THE MULTIPLE TESTIS OF URODELES. 47 



structure. The species possessing the multiple testes (Desmogna- 

 thus and Diemyctylus} are, in comparison with the above, small 

 animals, yet their testes may consist of from two to five lobes. 

 Plainly the size of the species is not to be regarded as a factor. 



That the elongated form of the body is responsible for the 

 multiple testis is likewise hardly conceivable. Among our Ameri- 

 can urodeles none are more slender and graceful of body than 

 Gyrinophilus and Spclcrpcs bislincatiis. Yet in these animals the 

 testis, though an elongated structure, is not divided into lobes. 

 Desmognathus and Diemyctylus, possessing multiple testes, are 

 comparatively stout-bodied animals. 



Champy's conclusion that the number of lobes in males of the 

 same species is proportional to their body length is only in a 

 limited sense correct, as the following figures from my records 

 show : The average length of eight Desmognathus males with testes 

 of two lobes each is 9.8 cm. ; the average length of six males with 

 testes of three lobes is 10.2 cm. or 0.4 cm. greater. Nevertheless 

 one of the individuals of the first group measured 12.2 cm., while 

 no animal of the second group exceeded a length of 10.5 cm. The 

 average body length for the same group of eight males having 

 testes of two lobes each is 9.8 cm. Within this group, however, 

 are lengths ranging from 8 cm. to 12.2 cm. Clearly other factors 

 than mere body length must be concerned, else we should find much 

 less variation within this group. 



Conceding that males which have been kept small by lack of food 

 may develop testes of few lobes, as Champy states, the writer is 

 forced, so far as Desmognathus is concerned, to doubt Champy's 

 further claim that animals well fed may become very large, and, at 

 one year of age, possess testes of numerous lobes. Since the 

 urodele male completes but one spermatogenetic cycle annually, the 

 males of one year referred to by Champy would of necessity reach 

 sexual maturity with testes of many lobes already developed. In 

 other words, the multiple testis, whatever its correlations, would 

 develop simply through the initiation of greater growth activity in 

 localized regions of the original cord of germ cells, as suggested 

 by Kingsbury ('02). A male approaching sexual maturity might, 

 on the basis of this theory, possess an even greater number of lobes 

 than one which had been sexually active for several seasons. 



