Nrw York AGRIcuLTURAL EXPERIMENT STATION. 561 
oped. With such varieties it cannot be maintained that self- 
sterility is due to imperfect development of the pollen. Most, if 
not all, of the grapes here listed as self-sterile have made a record 
of producing well-filled clusters of fruit in favorable seasons when 
so located that cross-pollination from some other variety could 
occur. ‘This is good evidence that the pistils are, as a rule, well 
developed. Plate L shows the results of preventing cross-pollina- 
tion with a practically self-sterile variety, the Eumelan. Not a 
fruit developed on all the covered clusters. At the close of the 
season only the stems remained, as shown in the illustration. 
Compare this with the clusters shown in Plate LIII which formed 
on self-sterile vines when the blossoms were exposed to cross-pol- 
hnation. Figure 1 shows a cluster of Herbert and Figure 2 a cluster 
of Barry. Compare Plate LI, Figure 1, which shows the three best, 
and with one exception, the only self-fertilized clusters of Brighton 
which have been obtained, with Figure 2, which shows a cluster of 
Brighton, which was open to cross-pollination. In view of these 
facts the second explanation for the self-sterility which is found 
among grapes cannot be accepted. 
By an examination of the parentage of the grapes which have 
been included in these investigations it is found that with the 
exception of the Eaton’ all of the thirty-three self-sterile varieties, 
Class 4, are hybrids, and of the twenty varieties listed in Class 3 
as perfectly self-fertile all but five are hybrids. It is well known 
that self-sterility is often found among plant hybrids. 
The evidence that self-sterility in general is due neither to de- 
fective pollen nor to defective pistils may be summarized as fol- 
lows: | 
1. Pollen is formed abundantly. 
2. Pollen retains its vitality till long after the pistil should be- 
come receptive. 

1Katon is a seedling of Concord, but whether from a seed produced from 
self-fertilized flower or from a flower open to cross-pollination, the records do 
not state. It is commonly classed as a pure lahrusea variety. 
36 
