HABITS, ETC., 0¥ OERBBRAT[JLUS LACTEUS. 113 



summers that it was very difficult to obtain perfect specimens 

 toward the close of the breeding season, though they were 

 fairly common at an earlier date. On the contrary, the 

 numberof specimens possessing regenerating papillge increased 

 as the season advanced. It was evident, therefore, that they 

 must dismember for some reason during the breeding season, 

 and it was suspected that the eggs might be discharged in 

 this way. During the past summer the matter was put to 

 test with the following results. 



A perfect male and female Cerebratulus were secured 

 during the latter part of June, and placed in an aquarium 

 half j&lled with the same mud from which they were taken. 

 They burrowed into this mud at once, and remained there 

 under fairly normal conditions, and apparently perfectly 

 healthy, until the 1st of September, when they were both 

 preserved on leaving. 



They were nearly sexually ripe when obtained, and after 

 being kept a fortnight the female laid three batches of 

 eggs at short intervals, each of which was fertilised by the 

 male. 



Soon after the last lot was laid they both dismembered, 

 leaving but a comparatively short piece of the body attached 

 to the oesophagus. This anterior portion at once began to 

 regenerate, and by the 1st of September, an interval of 

 three and a half weeks, it possessed a regenerating papilla 

 nearly 50 mm. long in the female, and about 40 mm. in the 

 male. 



The posterior fragments lived for two weeks, and then 

 died with no signs of regeneration. I have kept headless 

 fragments alive, however, under much less favourable cir- 

 cumstances for nearly a month, and Mcintosh in his mono- 

 graph (34) states that most of the pieces of a fragmented 

 Lineus were alive six months after dismemberment, and that 

 the sexual products were developed and discharged from 

 them in a perfectly normal manner. 



But this happened to be a male, and he could obtain no 

 eggs upon which to test the fertilising power of the sperms. 



VOL. 43, PART 1. — NEW SERIES. H 



