482 



THE AMERICAN NATURALIST [Vol. XLIV 



But if at any time individuals are taken out of this cul- 

 ture medium and put into spring water, males and sexual 

 females appear. By diluting the spring water with 

 varying amounts of culture water the number of sexual 

 forms that appear is directly proportional to the dilution. 



In this animal the individual that produces the male 

 eggs is the same individual that produces the sexual egg. 

 If she is early fertilized by a male, her eggs produce 

 sexual females. It is clear here that external condi- 

 tions change the cycle but do not determine sex. This 

 brief review will suffice to clear away the traditional 

 evidence supposed to support the view that sex is deter- 

 mined by the environment. 



Let us pass now to the results that seem to show that 

 there is an internal automatic mechanism that regulates 

 the production of males and females. I shall take up 

 the botanical evidence first. 



Con-ens' experiments with two species of Bryonia, 

 may be examined. Bryonia dioica is dioecious; B. alba 

 monoecious. Correns ' main experiment shows that when 

 dioica $ is crossed with alba <$, all of the offspring are 

 females, but when alba $ is crossed with dioica half 

 the offspring are male and half female. The results can 

 be explained by three assumptions. First that the male 

 condition dominates the female, second that the dioecious 

 condition dominates the monoecious; third, that the fe- 

 male is homozygous in regard to sex and the male heter- 

 ozygous. 



These conclusions are opposed to the interpretation of 

 other workers that make the female the dominant condi- 

 tion. It is also not clear from G. H. Shull's recent work 

 on a more extensive scale that the dioecious condition can 

 be assumed in general to be dominant. 



Whatever the correct interpretation may be, these facts 

 show at least that by treating sex as a character that 

 segregates in the gametes, as Mendelian characters in 

 general are assumed to do, the results can be accounted 

 for, provided one sex is assumed to be always heterozy- 

 gous and the other homozygous. 



