420 
variants might be compelled to alter their habits to some 
extent, and thus not to compete so severely with those 
retaining the normal form from which they were derived. 
There may be great use sometimes in the principle of 
reversion. It is quite possible that animals may suddenly 
find themselves in some condition of life unfamiliar to 
themselves and to their immediate forefathers, but iden 
tical with conditions to which a still more remote ancestry 
was subjected. If then, in consequence of their descent, 
their bodies should be found to possess a special facility 
of adaptation, or if some long latent instinct were revived, 
the result might be of a very beneficial character. 
Nor must we omit from this enumeration the fact that 
many of these influences are not only reproduced, but 
enhanced in the process of reproduction. They are almost 
sure to be reproduced by the action of heredity, when 
the modifying circumstances which affect all the indi 
viduals of a species, or most of them, remain the same. 
But they are reproduced with an added power. The 
habit becomes an instinct, a second nature ; the adapted 
organism is made more fit for its work by transmission 
through one generation to another. 
But if the changes due to transforming influences are 
sometimes useful, this disposes of the contention that all 
useful variations are due to Natural Selection. Yet the 
inexorable logic of the theory still asserts that the useless 
variation cannot be produced by Natural Selection. So 
far then as changed conditions produce transforming 
effects which are not useful and useful at every stage, to 
that extent the action of Natural Selection is limited. 
Opportunities of transmutation, open to transforming in 
fluences, are closed against Natural Selection. 
There is another case, to which further reference will 
be made hereafter, in which transforming influences may,