New York AcricutturaL Experiment Strarion. 3865 
which had held the cluster which furnished the pollen. After the 
bag had been closed and labeled, it was shaken so as to agitate the 
loose pollen which it might contain and possibly assist in more 
thoroughly distributing it among the open flowers of the cluster 
which was to be pollinated. 
In the Penn Yan tests and in those at the Station, in addition 
to the treatment above outlined, the open flowers of the cluster 
which were selected to supply the pollen were brushed over the 
open flowers which were to be cross-pollinated before fastening the 
two clusters together. With most varieties the interlocking of the 
branches of the two clusters was sufficient to hold the loose cluster 
in place, but in some instances it was necessary to tie the two to- 
gether. Judging from their appearance when the bags were 
opened after the blooming season had passed, the flowers of the 
detached clusters which were not open when the bag was finally 
closed and labeled, generally failed to open later. It appears, 
therefore, that the pollen which effected the fertilization of the 
flowers as shown by the later development of fruit, must have come 
almost entirely from blossoms which were open when the hand 
pollinating was done. With some of the varieties which were arti- 
ficially cross-pollinated after the manner above described, a con- 
siderable portion of the flowers were not opened when the hand 
pollination was done, and probably did not open for a period of 
from 24 to 48 hours thereafter. Had the bags been reopened on 
each of the two days following the one upon which the hand pol- 
linating was done, and the clusters brushed again with freshly 
opened blossoms, the result might have given a more accurate 1n- 
dication of the kind of clusters which such varieties may be ex- 
pected to produce when they stand adjacent to each other, and the 
process of cross-pollination goes on freely during the entire bloom- 
ing season. 
On the other hand it should be observed in this connection that, 
even with varieties which had not come into full bloom when the 
hand pollinating was done, some perfectly filled clusters were 
obtained, showing that the percentage of blossoms which became 
