The Inheritance of Quantitative Characters in Maize 13 
therefore represented by the presence of a single factor R and 
behaved as it should when it met its absence r. But since the 
cross between the two red varieties gave the dihybrid ratio in P 2 , 
one must suppose that, tho alike in appearance, they were not 
allelomorphic to each other. One was R^r.r, and the other 
i\)\R 2 R 2 . Each bred true to red within its variety, but when 
crossed, each factor met its absence and an ¥ x was produced 
heterozygous in both. The resulting F 2 generation for this 
reason was exactly the same as if a red R 1 R X R 2 R.^ had been 
crossed with a white >y W* 2 - 
Just imagine such a case when more pairs of factors are in 
action. If units A t A 2 A 3 a 4 met units a 1 a 2 a s A 4 in hybridization, 
there is again the apparent paradox of two individuals which 
have bred true for a particular character, giving pure recessives in 
F 2 w hen crossed. But this time the zygote without the character 
occurs only once in 256 progeny. When one considers the rarity 
with which dominants or recessives pure for all factors are 
obtained when three factors or more interact, he can hardly avoid 
the suspicion that here is a perfectly logical way of accounting 
for many cases of so-called atavism that are not explained by 
the interaction of two dominant factors. No definite cases can 
be pointed out, but it is not uncommon for a florist to have a 
Dew type appear with regularity tho with extreme rarity for many 
generations after a cross. No doubt proper analyses of con 
trolled cultures would elicit simpler explanations of most of 
these cases, but some of them surely must be referred to the 
class of phenomena we have just discussed. It is quite within 
the range of possibility that some of de Tries' Oenothera mutants 
have originated in this manner. 
The colored zygotes, in the glume color of oats and head color 
and seed color of wheat investigated by Nilsson-Ehle. have been 
treated as if they were exactly alike in appearance. This was 
not always the case. The color produced by additional factors 
was often somewhat cumulative, particularly so in the case of 
the colored wheat heads. Where the two dominant factors were 
present there was usually more color developed than where only 
one was present. The classes could not be distinguished visually, 
but the general tendency was cumulative. This fact is nicely 
illustrated by the F 2 segregates of the maize cross in which yellow 
endosperm behaved as a dihybrid. All of the classes having 
different gametic constitutions vary in the intensity of their 
yellow color. The yellow becomes lighter in shade in the follow- 
ing order: Y^Y^Y 2 Y 2i y l u^Y 2 Y 2 or YJ? x Y 2 y 2 . Y x y t or Y 2 y 2 , and 
These facts open up two new and important phases of Men- 
delian inheritance. The first is the possibility of having new 
