EVIDENCES OF MUTATION 81 



self -pollinated, may give a variable progeny in crosses, and 

 they seem to breed true merely because certain classes of their 

 progeny are too feeble to survive. For in some cases only a 

 fractional part of the seeds produced contain embryos 

 capable of survival. 



According to the views expressed above, Oenothera Lamarch- 

 iana is best interpreted as an impure or hybrid species which 

 only breeds true in a relatively high degree because of 

 extensive sterility, which eliminates large numbers of gametes 

 and zygotes that differ from the germinal cells which repro- 

 duce the Lamarckiana type. The " mutants " come from 

 occasional seeds of different types that survive the heavy 

 mortality which renders sixty per cent or more of the seeds 

 infertile and about fifty per cent of the pollen grains abortive. 

 If this is the correct explanation of the peculiar breeding 

 behavior of Lamarckiana, this plant is very far from being 

 representative of a pure species, as De Vries assumed it to be, 

 and is hardly suitable material for experiments designed to 

 give evidence of mutation. 



Even if we reject this explanation and consider that the 

 mutability of the evening primrose has no causal relation to 

 its hybridity, it by no means follows that mutation is a 

 general method of origin of new varieties and species among 

 animals and plants, which is the thesis of De Vries. In recent 

 years the expression "mutation theory" has been used in a 

 sense very different from that in which De Vries originally 

 used it, and implying merely the origin of new and stable 

 organic forms by change in single inheritance factors (genes), 

 whether these produce striking variations (sports) or varia- 

 tions so minute as to be scarcely observable. This form of 

 mutation theory will be discussed in a later chapter. To the 

 mutation theory of De Vries, as a general theory of evolution, 

 it seems to be a fatal objection that such mutation as it re- 

 cognizes is not general in occurrence. Crosses of species or 

 varieties as found in the wild state more often reveal the 

 existence of numerous minute genetic differences than a 

 single or a few striking differences. 



