592 Species According to the Theory of Mutation. 
the best varieties are, as a rule, not united by transitional 
forms with the parent species ; in the second place, trans- 
gressive variability tends to obscure boundaries where 
they really exist. These limits are often overlooked in 
the descriptive method, and the search for them can only 
be carried out on experimental and statistical lines. With 
good right DE CANDOLLE speaks in such cases of provi- 
sional species. 1 
5. THE PARALLEL BETWEEN SYSTEMATIC AND 
SEXUAL RELATIONSHIP. 
Since the doctrine of descent now commands general 
recognition, it is desirable that the systematic divisions 
should be an expression of the various degrees of rela- 
tionship. Even before the appearance of DARWIN^S 
works it was recognized that the task of systematic biol- 
ogy as a descriptive and classificatory science was differ- 
ent from the mere question of actual relationship. To in- 
vestigate this and, where possible, to bring the divisions 
of the natural system into harmony with it, these were 
the ends which the pioneers in the study of hybridization 
had continually in view. 
The result did not, however, correspond to this ex- 
pectation. We have not, as yet, succeeded in bringing 
into harmony the study of hybridization with that of 
systematic biology. NAGELI expressed this incompati- 
bility most clearly by introducing his conception of sex- 
ual affinity. The degree of this affinity between two 
types was judged first by the degree of their fertility 
when crossed with one another, and, then, by that of the 
fertility of the hybrids thus produced. 
1 ALPH. DE CANDOLLE, La Phy to graphic, pp. 98, 167. 
