34 
N. C. Agricultural Experiment Station 
pollination; and tlie second or check lot, in order to offset the effects of 
the covering in the first lot. The only difference in the resultant fruits 
must, then, be due to the origin of the pollen. 
The flower clusters were covered, when possible, before any of the 
flower buds had opened, and such as had opened were carefully re- 
Fig. 8.—A flower cluster of Munroe (dewberry), showing an abundance of strong, 
healthy stamens produced by a self-sterile variety of Rubus trivialis. 
moved, so as to forestall anv results due to activities of insects in their 
daily search for pollen and nectar. The hags covering the clusters of self- 
pollinated flowers were never lifted until comparisons of the inclosed 
fruits were made, while those bags which covered the cross-pollinated 
