385 
This correlation between similar variations and infertility 
with other members of the species not so modified might 
arise in three ways: the two phenomena might occur 
simultaneously ; or the variation might arise first and then 
the infertility with reference to the rest of the species ; or 
infertility might appear first and then variations might 
ensue. If partial infertility, with the rest of the species, 
appeared first, then what would happen? There would be 
no suppression of the sexual passion, and those variants 
would be as ready to unite as the others would be. They 
would interbreed with the average of the race, but they 
would have no offspring, or only a few ; and their pecu 
liarity would soon disappear through the inability to 
reproduce itself. If, on the other hand, the similar vari 
ation in structure appeared first, the isolating cause—viz., 
infertility with the rest—woujd be absent; the variants 
would intermarry with the rest of the species, and in the 
absence of prepotent power the variations would be 
swallowed up in the race. Hence it seems to me quite 
clear that in order that this principle should be effective 
the structural variation and the infertility with the rest of 
the species should arise simultaneously in the same 
organism, and moreover that the infertility should be 
complete from the first. If there was only a slight degree 
of infertility this would not prevent intercrossing ; and 
this slight variation might easily disappear as so many 
others do by the principle of regression to the average of 
the race. But what cause can we regard as efficient to 
produce at once a change of structure and absolute 
infertility with the rest of the species ? Assuredly such a 
modification could not arise from the mere union of two 
normal individuals. We must find its cause in some 
external influence brought to bear upon the parturient 
system in many individuals. Mr. Darwin argues that the 
Y