IQIS] 



DE VRIES—A MENDELIAN MUTANT 



341 



mut. semialta and mut. debilis. The latter is, on the average, about 

 half as high as the former. This curious segregation repeated itself 

 in the next generation in 1914, not from all the individuals, but 

 from only one of the two whose offspring have been tried in this 

 respect. 



Similar proofs of latent mutations of sexual cells may evidently 

 be expected to occur in other strains also and will have to be looked 

 for in all cases of an unexpectedly high degree of mutability. 



I will now return to my experiments on the production of dwarfs 

 by O. gigas. In order to obtain specimens of 0. gigas yielding a 

 high percentage of dwarfs from their seeds, I sowed in 191 1 seeds 

 of my pure strain, cultivated the plants as biennials, and fertilized 

 them in 1912 by their own pollen, in bags. They were vigorous 

 plants of the fourth generation (Gruppenweise Artbildung, p. 175), 

 and yielded a large harvest of seed, which was sown in 1913, and 

 served as a criterion, since no essential differences were to be seen 

 on the plants themselves. Moreover, I used the seeds of some good 

 biermial specimens of the previous or third generation. The 

 ancestors of all these plants had been fertilized by myself in bags 

 down from the mutant in 1896 which started the race. The harvest 

 of 1912 and 1910, sown in 1913, gave the result as shown in table II. 



TABLE II 

 A. Percentages op dwarfs among oitspring of 0. gigas 



From a second strain, derived from the same mutant and 

 described in my Gruppenweise Artbildung (p. 175), I had in 1911- 

 191 2 nine biennial specimens which yielded a sufl5cient harvest. 



