DR. E. B. WILSON ON THE DEVELOPMENT OF RENILLA. 739 
is important to note that whether eight or sixteen spheres are formed at the first 
cleavage, each contains a single nucleus only; for this shows that division of the 
vitellus does not always occur at the same stage in the division of the nuclei. 
The prominences soon become very marked, increasing in size at the expense of the 
central mass and becoming at length of a pyriform shape. The egg then consists of a 
central unsegmented mass containing no nuclei, and surrounded by partly formed 
spheres, each containing a single nucleus and connected by a broad isthmus with the 
central mass. In some cases, at any rate, the embryo now passes into a resting stage, 
as shown in fig. 95. The spheres flatten together and are separated by very distinct 
narrow clear spaces, which terminate abruptly some distance from the centre, thus 
leaving the central mass quite unsegmented, and continuous with the mass of the 
partially-formed spheres. In the figure some of the spheres appear to be completely 
separated from the central mass, and this may perhaps be the case with some of them. 
Others however are certainly not separated from the central mass. 
In the second stage of activity, or perhaps in some cases in the first, the spheres 
increase still further at the expense of the central mass, which becomes at length 
reduced to a very small remnant (fig. 92), to which the spheres are attached by narrow 
necks. Finally even this remnant disappears, and the completely formed spheres 
extend to the very centre of the embryo. A small mass of granular matter still 
remains in the middle of the embryo, and the spheres are attenuated at their inner 
ends (fig. 93). No segmentation cavity exists at this stage, but the inner ends 
of the cells soon become evenly rounded and a small segmentation cavity is formed 
(fig. 94), in which a quantity of granular débris usually remains. The spheres are 
destitute of cell-membranes, but are separated by a small quantity of intercellular 
substance. Their substance is completely similar to that of the unsegmented egg, the 
nuclei have the same appearance as in the latter, and are situated in the outer halves 
of the cells. The clear peripheral zone observed in the unsegmented egg is still very 
distinct in some specimens, but in others cannot be seen. It does not follow the 
lines of cleavage into the interior of the egg. 
c. Formation of the layers. 
The egg is now in the condition of a blastula in which the cells are not yet dif- 
ferentiated into ectoderm and entoderm. In the next change—which constitutes 
perhaps thé most important epoch in the development of the larva—the ectoderm 
and entoderm are separated by a process of delamination ; 7.e., the inner end of each 
sphere separates as an entoderm cell from the outer portion which remains as an 
ectoderm cell. A careful study of my sections taken in connexion with the external 
appearances, leaves no room for doubt that this is the mode in which the layers are 
separated ; but it is clear that the cells do not in all cases perform the delamination 
cleavage simultaneously. On the contrary there appears to be much irregularity in 
this process, which is not surprising in view of the other remarkable variations in the 
