174 MUTATION. 



that it is very probable that mutation, at least loss-mutation is 

 a phenomenon which occasionally occurs. We have tried to 

 show, what extreme difficulties lie in the path of the genetician, 

 who wants to obtain certainty, that the sudden production of 

 a novelty is not caused by redistribution of genes over the 

 gametes of a heterozygote individual. We saw that such a 

 redistribution of genes, Mendelian segregation, can in some 

 instances be deferred for several generations, making test-mat- 

 ings necessary in almost every instance. 



If we exclude as doubtful all those instances of alleged mu- 

 tation in which the necessary tests could not be, or simply were 

 not made, there remain only five or six instances. And as those 

 instances were all cases oi loss-mutations, or at. least cases in 

 which a recessive novelty was produced, which we have tried 

 to show means loss of a gene, we can only conclude that the 

 role mutation can have played and can still play in the evolu- 

 tion of species is at the most a very insignificant one. 



