86 BULLETIN 754, U. S. DEPARTMENT OF AGRICULTURE. 
The assumption can then be made that the white horny plant is 
producing gametes (Hr, Chr, cHr, and chr, with a correlation be- 
tween C and H. Such a plant, if self-pollinated, would produce an 
all-white ear with 3 horny seeds to 1 waxy. 
The colored waxy plant might be making gametes CRA Crh. A 
cross between a colored waxy plant of this nature and a white horny 
plant such as has been described would result in an ear with four 
classes of seeds: CH, 7; CX,5; WH,1; WX, 38; correlation, 0.615. 
In this case, however, the percentage of white seeds is but 25, and 
_ the correlation is 0.615, while in the two ears reported by Collins, one 
had 380.1 per cent white with a correlation of 0.3730.057, and the 
other had 48 per cent white and a correlation of 0.470.056. It is 
at once apparent that neither of these ears could be looked upon as 
approximations of the above zygotic arrangement. 
From the second ear of the hybrid Dh 2387, No. 1130, from which 
the six classes of seeds were planted, three self-pollinated ears from 
colored horny plants were secured that had four classes of seeds 
(Table XX XIX). These three ears had a correlation of 0.7250.017, 
slightly lower than the expected 0.766. None of the three ears devi- 
ated from the expected in excess of three times the probable error, 
though all three were below the expected. As the result of crossing 
plants grown from white waxy seeds with plants grown from colored 
horny seeds of ear No. 1130, six ears were obtained by using the 
white waxy plants as the female parents, and five ears were obtained 
by using the white waxy plants as the male parents. These 11 ears 
are also shown in Table XX XIX. 
The 6-ear group had a correlation of 0.809+0.01, which is a very 
close approximation of the expected 0.8. None of the individual ears 
of this group deviated from the expected correlation by as much as 
three times the probable error. 
The ears of the 5-ear group had with one exception lower correla- 
tions than the lowest obtained in the 6-ear group. The correlation _ 
for the five ears was 0.76+0.01, which differs from the 0.809 correla- 
tion obtained in the reciprocal ears by 0.049-+-0.015. None of the 
ears in the 5-ear group deviated by as much as three times the prob- 
able error from the expected correlation. These two groups, together 
with the three self-pollinated ears from the colored horny plants, are 
certainly very close approximations of the expected results where the 
coupling is of the form 3-1-1-3. The individual ears exhibit a re- 
markably uniform grouping of the four classes of seeds. 
Crosses were also made between plants from the colored waxy seeds 
and plants from the white horny seeds, which should result in ears 
with no correlation. 
