"90 MASS. EXPERIMENT STATION BULLETIN 355 



are harvested, it is too late to sow buckwheat and get much growth. Barley was 

 tried one year but the catch was spotty and growth poor. 



The use of a light one-horse disc harrow for cultivation is not satisfactory on 

 the type of soil at Amherst. Where cultivating is to be done with a tractor har- 

 row, rows spaced eight feet are too close; rows spaced ten feet are satisfactory. 



The summer of 1938 was very rainy. Except for a few spots where the soil is 

 too wet in a rainy season, blueberry plants made very good growth. Rains inter- 

 fered with the picking considerably, especialh- about the time that Rubel began 

 to ripen. Many berries dropped before they could be picked, making yield records 

 less reliable than in years of normal rainfall. 



Studies of the relation between size of the bush and yield showed a high cor- 

 relation between maximum spread (measured in one direction) and yield but a 

 low correlation between maximum height of bush and yield. Since the latter is 

 ' equivalent to correlating yield with the longest branch on the bush, the results 

 are not surprising. 



Experiments with honey bees as pollinators for blueberries were continued 

 in cooperation with Dr. Frank Shaw of the Department of Entomology. Three 

 aster-cloth tents were built, each covering two blueberry bushes of different 

 varieties, and a nucleus hive of bees was placed under each tent. The set pro- 

 duced by the bees is compared in the following table with that of open-pollinated 

 flowers and of branches covered with aster cloth to exclude insects. Varieties 

 placed together in each tent are indicated in the table. 



In Tents Open Covered to 



Variety Bee Pollinated Pollinated Exclude Insects 



% Set % Set 



Number Number 



of of 



Clusters Berries 



Cabot 94.8 75.8 20 10 



Pioneer 80.7 68.8 18 10 



Pioneer 94 . 8 23 5 



Jersey 94.9 77.4 13 3 



Pioneer 87.0 U 



Rubel 93.8 89.4 30 9 



The figures show that honey bees have again proved their value as pollinators 

 of blueberries. 



The hay mulch used in blueberry plot B seems to be very satisfactory except 

 that it formed excellent cover for mice. Several plants were slightly damaged by 

 mice during the winter of 1937-38. Anyone using a hay mulch should carry on a 

 vigorous poisoning campaign in the fall. 



Premature Dropping of the Mcintosh Apple. (L. Southwick.) The September 

 hurricane disrupted the experimental work, although valuable data were obtained 

 prior to the storm. During the latter part of the season, daily growth measure- 

 ments were made with an improvised circumeter, for the purpose of correlating 

 growth rate with abscission initiation. Atmometers were employed for the first 

 time this year to measure the evaporating power of the air at frequent intervals. 

 The possible effect of this factor upon preharvest dropping will be studied. 



Analysis of data up to this year warrant the following tentative conclusions 

 regarding premature Mcintosh drop. The evidence points to a complexity of 

 interacting causes. 



Premature dropping seems to increase in intensity from Massachusetts to the 

 southern limits of the Mcintosh belt; it does not decrease so markedly as the 



