CONTINUITY IN EVOLUTION S24 
contains fulfil their biological end or perish with it. Natural 
selection secures the survival of those animals which bear the 
seed from which their like will be developed. 
On this view all variation arises within the germinal sub- 
stance, but it is manifested in the body which is its product. 
How variations arise we do not know with any exactness of 
detail. That the germinal substance is influenced in its nutri- 
tion and in other ways by the surrounding tissues is highly 
probable ; ana this influence may lead to changes which are 
the source of variations ; but it is very doubtful whether such 
influence can be what we before termed “‘ homceopathic.”* It 
is improbable that the formation of the nerve-connections 
involved in intelligent behaviour which has grown habitual 
through repetition, can so influence the germinal cells as to 
give rise to variations of like nature. In other words, acquired 
habit is probably not a direct determinant of an inherited 
variation of like nature in instinctive behaviour. Apart from 
such influence the only source of variations which can be 
assigned is either the differential division of nuclei in prepara- 
tion for the process of fertilization,f or the process of fertiliza- 
tion itself. The union of perhaps differentiated germinal 
substance from two distinct parents affords the opportunities 
for the admixture and compounding of hereditary qualities in 
the two samples, from which variations favourable or the 
reverse may arise. 
It is now generally recognized, however, that the origin of 
variations is a problem quite distinct from that of the survival 
of those whose direction is favourable to that end. The theory 
of natural selection, as such, does not pretend to offer any 
explanation of the manner in which variations arise ; though 
of course a complete theory of organic evolution must assign 
the antecedents and conditions of organic progress in all its 
varied phases. We know that variations do occur ; we know, 
too, that more individuals are born than survive to procreate 
their kind ; and, on the theory of natural selection, we draw 
from these data the conclusion that, on the average, the 
* Supra, p. 36. ¢ Supra, p. 13. 
