262 
GENETICS: W. E. CASTLE 
agency is genetic or purely physiological (like that which produces the 
large size of Fi) is at present uncertain. If it should prove to be genetic, 
it may conceivably be a MendeHzing factor of greater or less stabiHty, 
like those which affect the shape of fruits in tomatoes, peppers, and 
squashes (Gross, Emerson). But this will remain a matter of uncer- 
tainty until further evidence is forthcoming. Aside from these two 
individuals with skeletons of unusual form, everything so far indicates 
that the inheritance of skeletal dimensions is completely blending with 
physiological increase of size in Fi, this being however not a matter of 
heredity. 
We have already anticipated the result of the Cutleri X race B cross 
as regards skull- width. See Table 2. Here also the parent races 
scarcely overlap in range of variation. The mean of Fi slightly exceeds 
that of race B. F2 is intermediate between the parent races and scarcely 
more variable than Fi, with no very aberrant individuals. 
In the lower portion of Tables 1-3 will be found the bone-measure- 
ments for the Fi and F2 generations produced by a cross between race 
B females and the Arequipa male, 1002, still living and so not available 
for bone measurements. 
But bone measurements of a pair of animals of the Arequipa race are 
slightly less than the corresponding average measurements of the Fi 
animals. Hence it is probable that a considerable physiological increase 
of size occurs in Fi in this cross, as in the Cutleri cross, causing the pro- 
duction of an Fi larger than either parent race. Comparison of the 
growth curves of Fi and F2 (figs. 3 and 4) and of their respective bone 
measurements confirms this idea. The mean of F2 is in all measure- 
ments less than that of Fi, the difference amounting to one or two 
millimeters. This indicates that the increase of size due to crossing, 
seen in Fi, does not persist in F2. 
But the important question theoretically is whether there is evidence 
of the segregation or recombination in F2 of distinct genetic size factors. 
If such segregation occurs, it might be expected to show itself either 
(1) in increased amplitude of the variation, provided that multiple fac- 
tors occur which lack dominance but segregate and recombine inde- 
pendently of each other, or (2) in the formation of multi-modal F2 
variation curves with the production of isolated aberrant individuals, 
provided that the genetic factors concerned are few in nxmaber or show 
dominance. 
As regards the first possibility, the F2 animals do show somewhat 
greater variability than the FiS, though the difference as measured by 
the standard deviation (o-) is not great. In the case of Cutleri X race 
