354 



The Journal of Heredity 



the germ-plasm, in other words is de- 

 termined as the result of a certain 

 quantitative balance between these 

 genes; (2) that the rate of working 

 of these factors is important, since 

 their effect is ultimately exerted 

 through their chemical activity; (3) 

 that by making, through racial or 

 species crosses, certain combinations 

 of these factors, it is possible to pro- 

 duce a consecutive intersexuality — i. e., 

 sex-dimorphism in time in one and 

 the same individual, one sex phase 

 giving place to the other after a cer- 

 tain moment. This effect is due to 

 combining relatively slow-working 

 factors for one sex with relatively 

 quick-working factors for the other, 

 and to the consequent intersection of 

 the curve of production of their sex- 

 determining substance or substances. 



Having established these points, we 

 may profitably consider the last a 

 little further. In the first place it 

 should be noted that various animals 

 besides the Gipsy Moth may show a 

 similar sex-reversal during develop- 

 ment ; but in animals in which no 

 hard parts connected with sex are 

 laid down in permanent and unal- 

 terable form, the transformation from 

 one sex to the other will almost al- 

 ways be eventually complete, since the 

 soft parts of one sex will be absorbed 

 or remodeled and those of the other 

 sex laid down in their place. This, as 

 a matter of fact, occurs in Frogs, as 

 Witschi,^ Crew" and others have 

 shown. To get evidence of the sex- 

 change here we must either argue in- 

 directlv from sex-ratio, or follow the 

 indiviclual during its development. 



In the second place, there are a 

 great many forms of life in which a 

 consecutive sex-dimorphism — in other 

 words, sex-reversal at a certain point 

 of time diiring development normally 

 does occur. There are the numerous 

 bisexual forms that rise by male or fe- 

 male individuals taking on the char- 

 acters of the other sex. and giving 

 rise to ])rotandric and protogynous 

 hermaphrodites, the former kind be- 

 ing the more abundant. This ques- 



tion has been treated at length by 

 Goldschmidt in the work previously 

 cited". Suffice it to say that frequent- 

 ly the change of sex appears definitely 

 to be associated with and determined 

 by a change in metabolism, such as 

 that inevitably brought about by in- 

 crease of size and age. Here one 

 might speak of normal and abnormal 

 sex-metamorphosis, normal in herma- 

 phrodites, abnormal in Goldschmidt'.-: 

 moths and Crew's frogs. 



This brings us back to the question 

 with which we started — a question 

 which this, I fear somewhat lengthv 

 digression, has been intended to il- 

 luminate. In these cases which have 

 be'^'n properly analysed, sex-m.etamor- 

 phosis has been shown to depend on 

 the intersection of the curves of pro- 

 duction of certain substances. The 

 production of these, in their turn, de- 

 pends upon the relative amounts or 

 rates of working of certain factors in 

 the organism's constitution. Why, we 

 may then ask, if it is possi1)le to have 

 such sex-metamorphosis, should not 

 a similar state of affairs be supposed 

 to exist in true metamorphosis? I 

 will endeavor to show that such a state 

 of affairs may very probably be sup- 

 posed to exist. 



We have first to suppose — what is 

 clearly the case — that the germinal 

 constitution of the Anura (which I 

 choose because the problem has in 

 them been most intensively studied) is 

 such that it interacts with the normal 

 environment to produce the tadpole 

 type of organization, and this wJiether 

 a thyroid be present or not. The ex- 

 periments of Allen" and others witli 

 thyroidectomized tadpoles show con- 

 clusively not only that normal healthy 

 growth is possible in tadpoles with- 

 out a thyroid, but that the tadpole or- 

 ganization will persist and grow to a 

 size far above normal when no thyroid 

 is present. Some of Allen's thyroid- 

 less tadpoles grew to a length of sev- 

 eral inches and became almost as large 

 as the normal metamori)hosed adults 

 of the species. It is quite clear, then, 

 that the tadpole type of organization 



