Feb. 9,1924 
Selective Fertilization in Cotton 
339 
and possibly not a significant advantage in favor of the like pollen. 
The two kinds of pollen, therefore, do not differ appreciably in com¬ 
patibility. 
Tests made in sugar solution showed no important differences in the 
viability of the several pollens, and the very fact that the Egyptian and 
the Upland flowers, when pollinated with a mixture of pollen of the 
two types, yielded approximately equal mean percentages of hybrids 
indicates that the average viability of the two kinds of pollen was prac¬ 
tically the same. 9 
The case of selective fertilization here described is clearly not attrib¬ 
utable to differences in the viability or in the compatibility of the two 
kinds of pollen. The evidence is against the assumption that the phe¬ 
nomenon is one of selective survival of the homozygotes, at any stage 
from the formation of the zygote to the attainment of the adult state. 
So far as could be ascertained, the unlike pollen is not inferior to the 
like pollen in the rate of development of its tubes and in ability to unite 
with the female gametes. How, then, is the fact of selective fertilization 
in cotton to be explained ? 
The only hypothesis which seems to fit the observed facts is one pre¬ 
viously advanced (7, p. 48) that the presence of like pollen in some 
way prevents the germination or subsequent development of many of 
the unlike pollen grains when both kinds are present on the stigmas. 
That the inhibiting factor does not reside in the stigmas themselves 
when like pollen is absent seems clear from the fact that when applied 
separately the unlike pollen is not inferior to the like pollen in rapidity 
of development and ability to effect fertilization. 10 
It is conceivable, however, that the presence of pollen of the same 
type may induce a physiological reaction in the stigmas which makes 
them a relatively unfavorable medium for the germination or growth of 
pollen of a different type. The further assumption must be made that, 
in spite of this unfavorable condition, some of the unlike pollen grains 
are able to accomplish fertilization, possibly because they are more 
resistant, possibly because they happen to be so placed as to avoid the 
tracts of stigmatic tissue affected by contact with the like pollen. It 
would seem that such of the unlike pollen grains as succeed in avoiding 
or overcoming this obstacle develop their tubes as rapidly as do the 
pollen grains of the same type, and that there is no appreciable differ¬ 
ence in the readiness with which the two kinds of male gametes unite 
with the female gametes. 
The double-pollination experiments with maize carried out by Jones 
involved pairs of types which showed various degrees of genetic differ¬ 
ence, some being nearly and others very remotely related. A pronounced 
positive correlation was found to exist between the degree of selective 
fertilization in favor of like pollen and the genetic distinctness of the 
types, as indicated by the degree of heterosis manifested by the crosses 
between them. As Jones expresses it, “the handicap placed upon the 
foreign pollen is proportional to the germinal unlikeness” (4, p. 283). 
The experiments with cotton described in this paper involve only unre¬ 
lated types which differ profoundly in their morphological characters. 
* It is, of course, possible that, even with a difference in the pollen viability of two varieties, equal per¬ 
centages of hybrids might result from reciprocal double*pollination if selective fertilization in favor of 
the like pollen were more pronounced in the variety having the less viable pollen. 
10 This fact is also difficult to reconcile with the suggestion made by Jones (4, p. 284, 28s) that we are 
concerned here with a phenomenon analogous to the toxicity of foreign proteins. 
