The Explanation of Adaptations. 609 
A strong argument for my view was put forward by 
ROSA and CATTANEO. 1 According to these authors the 
extinction of large groups of species proves that the 
variability resident in them was powerless to adapt them 
to the changing conditions of life; and from this con- 
clusion they infer that the ordinary variability, as it is 
always manifested, is not sufficient for this purpose. Ob- 
viously some other process is necessary. 
II. Fluctuating variability is linear] it oscillates only 
in a phis and a minus direction, whilst adaptations demand 
a variability which will produce variations in all direc- 
tions. 2 On this point also I have expressed my opinion 
in the first volume (p. 118). It constitutes, in my opin- 
ion, one of the strongest objections to the prevailing 
view ; and it also shows more clearly than anything else 
how far DARWIN'S adherents have departed from the 
views actually expressed by him. To DARWIN'S mind 
the essential point was that the struggle for existence 
should have to select from material supplied by an in- 
determinate variability. Natural selection is a sieve. It 
creates nothing, as is so often assumed; it only sifts. 
It retains only what variability puts into the sieve. 
Whence the material comes that is put into it, should be 
kept separate from the theory of its selection. How the 
struggle for existence sifts is one question; how that 
which is sifted arose is another. In both respects, DAR- 
WIN'S original view is still the best, but the point at issue 
has been often obscured by later writers. The meshes 
of the sieve are not such as to separate only the very best ; 
on the contrary natural selection only throws out some 
part of the individuals, and amongst them the worst, i. e., 
1 See below in 12. 
2 GusT,\v WOLFF, Dcr gegenwdrtige Stand dcs Dari^'inisinns, 1896. 
