24 
V. C. Agricultural Experiment Station 
ing of the total number of apparently perfect and the total number of 
defective pollen grains that were found in ten separate fields under the 
microscope. From these figures the percentages of good and of defect¬ 
ive pollen grains were calculated. 
This method of procedure was repeated for each variety, and the con¬ 
dition of its pollen determined. It must be stated that pollen grains 
which appeared normal regardless of actual viability were classed as 
good grains, and only small, hollow, and shriveled grains were counted 
as defective. With some varieties this distinction becomes somewhat diffi¬ 
cult because of the production of a great many angular grains which 
closely resemble the defective ones. 
£ I)rpm the tables we learn some interesting facts: first, that the self- 
varieties of Rubus villosus produce a comparatively high (over 
90%) percentage of normal pollen grains, and that Premo, the only 
lineal descendant grown here which does not produce such a quantity of 
normal pollen, is self-sterile; second, that the other self-sterile varieties, 
all of which, with the exception of the hybrids, are representative 
varieties of Rubus trivialis, vary considerably in the amount of defective 
pollen that they produce. Such varieties as Limekiln, Munroe, Rogers, 
and White produce a very small percentage of defective pollen, less than 
8 per cent. In this respect they correspond very favorably with those 
varieties that are of Villosus origin and which are self-fertile. Other 
varieties of this same species, such as Chestnut, Manatee, and San 
Jacinto, show a very much larger per cent of defective pollen. With 
the exception of Manatee, these varieties produce approximately from 
50 to 80 per cent of normal pollen. Manatee falls far below this class 
in that it sometimes produces as high as sixty and three-tenths (60.3) per 
cent and sometimes as high as ninety-four and one-tenth (94.1) per cent 
of defective pollen. All of these varieties are self-sterile. Cox and Ruth, 
two self-fertile varieties of hybrid origin, while perfectly self-fertile, 
produce as high as approximately twenty-two and three-tenths (22.3) 
per cent and twenty-five and four-tenths (25.4) per cent of defective 
pollen grains; while Haupt, a hybrid, which also produces an equally 
large amount of perfect pollen, is self-sterile. 
From these results it becomes quite evident that the quantity or the 
number of defective pollen grains in the flowers of our self-sterile 
varieties of dewberries and hybrids plays no fundamental role in the 
cause of self-sterility of such varieties. The production of a compara¬ 
tively high or a comparatively low percentage of defective pollen grains, 
however, becomes of vital importance to the commercial fruit grower 
in the selection of pollinators for the berry plantation. For, other con¬ 
ditions being equal, the greater the per cent of pollen produced, the 
greater are the chances for perfect cross-pollination and a “bumper” 
crop. The subject of the commercial importance of varieties which 
habitually produce a high per cent of perfect pollen will be discussed 
later under the subject of pollinators for the plantation. 
