124 Journal of Agricultural Research voi. xvii, No. 3 



occurred on May 21 and 25, as well as the interference with insect flight, 

 pollination would appear to be uncertain because pollen was not avail- 

 able for dissemination a large part of the time. This year, therefore, it 

 appears that pollination was uncertain instead of fertilization, as was 

 the case the year before. At any rate, during these two seasons the 

 temperature at bloom was very different. In 191 7 rain, high winds, low 

 temperatures, and even frost, were prevalent during bloom, and at the 

 close of bloom there were nearly 3 days of cool, rainy weather which 

 came at a critical time during tube growth. In addition to this, frequent 

 rains and a relatively low temperature at the latter part of the lo-day 

 period following bloom supplemented the retarding effect of the 3-day 

 rainy period. The wind on May 20, 21, 22, and 26 was strong enough to 

 interfere with the work of bees. Both pollination and fertilization were 

 uncertain this year. 



In contrast to the slight set of these three seasons there was a good set 

 in 1912, 1913, and 1918, and a heavy set in 1914. It now remains to be 

 seen whether there were conditions at bloom these seasons which differ 

 markedly, as far as the influence on pollination and fertilization is 

 concerned, from the others. In 191 2 the temperature was relatively 

 high, except for three days, during the entire period. The rains were 

 slight at bloom. Also, in 1913 the temperature was within the range 

 of fast tube growth a good part of the time and rains were unusually 

 scant at bloom. The unusually high temperature in 19 14 is in marked 

 contrast to the low temperature the following year, and in the absence 

 of heavy rains there was the greatest setting of fruit as well as the heaviest 

 crop of all season included. The high temperature at the beginning of 

 bloom in 1918 gradually fell toward the end and there was a frost the 

 night of May 12. The rains were not prolonged during bloom, but the 

 heavy rain of May 9 delayed pollination in the later blooming varieties. 

 The warm period following bloom, however, counterbalances the cooler 

 4-day period at the end of bloom, so that the rate of tube growth was in 

 general increased. The setting of fruit was sufficient for a good crop this 

 season. 



It will be seen from this brief analysis that there are conditions each 

 season v/hich can be correlated with the set of fruit. With a light set 

 it is impossible to get a heavy crop. As early as the 5- or 6-week period 

 the possibilities of a crop are determined. 



SUMMARY 



(i) Unfavorable weather at blooming time may completely prevent 

 the setting of fruit in the plum, even though there be a full bloom, A 

 study of the manner in which weather affects the processes at bloom 

 shows that rain and low temperatures are the most important factors, 

 although strong winds when prolonged are also important. 



