PROBLEMS OF MODERN SCIENCE 



inheritance, but also the origin of all Mendelian 

 characters. Why this should be so would take 

 us too long to discuss here. Nearly all of the two 

 or three hundred different mutations which have 

 appeared in the fruit-fly Drosophila in the experi- 

 ments of Morgan have been of this Mendelian type. 



Soon after the work of de Vries appeared, the 

 objection was raised that (Enothera Lamarckiana 

 was probably a garden hybrid. But this objection 

 has been set at rest by the discovery of the same 

 process of mutation in various other species, in- 

 cluding wild species with small flowers studied 

 by Bartlett in America, in which the chances of 

 crossing are very remote. It has also to be re- 

 membered that complete absence of crossing is an 

 exceptional condition in plants and an impossible 

 condition in bisexual animals. However evolution 

 has taken place, the unit in which it occurred must 

 have been an interbreeding population of closely 

 related forms. Evidence is accumulating that 

 some apparently * good species,' for example among 

 the roses, are really crypthybrids, but this does not 

 necessarily lessen the evolutionary significance of 

 the variability which they exhibit. 



A more serious objection to mutation as an 

 evolutionary factor is that the new forms which 

 appear are all abnormal, or pathological, or at 

 any rate too weak to compete with their neigh- 

 bours. It is undoubtedly true that many muta- 

 174 



