2 BULLETIN 754, U. S. DEPARTMENT OF AGRICULTURE. 
In subsequent breeding work this waxy endosperm was found to 
be alternative to horny endosperm, behaving as a recessive character 
when crossed with American flint and dent varieties. The new 
endosperm character has proved to be adapted for investigation of 
some of the problems of heredity in maize. The inheritance is defi- 
nitely alternative, no blending or intermediate stages having been 
found. This absence of intergradations between the new waxy 
endosperm and the other endosperm textures simplifies the classifi- 
cation of material and renders the numerical relations of this char- 
acter more definite and hence of more significance than those ob- 
served with other Mendelian characters of maize thus far studied 
where intermediates are comparatively common. 
In the first crosses that were made the behavior of the waxy 
endosperm was very similar to that of a Mendelian unit character, 
. the approximation being so close that the deviations were at first con- 
sidered accidental. But with the progress of the investigation and 
the accumulation of larger numbers of individuals it became apparent 
that the deviations from the theoretical Mendelian ratios, though 
never large, were too definite to be ascribed to chance. It further 
developed that a correlation existed between the waxy texture of the 
endosperm and the red and blue colors in the aleurone cells of the 
Chinese maize in crosses with other varieties. 
This bulletin reports the results of a series of crosses between the 
Chinese variety and an African pop corn made for the further study 
of the correlation between the endosperm texture and the color of 
the aleurone. In these experiments large numbers of seeds were 
classified with respect to the endosperm and aleurone characters. 
While the results are in the main similar to those previously reported, 
the experiments now comprise such large numbers of individuals as 
to place the deviations from the expected ratios beyond question and 
also to establish more definitely the existence of the correlation be- 
tween endosperm texture and aleurone color. 
The subject became somewhat involved when it was found that the 
aleurone color had to be resolved into two factors, and it became 
necessary to analyze the behavior of correlated characters where the 
correlations are between factors. In attempting to understand the 
correlations of the characters, both the theory of reduplication (1) 
and the linkage theory (11) have been kept in mind. 
The experiments have yielded a body of evidence that should be 
of value in testing the general applicability of current theories of the 
segregation or alternative inheritance of characters and also in test- 
ing explanations that may be advanced in the future. To be of 
value in this connection, it is necessary that the data should be placed 
