dates, Tetraploid Mutants and Chromosome Mechanisms. 



characters", is obviously not in harmony with the usual biological 

 conceptions. For that term in biological usage, means the occur- 

 rence of a somatic modification, and its subsequent reflexion back 

 into the germ plasm. But a megaspore mother cell or a fertilized 

 egg is germ plasm par excellence, and any chromosome-doubling 

 occurring here is obviously a germinal change, and as such likely 

 to be inherited. We may contrast with such a condition the one 

 obtained experimentally by Nemec (1910), in which, by treatment 

 of root tips, a doubling in the number of chromosomes in various 

 cells of the growing tip is obtained. There is obviously no pos- 

 sibility of such a change being passed on to the next generation, 

 and as a matter of fact the tetraploid condition of these cells 

 gradually disappears, although opinions differ as to how the return 

 to the diploid number takes place. De Vries cites in comparison 

 with yiyas the binucleate cells of Spirogyra obtained by Geras- 

 simow, which afterwards returned to a uninucleate condition. But 

 the recent important work of the Marchals, already cited, in obtain- 

 ing diploid moss gametophytes by wounding the sporophyte, is a 

 closer parallel, and those experiments frequently gave rise to con- 

 stant tetraploid races. The statement of de Vries, that my point 

 of view is ,,vollig widerlegt" by the newer facts, is therefore scar- 

 cely in accord with the evidence. 



De Vri,es(l. c., p. 35) calls the triploid forms \\a\i-giyds mutants 

 and states that they agree in character with 0. yigas X Lamarck- 

 nnta, a result which would be expected. He finds that when 

 Lamarcldana is crossed with pollen from cruciata, muricata or 

 Millersi (nov. sp.), most of the seedlings produced are yellowish, 

 the occasional deep green ones (15000 yellowish to 45 green), called 

 Hero, having 21 chromosomes. This gives a mutation-coefficient 

 of about 0.3 per cent., which is assumed with probability to represent 

 the frequency of diploid eggs in 0. Lamarcldana. This of course 

 furnishes no evidence of diploid pollen grains. If such really occur 

 in Lamarcldana and are functional, it would seem probable that 

 their frequency might be determined by making the reciprocal 

 crosses, with Lamarcldana as the pollen parent, but this does not 

 seem to have been done. We must conclude, then, that the inter- 

 esting evidence offered by de Vries shows only the frequency of 

 diploid egg cells, the occurrence of which we already had some 

 reason to believe in through the observation of Geerts, but offers 

 no support whatever for the occurrence of functional diploid pollen 

 grains. 



In discussing the status of 0. yigas, de Vries (1. c., p. 36) 

 regards it, and I believe rightly, as ,.einegute progressiv entstandene 

 Art". He believes also that many of the differences from its parent, 

 0. Lamarcldana, cannot be explained as a result of the original 



