Self-sterility in Dewberries and Blackberries 
23 
of the variety Premo, which presumably is a direct descendant from the 
self-fertile species Rubus villosus. Third, that true hybrids between self- 
fertile species of Rubus (examples of which are the varieties Cox, Ruth, 
and Wilson) may retain complete self-fertility, but sometimes only 
partial self-fertility, as in the case of Rathbun. Fourth, that true hy¬ 
brids of Rubus species with self-sterile species may be self-sterile, 
examples of which are Haupt, Spalding, Sorsby, and possibly McDonald. 
These hybrids are not self-sterile, however, because of the fact that 
they are of hybrid origin, but, rather, because sterility was imparted to 
them directly by one of their parents, presumably some variety of Rubus 
trivialis. 
Since some of the varieties of true dewberries which show no hybrid 
origin are utterly incapable of setting fruit by themselves, we conclude 
that the true cause for the sterility of the dewberries and their hybrids 
must be sought elsewhere. 
Amoimt of Defective Pollen ( Miscroscopical Test) 
Table No. 6 
PERCENTAGE OF NORMAL POLLEN GRAINS. 
V ariety 
Species 
Count 
Number 1 
Count 
Number 2 
Count 
Number 3 
Per Cent. 
Per Cent. 
Per Cent. 
Austin _ _ 
R. villosus._ _ 
92.1 
95.9 
95.6 
Chestnut _ _ 
R. trivialis_ 
73.5 
77.9 
83.7 
Cox__ . 
Hybrid.. _ ... _ 
78.6 
76.8 
83.7 
Ruth_ 
Hybrid-. _ _ _ __ 
74.6 
82.4 
Grandee __ _ _ __ 
R. trivialis, _ _ _ _ 
41.2 
69.6 
51.1 
Limekiln _ _ . _ _ _ 
R. trivialis_ _ _ __ 
93.4 
92.0 
91.3 
Lucretia 
R. villosus _ _ _. 
92.7 
95.8 
Haupt _ ______ ._ 
Hybrid _ _ _ _ _ 
38 .2 
77.4 
82.9 
Manatee, 
R. trivialis_ 
95.9 
39.7 
36.8 
Munroe _ 
R. trivialis ._ __ 
96.5 
96.3 
96.5 
Premo 
R. villosus _ _ 
94.8 
94.5 
Elijah, No. 2 _ __ 
R. trivialis._ __ 
78.0 
96.3 
93.2 
Rogers. _ _ 
R. trivialis _ 
94.5 
91.9 
92.5 
San Jacinto _ _ _ _ _ 
R. trivialis _ 
68.9 
79.6 
91.7 
White_ 
R. trivialis __ 
95 .2 
96.4 
92.8 
Spalding 
Hvbrid_ 
56.4 
Rathbun 
Hvbrid _ 
57.5 
M eTJonald 
Hybrid 
59.5 
Sorsby 
Hybrid.. _ _ 
76.3 
Wilson 
Hybrid 
53.2 
The method used to secure the data in the above table was as follows: 
ten normal flowers, selected quite at random from different plants of 
the same variety, were used to supply the representative samples of 
pollen. The pollen from the ten flowers was shaken into a glass vial, 
thoroughly mixed by mechanical means, and then subjected to a miscros¬ 
copical examination. This examination consisted in counting and record- 
