386 
same causes which led to sterility may lead to other modi 
fications of structure; and surely it is admissible to believe 
that the reverse may also take place, and that similar 
variants may possess fertility inter se and infertility with 
the rest of the species. But if we grant, for the sake of 
argument, that this correlation is possible, so far as the 
variation due to sexual reproduction is concerned, there is 
no conceivable reason why such correlation should occur 
except in the most casual way. Mr. Catchpole assumes 
that this correlation may arise in only two individuals, and 
that these individuals will marry one another.* But if the 
accident of birth is the only cause of this correlation, these 
two individuals may be living far apart and may never 
meet. In this case they have a potential power of be 
coming the ancestors of a new species ; but it will avail 
nothing in the actual life^ If they do unite, it would 
probably be on account of their similar variation which 
led them to segregate themselves, in which case their 
infertility with others would not be of any use. Many 
similar variants, all of whom were only fertile inter se, 
would not avail if they occurred at random, unless they 
could be brought together by segregation or some other 
means. If they were not, the variants would die out either 
because they did not find partners or because they inter 
married with those with whom they would be infertile. 
But if the infertility of variants with the rest of the 
species be due to some definite cause, acting on a given 
number of individuals, this will probably arise from the 
fact that, living together in the same locality, they are 
subject to the same modifying influences. And inasmuch 
as they are already living together by the hypothesis, 
they will intermarry with one another; and then their 
Nature, vol, xxxi., p. 4.