Feb. 9,1924 
Selective Fertilization in Cotton 
333 
both experiments, although the differences were significant in only two 
of the four comparisons, that of the Pima flowers in experiment 1 and 
that of the Upland (Acala) flowers in experiment 2. The differences in 
these cases were, respectively, 6.3 and 8.4 times the probable error of 
the difference. The probable explanation is that in some of the flowers 
the surface of the stigmas was so well covered by the pollen applied first 
that many of the grains of the other kind of pollen did not come into 
contact with the stigmatic surface and hence were unable to effect fer¬ 
tilization. It should be noted that the highest percentage of hybrids 
resulting from application of the unlike pollen first (Acala with Pima 4 - 
Acala in experiment 2) is significantly lower than the 50 per cent 
expected if there had been no selective fertilization, the departure having 
been 7.8 ±1.76. 
Table II. —Percentages of hybrids resulting from double pollination with (a) the like 
pollen applied first , (b) the unlike pollen applied first 
Sequence of application of the like and unlike pollens. 
Number of— 
Percentage 
Plants. 
Hybrids. 
of hybrids. 
EXPERIMENT I 
(a) Pima with Pima-J-Lone Star. 
741 
200 
27. 0±I. IO 
(b) Pima with Lone Star-f-Pima. 
Difference in favor of unlike pollen applied first.... 
640 
242 
37. 8±i. 29 
10. 8±i. 70 
21. 2 ± 1. 26 
(a) Lone Star with Lone Star+Pima. 
481 
102 
(b) Lone Star with Pima+Lone Star. 
Difference in favor of unlike pollen applied first... . 
271 
64 
23. 6±i. 74 
2. 4±2. 15 
EXPERIMENT 2 
(a) Pima with Pima+Acala. 
627 
IOO 
15- 9 ±°- 9 8 
(b) Pima with Acala+Pima. 
Difference in favor of unlike pollen applied first.... 
341 
63 
18. 5 ±1.42 
2. 6±i. 73 
(a) Acala with Acala+Pima. 
309 
69 
22. 3 ± 1. 59 
(6) Acala with Pima+Acala. 
358 
151 
42. 2 zfc 1. 70 
Difference in favor of unlike pollen applied first.... 
19 * 9 ±2. 37 
SELECTIVE SURVIVAL AS A FACTOR AFFECTING THE PERCENTAGES 
OF HYBRIDS 
The percentage of hybrids in an adult population from double-polli¬ 
nated flowers would not represent the actual degree of selective fertiliza¬ 
tion if there had been selective survival in favor of either the homozygotes 
or the heterozygates. It is possible, even, that what would appear to be 
a case of selective fertilization in favor of like pollen might be, in reality, 
merely a result of pronounced selective survival of the homozygotes. It 
is important, therefore, to examine critically the evidence as to whether 
there had been selective survival in the experiments under discussion. 
Whether selective survival took place during germination and in the 
early stages of seedling growth is the first question to be examined. Had 
this been the case it seems probable that the heterozygotes rather than 
the homozygotes would have benefitted, since the Pima-Upland hybrid 
plants soon become conspicuously larger and more vigorous than the Pima 
or the Upland plants resulting from fertilization by like pollen. If there 
