304 
contained favourable variations, and rejected all the column 
to the left because it contained prejudicial variations, 
such selection would be apt to promote the transmutation 
of species rather than the stability, as Mr. Meldona 
suggests. 
“ It is obvious that many individuals are similarly and simultane 
ously modified, and since the action of selection in the conservative 
sense is exerted upon individuals above the mean, while the loss by 
destruction falls upon those which are below the mean, it follows that 
in the next generation the line of mean variability will be raised— 
i.e.y the species will have come into closer harmony with the external 
conditions, and so on in successive generations till equilibrium is 
reached.”—(Nature. vol. xliiip. 410.) 
Clearly the writer is here picturing this selection of one 
column and rejection of another as a means of producing 
a modification of species by raising the line of mean 
variability. But if this is the effect of such selection, 
how can the perfect adaptation of species to existing 
conditions be maintained by the same process ? 
There is also one other consideration which must not 
be lost sight of in this connection. I am not aware that 
it has been ever proved that the one column is composed 
of useful and the other of injurious variations. Many 
differently correlated organisms may be equally efficient 
in fighting the battle of life. Moreover, we must not sup 
pose that they will be rigorously selected in the face of 
accidental death ; or that survivors will not intermarry 
promiscuously, simply because they are represented as 
occupying different sides of a straight line drawn on a 
diagram to illustrate the classification of variants, which 
is justified by the researches of scientific men. 
If, then, in the face of unchanged conditions the prin 
ciple of regression to mediocrity has produced the stability 
of species, and Natural Selection has not been active 
in maintaining the ideal excellence of a species, the idea