Self-sterility in Dewberries and Blackberries 
11 
under certain conditions accept pollen from the same flower or from 
flowers of the same variety. 
Of the flowers that were allowed pollen from the same flower and 
flowers of the same variety we find that, in the case of White, out of 
fourteen flowers inclosed, six produced fruit composed of a few druplets, 
and eight remained absolutely unaffected. The number of druplets 
that were produced by the flowers varied from one to five. Of the ten 
flowers of Rogers, six developed a few druplets and four developed none; 
while of the ten flowers of Munroe, only one flower developed a few 
druplets and nine absolutely refused the pollen. 
That the injury done to flowers in the emasculation process, if care¬ 
fully done, is negligible and does not affect the fruit formation is seen 
in the last part of the table where fruit buds of White were emasculated 
and crossed with pollen from the variety Rogers. Each of the four 
flowers that were emasculated and then cross-pollinated set and developed 
a normal good-sized berry. The three buds of Munroe that were treated 
in similar manner and then cross-pollinated with pollen from the 
variety of Chestnut also developed berries, normal in every respect. 
The three buds of Rogers that were emasculated and crossed with pollen 
from White produced no fruits. This result at first glance would indi¬ 
cate a positive reaction to emasculation injury, but a careful examina¬ 
tion of the flower remains indicated that the trouble lay with the pollen 
rather than with the treated flower. 
All these results seem to point toward the fact that a few indi¬ 
vidual pistils of flowers of these otherwise self-sterile varieties will 
accept some pollen from the same flower in order to effect fertilization. 
With these self-sterile varieties, this phenomenon seems to he the sole 
remaining sexual means of perpetuation of the species in case of 
isolation. 
With this explanation as to the cause of imperfect fruits, we get a 
better understanding of the real significance of the data in Table No. 1. 
The varieties of dewberries then can be grouped into self-fertile and 
practically self-sterile classes, which are as follows: as self-fertile, we list 
the varieties, Austin, Cox, Ruth, and Lucretia; as self-sterile, Chestnut, 
Grandee, Limekiln, Haupt, Manatee, Munroe, Premo, Elijah No. 2, 
Rogers, San Jacinto, and White. 
If now we turn to the column of species, we shall discover that all 
of the varieties that are self-sterile, with only two exceptions, fall into 
the species Rubus trivialis. It will be noticed that this species contains 
no self-fertile varieties, and this is a good indication that the species 
itself if self-sterile. The two self-sterile varieties that are not classified 
with this species are Haupt (a hybrid) and Premo, the only self-sterile 
variety of the species Rubus villosus that we are growing. 
Of the self-fertile varieties, Cox and Ruth are hybrids, and Austin 
and Lucretia belong to the species Rubus villosus. 
