Recurrent Mutations 
pure ancestors and breed true zuter se. Nor is 
this all. 
Experience shows that where a mutation, or 
sport, or discontinuous variation occurs, it fre- 
quently repeats itself; for example, the black- 
winged sport of the peafowl has occurred several 
times over and in different flocks of birds. The 
sport or mutation must have a definite cause. 
There must be something within the organism, 
something in the generative cells, which causes 
the mutation to arise; and hence, on a przore 
grounds, we should expect the same mutation to 
arise about the same time in many individuals. 
It seems legitimate to infer that things have 
been quietly working up to a climax. When 
this is reached there results a mutation. There- 
fore we should expect sudden mutations to appear 
simultaneously in a number of individuals. To 
this important subject we shall return. 
8. An almost insuperable objection to the 
theory that species have originated by the action 
of natural selection on minute variations, is that 
such small differences cannot be of a life-or-death 
value, or, as it is usually called, a survival value 
to their possessor. But if evolution is the result 
of the preservation by natural selection of such 
slight variations, it is absolutely necessary that 
each of these should possess a survival value. 
As D. Dewar has pointed out, on page 704 of 
vol. ii. of The Albany Review, it is only when the 
43 
