102 



White. 



Table E. 



No plant in any of the four families was any less normal than 

 the parents, and there were many more extreme abnormals in all the 

 selections than were present in either of the 1910 families. Roughly 

 301 — 1 — 29 appeared to be more abnormal in stem-fasciation then 

 301 — 1 — 2, but this was not true of the families 303—1 — 14 and 

 303 — 1 — 12, although the two parent plants represented extreme con- 

 ditions. If one may draw conclusions from such scant data, I should 

 interpret these results as showing the ever -varying nature of the 

 character, not ever-varying however, in the sense of de Vries. No plant 

 ever approached the normal, and were it not for labels, I should have 

 been unable to have distinguished the two 301 — 1 selections from the 

 parent strain growing beside them. De Vries, too, after two or three 

 years, found selection of little value, and this was especially true in 

 his attempt to produce a normal cockscomb by selection from an 

 abnormal race. The value of selection for the first two or three years 

 in de Vries' cultures may be accounted for by the fact that his plants 

 came from the wild, of whose immediate ancestry he was ignorant. 

 His fasciated races were also plants which as a rule were cross-fertilized. 

 Selection work on this Nicotiana race should be continued, and careful 

 detailed records taken during many years before a dogmatic decision in 

 regard to selection of an abnormal from a normal and a normal from 

 an abnormal could be made. The material is ideal for such work, as 

 it fills the conditions called for by the advocates ofJohannsen's pure 

 line theory admirably, and the objection of bisexual inheritance is not 

 here applicable as in the case of animals. Starting with what is in 

 all probability a natural pure line, made homozygous by thousands of 

 generations of inbreeding, and making use of a character that acts very 

 clearly as a single unit, it would seem that the question of the power 

 of selection to modify this particular gene could be irrevocably settled. 



