1882.] The Development of Renilla. 385 



I. 



(1.) As in other Alcyonaria, Renilla is dioecious, and fertilisation is- 

 effected in the water after discharge of the generative elements 

 through the mouths of the feeding polyps. The ovarian or spermatic 

 follicles are ruptured and cast aside at the time of spawning. 



(2.) The action of the vitellus during segmentation is extremely 

 variable, while the division of the nuclei appears to be nearly regular. 

 The vitellus, after a preliminary division of the nuclei, may divide at 

 the first cleavage into two, four, (?) eight, sixteen, or thirty-two 

 spheres, which may be equal or noticeably unequal. In some cases 

 cleavage begins at one pole a considerable time before any sign of 

 activity is shown at the opposite pole, so that the segmentation appears 

 precisely like that of a true meroblastic egg. In others, again, the seg- 

 mentation is irregular and extremely unequal, so as to appear quite 

 like that of an epibolic gastrula. All of these forms gave rise, how- 

 ever, to quite similar larvae, which were proved by isolation in small 

 aquaria to be normal and healthy. 



In several cases of division into sixteen and thirty-two spheres, the 

 vitellus was observed to undergo slight changes of form some time 

 previous to actual division. These changes appear to be the expres- 

 sion of attempts at division on the part of the vitellus, which has not, 

 however, energy enough to carry out a complete cleavage. In other- 

 cases the attempts are partially successful and the egg divides incom- 

 pletely into spheres which do not become clear and well defined until 

 the following cleavage. The series of forms appears to be produced 

 by variations in the activity of the vitelline protoplasm or in the 

 resistance (in the form probably of deutoplasm) which is opposed to it. 

 The phenomena are of considerable interest, as showing how natural 

 selection may find a field for action even in the earliest stages of an 

 organism, and as a caution against drawing too hasty conclusions in 

 regard to the character of segmentation from the study of a few indi- 

 viduals only. It is further remarkable to find the action of the vitellus 

 and of the nuclei as independent of one another as these facts seem to 

 indicate. 



(3.) Cleavage is at first superficial, a considerable central mass 

 remaining unsegmented. The furrows finally extend to the centre, 

 and a small temporary segmentation cavity is formed. 



(4,) The layers are separated by a process of delamination. The* 

 endodermic mass is at first solid and is not separated by a supporting 

 lamella from the ectoderm. 



(5.) The supporting lamella is derived mainly from the ectoderm 

 by a very peculiar process of secretion at the inner ends of the cells. 



(6.) The gastric cavity is formed by absorption of the central endo- 

 derm cells by those which are more peripherally placed. The former 

 undergo a peculiar process of disintegration and form a granular debris,.. 



