744 
DR. E. B. WILSON ON THE DEVELOPMENT OF RENILLA. 
variations in the distribution of the deutoplasm. As shown in figs. 55-57 certain 
spheres may be slower in their development than others, so that their descendants 
are larger, a fact long since observed by Allman in Hydroid eggs. This is probably 
caused by the presence of a larger amount of deutoplasm than common, though possibly 
to the tardy division of the nuclei. In the forms of “ partial ” segmentation shown in 
figs. 59-67, the large unsegmented mass must contain a number of fully formed 
nuclei, since it breaks up almost at once into several spheres. Hence the delay is 
caused, apparently, by some obstacle in the vitellus, which we may suppose to be an 
especially great amount of deutoplasm in one half of the egg, as is normally the 
case in the entodermic pole of an epibolic gastrula. It is possible in this case also to 
suppose that the delay is due to tardy multiplication of the nuclei, but this explana¬ 
tion seems less probable than the other. In some cases the small spheres are gradually 
constricted olf from the unsegmented part, and the egg may pass into a resting stage, 
leaving a number of spheres only half formed (see figs. 59, 60, 63, 64). This fact 
strongly indicates that there is some resistance to be overcome in the vitellus, for there 
can be little doubt that the half-formed spheres contain fully developed nuclei. 
There are a number of other facts which point in the same direction. The first 
formed cleavage furrows penetrate very slowly towards the centre of the ovum and, 
in some cases at least, do not reach the centre during the first stage of activity. The 
segmentation is at first, therefore, of the type which Balfour has termed centrole- 
cithal, the egg consisting of a peripheral layer of partially-formed cells and a solid 
central yolk-mass. The egg differs somewhat in structure, however, from a typical 
centrolecitbal ovum; for the central yolk-mass, so far as can be determined, does 
not contain at this period a greater proportion of deutoplasm than the peripheral 
parts, though it does so at a later stage. The failure of the cleavage furrows to reach 
the centre of the egg seems to be due either to the resistance being greater in the 
central parts, or to the exhaustion of the energy of the protoplasm before the inertia of 
the entire mass of deutoplasm has been overcome. 
It is interesting to compare Renilla in this respect with Clavularia on the one hand 
and Alcyonium on the other, as described by Kowalevsky. In Clavularia the resist¬ 
ance of the entire mass of deutoplasm would seem to be less than in Renilla, as the egg 
divides completely and regularly from the first. In Alcyonium, on the other hand, 
the resistance in the central mass is greater than in Renilla, and the segmentation 
does not affect the central portions of the egg for some time. Irregular protoplasmic 
protuberances separate themselves from the yolk to form segmentation spheres, wdiich 
after a time arrange themselves in a simple regular ectodermic layer. The central mass 
remains, for a considerable time, quite unsegmented, but finally breaks up into large 
rounded entoderm cells. Hence it appears that the cleavages do not reach the centre 
of the egg until the delamination takes place; and in this case the cause seems pretty 
clearly to lie in the greater abundance of deutoplasm in the central portion of the egg. 
The principle that unequal distribution of deutoplasm produces unequal rates of 
