 13 are markedly skew in opposite senses. 
This is not because of overstepping of the physiological limits, as there 
were no solid colored guinea-pigs in family 39, and only 8% black-eyed 
white in family 13. The curves appear to be compressed at the ends of 
the range, indicating that a unit of variation in these regions means more 
than in the middle of^the range. It is especially difficult, so to speak, to 
remove the last bit of color, or the last bit of white. This phenomenon is 
shown by all of the stocks in which there is not an approximately^equal 
average amount of color and white. 
