154 BERTRAM G. SMITH. 



ing and underlying gravel this hollow looked as if it had been 

 dug out by some animal. The eggs had been disturbed and 

 scattered considerably ; the strings were short, as if they had 

 been much broken up, and many separated eggs were found. 

 The eggs were thickly covered with silt. Both the envelopes 

 and the contained eggs were, as a rule, slightly larger than those 

 previously found. No adults were seen in the vicinity. 



The eggs obtained on September 6 from a specimen in cap- 

 tivity were found rolled and tangled together in such an intricate 

 manner that they seemed to occur in clusters or masses rather 

 than in strings. The number was estimated at about 80, nearly 

 all in one oblong mass. The solitary female did not appear to 

 care for the eggs in any way. 



On September 8 about 300 more eggs were laid by the same 

 female specimen in captivity. The strings of eggs were aggre- 

 gated in one large mass, but they were not so much tangled as 

 in the case of those laid previously. There were present at the 

 time in the same aquarium three male specimens, but so far as 

 was observed, none of the adults paid any attention to the eggs. 

 When all four specimens were killed a few days later for the 

 purpose of determining the sex by dissection, the stomachs of 

 the males were found distended with undigested eggs. 



Upon examination, the female, which during captivity had laid 

 nearly 400 eggs, was found to contain, at a rough estimate, seven 

 or eight hundred more, in a state of development whiqh indicated 

 that had she been allowed to live they would all have been laid 

 during the same season. Evidently the eggs are, in some cases at 

 least, matured and laid in batches of a few hundred at a time. 



According to Gadow ('01), Amblystoma alone among the Uro- 

 deles lays as many as 1,000 eggs in a single season. The facts 

 stated above make it probable that CryptobrancJius rivals Ambly- 

 stoma in the number of eggs laid. Kerbert's specimen of C. 

 japonicns is reported to have laid 500 eggs in the fall of 1902, 

 and about 900 in 1904. The eggs of AmpJiinma found by Hay 

 numbered about 150. 



The absence of any evidence of brooding habits of either the 

 male or the female Cryptobranchus is rather unexpected in view 

 of the possession of brooding habits by closely related forms. 



