EXPERIMENTAL EVOLUTION 537 
found that they bred true (except for additional muta- 
tions) when propagated by seed for over 25 years that 
is, they were true mutations. 
457. Relation of Mutation Theory to Darwinism. The 
mutation theory is not intended by de Vries to supplant 
the theory of natural selection, but to demonstrate that 
the materials upon which selection acts in the formation 
of new. species are mutations, and mutations only never 
fluctuating or individual variations. In the second place 
the mutation theory explains away numerous objections 
to natural selection. It shows how characters that are 
never of vital importance 1 i.e., matters of actual life or 
death- to a species may arise and be perpetuated. With- 
out mutation this is difficult to explain, 2 and yet many, if 
not most, of the characteristics by which different species 
are distinguished from each other are of this kind not, 
so far as we can see, absolutely essential to the life of the 
species. Mutation also offers a method by which evolu- 
tionary changes may take place within a much shorter 
time-period than was demanded by the natural selection 
of fluctuations. 
Incidentally, the mutation theory clearly shows that the 
absence of "connecting links" between species is no argu- 
ment against evolution, but is, on the contrary, just what 
we might expect to find. 
458. Value of the Mutation Theory. As stated above, 
the elaboration of the mutation theory has furnished the 
1 As required by Darwin's theory. See quotation on p. 516. 
'* Other explanations have been offered. For example, sometimes two 
characters appear to be always associated, so that the presence of one 
involves the presence of the other; as a mane and maleness in the lion, 
dicotyledony and exogeny in Angiosperms. 
