327 
the parent species. It is obvious at the outset that if the 
variants have less advantage than the parent stock, they 
will, according to the theory, be destroyed, and then the 
stability of species, rather than its transmutation, will 
ensue. If, on the other hand, the variants, which pre 
sumably will be in the minority, have the advantage, and 
if that advantage is gained by the survival of the best 
among themselves and the destruction of all the rest, the 
process may be too costly to secure their preservation as a 
race. It is possible that they might supplant the parent 
species and yet not permanently establish themselves. 
The process would always be liable to produce that rarity 
which is the precursor of extinction ; or, if rarity already 
existed, to convert that rarity into extinction. Other 
methods of transmutation, on the other hand, which are 
not dependent upon the accidental emergence of favour 
able variations, and which do not work by life and death, 
are not so obnoxious to this danger. 
The connection between the extinction of species and 
Natural Selection appears to be this. A quasi extinction 
of species is the prelude of transmutation, when all the 
members of a species undergo modification ; but this fact 
does not necessarily imply that Natural Selection has been 
at work. The extinction of a whole species obviously 
prevents the transmutation of that species, whether by 
Natural Selection or any other cause. There is no reason 
to suppose that the bifurcation or ramification of species 
has been produced by Natural Selection ; nor does it 
follow that species and variants will necessarily compete 
with one another. This conflict can be, and often is, 
avoided. If conflict did arise, it would either lead to the 
stability of species, if the parent species survived, or 
threaten the few surviving favourable variants with 
extinction. Finally the extinction of species reminds