60 JouRXAL OF THE IMiTCiiELL SOCIETY [Septcniher 



lization. This thoory is contrary to the modern concept of ferti- 

 lization, and seems untenable, at least in the case of C. cinerea, al- 

 though there is no specific evidence bearing on the case. 



There is, however, some contributory evidence. If turtles be taken 

 from their native haunts before the spring copulation season, they 

 produce wholly infertile eggs, and it is only after spermatozoa begin 

 to appear in the oviducts of turtles in their native habitat that fertile 

 eggs are found. It would seem reasonable that the fall copulation 

 should be the fourth for some considerable number of individuals, 

 and hence should initiate cleavage. But no such cleavage is found. 



The foregoing observations lead also to the conclusion that the 

 fall copulation is ineffectual for fertilization. The idea that the 

 spermatozoa are retained in the body of the female during the winter 

 and are capable of fertilizing the mature eggs in the spring is also 

 erroneous. In a few weeks after the close of the fall season a careful 

 examination of the uterus, oviducts, bladder and body fluids of a 

 number of specimens failed to reveal any spermatozoa. This leaves 

 still unexplained the presence of a "food mass" observed on the 

 spermatozoa by Glascock (not yet published) which might have been 

 interpreted as an adaptation for this purpose. 



As had been suggested, the males and the females are of con- 

 siderably different sizes when copulation begins. "Whether this is 

 due to a differential growth rate or a different age for reaching sexual 

 maturity is undetermined, but the evidence presented here appears 

 to sustain the former idea. That copulation does not begin until 

 sexual maturity has been reached by the females, at least, is evi- 

 denced ])y the fact that the spermatozoa were not observed in the 

 oviducts of sexually immature individuals. One turtle 125 mm. 

 taken after the copulation season was well under way showed no 

 traces of spermatozoa while another, 130 mm. in length, taken five 

 days later, produced fertile eggs. Since, as already shown, the males 

 may have active spermatozoa in the epididymus when they reach a 

 length of 88 mm., it is safe to assume that copulation begins in the 

 male when a carapace length of 85 mm. has l)oen reached. In the 

 case of females, no spermatozoa were found in the turtles with a 

 carapace length of less tlian 1:50 mm., therefore copulation must 

 begin about that time. 



Development to a certain stage takes place in the oviduct. The 

 great majority of eggs removed from the oviduct had attained the 



