POTATO BREEDING AND SELECTION. 21 
The results secured from our own studies as presented in Table IV 
are not entirely in accord with Salaman's deductions. For example, 
in a population of 1,425 seedlings from a cross between Irish Cobbler 
and Irish Seedling, the first parent having a creamy white skin and 
purplish tinged sprouts and the latter with flesh-tinted skin and 
purplish sprouts, color was absent in 70.2 per cent of the tubers. 
Of those showing color, 36 were mottled with white, 229 were flesh, 
104 were red, 55 were purple, and 1 was violet-black. In another 
instance, out of a population of 870 seedlings of Irish Cobbler crossed 
with Keeper, color was absent in 69.7 per cent of the tubers. The 
pollen parent, Keeper, being a red-skinned variety, it would seem 
that if white were recessive a larger proportion of the ¥ t generation 
should have shown color. Almost identical figures were secured 
when Extra Early Eureka and Keeper were crossed. In this case, 
in a population of 680, color was absent in 70.6 per cent, or a difference 
of but nine-tenths of 1 per cent between the two crosses. In view of 
the fact that the writer regards Irish Cobbler and Extra Early 
Eureka as one and the same variety, the parentage of the two crosses 
is thus identical. When it is considered that the data for the two 
crosses were secured and tabulated independently of each other, 
the uniformity of the data is all the more remarkable. When the 
Irish Cobbler is crossed with a Chilean seedling, a different set of 
data is obtained. In a population of 214, only 34.1 per cent of the 
tubers were free from color. 
When white-skinned and white-sprouted varieties were crossed 
with red-skinned varieties, the proportion of white-skinned seed- 
lings was larger, as is shown in the following crosses: Green Mountain 
X Keeper produced 88 seedlings, of which 79.5 percent were devoid 
of color; of those which did show color, 11 were mottled, 6 flesh, 1 
russet, and 1 red. Gold Coin X Keeper produced 322 seedlings, of 
which 79.8 per cent were devoid of color; and of those which showed 
color 28 were mottled, 32 were flesh, and 5 were red. The remark- 
able similarity of color expression of these two crosses is again 
accounted for by the fact that the varieties Green Mountain and 
Gold Coin are regarded as identical or so nearly so as to be practically 
indistinguishable from each other. 
While numerous other examples might be cited, it is believed that 
sufficient data have been presented to justify the assertion that 
white is not a recessive character in the seedlings of the crosses just 
named. 
POTATO IMPROVEMENT BY SELECTION. 
The improvement of the potato by selection alone restricts the 
operator to such variations as may occur within the variety. Fortu- 
nately, this field of endeavor is not such a limited one as it might at 
first appear to a layman, since considerable variation already exists 
