A. B. Stour 89 
Eleven cultivated varieties directly derived from the blackberry and 
propagated vegetatively were found to be self-compatible, while of 
twelve varieties descended from the dewberry ten are self-incompatible. 
Wild plants of the blackberry (Rubus villosus) were found to be self- 
_ compatible. Thus it appears that R. villosus is a self-compatible species 
while R. trivialis is strongly self-incompatible. : 
Of the varieties known to be hybrids between these two species, 
three (McDonald, Sorsby and Spalding) are self-incompatible and one 
(Rathbun) is partially self-incompatible. . : 
Further data are needed to determine fully whether the compati- 
bility or incompatibility of the hybrid varieties is ever complete and 
absolute, but it is clear that the self-incompatibility characteristic of 
one parent appears in various degrees in the different individuals of the 
F,. It is reported that the self-compatible plants when self-pollinated 
“produce apparently as good fruits as when cross-pollinated.” In regard 
to cross-compatibility it is reported that most crosses give successful 
fertilization ; evidently no decided cases of cross-incompatibility were 
found. | 
Detjen finds that, in addition to sterility from physiological incom- 
patibility, certain hybrid varieties of the dewberry-blackberry cross are 
also more or less sterile from impotence of sex organs, as is common in 
inter-specific hybrids. 
A most decided case of variation in sex relations has been reported 
by Sirks (1917) in the study of cross-incompatibilities in the herma-. 
phrodite species Verbascum phoeniceum. Here the reciprocal crossing 
between pairs of plants gives all grades of opposite results in immediate 
fertility. In the extremes, both crossings may be highly and perhaps 
absolutely compatible, or they may be incompatible, or one may be 
compatible and the other incompatible. In.the latter case using one 
plant as a male and the other as a female gives full compatibility ; 
reversing the relation gives complete incompatibility. Furthermore 
there are many grades in the comparative fertility of reciprocal crossings 
between two plants. - 
This evidence agrees with that which the writer (1916) has presented 
for chicory, but Sirks’ data are much more extensive and conclusive. 
It may be stated here that while Sirks has not reported on the self- 
compatibilities of V. phoenicewm, my own studies with this species, to be 
reported later, show that in my strains at least there appears to be no 
tendency for an end-season change in compatibility that might lead to 
