INHERITANCE OF WAXY ENDOSPERM IN MAIZE. 95 
proximation of the expected 0.570 when circle 15 is self-pollinated. 
The remaining ears of the cross Dh 334 also seem to approximate the 
expected correlations. 
Ears Nos. 1906, 1904, and 1922 are all close approximations to the 
expected results of crossing circles 4, 7, 12, or 15 with circle 8 or 4 
or circle 7 with 16. Ears Nos. 1906 and 1922 are exact reciprocals, 
and as such agree as well as could be expected. The colored horny 
parent of these last-mentioned ears had a self-pollinated ear, No. 
1923, which could have been the result of the self-pollination of 
circles 4 or 7, so that ears Nos. 1906 and 1922 are probably the result 
of crossing circles 4 or 7 with circles 8 or 16. Ears Nos. 1909 and 
1924 are exact reciprocals, and with the self-pollinated ear from the 
colored horny parent, No. 1925, make a complete group. The self- 
pollinated colored horny parent is a very close approximation to the 
result expected when circle 13 is self-pollinated. The reciprocals, 
Nos. 1909 and 1924, are in fairly close agreement with each other 
and are close approximations to the expected result of crossing circle 
13 with either circle 14 or 16. 
SUMMARY OF THE CORRELATIONS BETWEEN ENDOSPERM 
TEXTURE AND ALEURONE COLOR. 
The preceding experiments were undertaken in the spring of 1910 
with the view to determining whether the previously observed 
coherence between endosperm texture and aleurone color could be 
referred to the 3-1-1-3 reduplication series of Bateson and Punnett. 
The intention was to obtain a number of individuals large enough 
to determine definitely the degree of reduplication, as well as to 
decide whether all plants gave the same reduplication ratio. 
The experiments have now been carried to the third generation, 
and the degree of correlation with 17,015 seeds has been found to be 
0.762+0.0057. This correlation is a very close approximation to 
the 0.766 expected if the reduplication was of the series 3-1-1-3. 
Some plants have been found, however, which gave reduplications 
other than 3-1-1-8, and it is possible that succeeding generations of 
these will throw additional light on this complicated subject. In 
some cases the departures could be readily explained by slightly 
modifying the theory proposed by Punnett as to the type of cell 
division, which resulted in the 3-1-1-3 ratio, but in other cases an 
extremely complicated theory of cell divisions would be required to 
give an adequate explanation of the observed ratios. 
To explain the aberrant cases by Morgan’s linkage hypothesis, we 
would be forced to believe that at least one of the characters was 
insecurely located on the chromosome, since with the widely different: 
correlations obtained it is necessary to assume that the mutation of 
the locus of at least one of the characters is of frequent occurrence. 
