STERILITIES OF WILD AND CULTIVATED POTATOES. 23 
the same time with the same media and in the same moist chamber 
gave the excellent germination characteristic of them. Flowers, 64 
in number, representing 16 different plants, were inbred by pollina- 
ting with pollen from the same flower cluster. No evidence of seed- 
ball production was found. The results obtained from the germina- 
tion and pollination studies indicate that a high degree of male 
sterility, possibly complete, exists in these hybrids. 
When these hybrid plants were crossed with parent plants of 
Solarium fendleri failure resulted in every case, but fruit containing 
no seed was produced by several of the hybrids when S. chacoense 
was used as the pollen parent. 
STERILITY FROM INCOMPATIBILITY. 
INCOMPATIBILITIES IN CULTIVATED VARIETIES. 
The data presented for cultivated varieties of potato by Taylor (26) 
and the discussions by Stuart (24) raise the question whether certain 
failures to set fruit, to cross, and to self-pollinate involve differences 
in compatibilities of fertilization or whether they merely involve impo- 
tence of either the pistil or the pollen. To be certain that incompati- 
bilities exist one must be sure that the pollen and the pistils concerned 
are sufficiently potent to function in certain relations and that the 
pollinations are properly made. As pointed out in preceding pages, the 
cultivated varieties which completely fail to produce fruit to pollen 
known to be viable are indeed few, if any. It is clear that in respect 
to crosses between varieties the failures involve most frequently a par- 
ent which is impotent as a pollen parent or a parent somewhat feeble as 
a seed producer. When these grades of impotence are considered there 
appears to be no evidence of discriminating fertilizations or incompati- 
bilities. The successes in crosses between varieties that are in class 
A, in regard to seed production, and those that are in class 1 in regard 
to potency of pollen, average about 40 per cent, a rather high average 
for plants which have such a decided tendency to the ready abscission 
of flowers. 
The data from controlled pollinations are not sufficient to judge 
the self-compatibility of the varieties known to be both good seed 
producers and productive of viable pollen (classes A and 1), but it is 
such varieties and seedlings which often produce seed balls in abun- 
dance to open field pollination. 
INCOMPATIBILITIES IN WILD SPECIES. 
During the seasons of 1921 and 1922 a large number of hand pollina- 
tions were made for the purpose of determining whether incompati- 
bilities exist in certain of the wild species and also to test cross 
compatibility between these species and between these and cultivated 
varieties. The work was started in 1921 with plants of three wild 
species, Solarium chacoense, S. fendleri, and S. jamesii, for all of 
which the pollen produced is highly potent. Most of the work was 
performed in the greenhouse, though a few plants in the field were 
used. The plants of S. chacoense and S. jamesii were grown from 
tubers. It is not known whether the seed tubers of the species first 
named were produced by one or by several plants. Those of the 
species last named were from a composite lot of several hills grown 
in the field the previous season. The plants of S. fendleri were grown 
from seed. 
