301 
supposed selection, all investigation goes to prove that 
there is a great deal of variation in all species, however 
fixed. This fact is generally urged to show the possibility 
of transmutation of species by means of Natural Selection, 
in so far as it offers material for selection to act on. But 
there is another way of looking at this matter. If, in 
those cases in which the species has remained stable, we 
take a given number of adults who have passed through 
the period when destruction has been most at work—we 
must look upon the relative differences as the result of the 
action of Natural Selection. We do not find what we 
should expect if Natural Selection were preserving the 
species at its highest possible excellence ; but we do find 
that enormous difference between individuals, which is 
consistent with the stability of a species. 
Now it follows, necessarily, that if Natural Selection has 
not maintained the highest efficiency of a species, the cessa 
tion of Natural Selection will not lead to that deterioration 
which might otherwise have been anticipated, but rather to 
that stability of species which the principle of regression 
can of itself maintain. 
It is interesting to note the reasons which are assigned 
for the supposed deteriorating results of the cessation of 
Natural Selection. It was taken for granted that there will 
be phis and minus variations on either side of the average. 
Mr. Herbert Spencer thereupon assumed that in the ab 
sence of Natural Selection these plus and minus variations 
would neutralise one another, and so the stability of the 
species would be maintained. But Mr. Romanes replied 
that in the absence of Natural Selection there would be 
an excess of minus variations. Whereupon Mr. Spencer 
asked for proofs, and Mr. Romanes adduced two. 
“(i). When selection is withdrawn there will be no excessive plus 
variations, because so long as selection was present the efficiency