MUTATION 



263 



Now this process is not what is ordinarily meant by 

 mutation, for we mean by the latter that a new type has 

 suddenly arisen in the sense that some change has taken 

 place in the germ-plasm — a new gene has been formed. 

 The process here described is one of recombination of 

 genes shown by Mendelian hybrids, the only unusual fea- 

 ture being that all the phenomena involved do not come to 

 the surface because many classes are destroyed by lethals. 



The results are interesting also in another way. It has 

 been assumed by those who think that 0. Lamarckiana is 



Fig. 115. — Rosettes of the twin hybrids of the evening primrose, the plant to the left is 

 called laeta, and that to the right velutina. (After De Vries.) 



a hybrid that the mutant types are only the segregation 

 products of the types or combinations that went in to pro- 

 duce the hybrid. But the Drosophila cases show that 

 balanced lethal stocks may arise within stocks themselves 

 by the appearance in them of lethal factors closely linked 

 to other factors — new or old ones. When new genes arise 

 in such lethal stocks the process may be one of true muta- 

 tion, but the revelation of the presence of the gene is 

 hindered by the lethal factors, so that when the character 

 appears, it appears as a ^'new'' mutant, but is in reality 

 due to recombination of mutant genes that had arisen in 

 an earlier generation. As a matter of fact, the first 



