82 BULLETIN 754, U. S. DEPARTMENT OF AGRICULTURE. 
This difference between the male and female gametes in the degree 
of reduplication could be brought about by a higher death rate pre- 
vailing in the pollen grains carrying white horny and colored waxy, 
the combinations that came from opposite parents. It might-be urged 
that these combinations reacted in an unfavorable way, lessening the 
life of the pollen grain or so reducing it in vigor as to have a larger 
proportion of the ovaries fertilized by the parental, white waxy and 
colored horny combinations. 
It would be necessary that the death rate be alike for the two com- 
binations, colored waxy and white horny, or the percentages of 
white and waxy would be found deficient. This is in itself an argu- 
ment against this explanation, since it seems improbable that CY and 
WZ should happen to be equally weak. The percentages of white to 
waxy seeds for the three ears are sufficiently close approximations to 
the expected to indicate that the two combinations were retarded 
about equally. 
Ear No. 1764 had 50.2 per cent waxy and 50.4 per cent white seeds, 
certainly very close to the expected 50 per cent. The reciprocal ear, 
No. 1799, had 51.4 per cent waxy and 53.2 per cent white seeds. The 
differences between white and waxy percentages of these two recipro- 
cals are not significant. The percentage of waxy seeds on the self- 
pollinated ear (No. 1800) is almost exactly the expected percentage, 
while the percentage of white seeds is in excess of the expected by 
more than three times the probable error. That the excess of white 
seeds on this ear is not brought about by a failure of the pollen grains 
carrying colored waxy combinations is demonstrated by the percent- 
age of waxy seeds, which is normal. In other words, if the high 
percentage of white seeds is due to a high mortality of the pollen 
grains carrying colored aleurone, this mortality must have been equal 
in the pollen grains carrying colored-waxy and colored-horny com- 
binations. ; 
The expected grouping compared with that observed for the 755 
seeds is as follows: 
Number of seeds. 
Class, 
CH. Cx WH. Wx. 
pected tes sie Ser psee ters teptetars ciclo wisteteclete 'eision Pe sasoosacs 484 82.5 82.5 106 
COLORECT Z10 Waa SR See A Sea One EERE il OAc 476 64 88 127 
As can be seen, there was a failure of both colored horny and col- 
ored waxy gametes, but greater in the colored waxy than in the col- 
ored horny, while a corresponding increase is noted in both the white 
horny and white waxy combinations, though the increase is much 
larger in the latter case. The figures would seem to indicate that the 
