740 DR. E. B. WILSON ON THE DEVELOPMENT OF RENILLA. 
segmentation which have been described. Thus, in the same specimen, some of the 
cells may be undivided and contain simple nuclei; others contain delamination 
amphiasters (7.e. those whose long axes are radially directed); while others have 
completely divided into ectodermic and entodermic moieties. Moreover, delamination 
cleavages may be in progress in some of the cells, while in others the cleavages are 
taking place in vertical planes. This is shown, for instance, in fig. 99, where two of 
the cells (a, 6) contain delamination amphiasters, as shown by the direction of their 
long axes, while a third cell (c) is about to divide in a vertical (or radial) plane, as 
shown by the position of the amphiaster. 
This suggests the interesting question as to whether delamination cleavages really 
take place in all of the cells, or may not rather be limited to the cells over a certain 
area. My sections are inconclusive on this point, which is of great importance in its 
bearing on the mode of transition between the invaginate and delaminate modes of 
development. (See Barrour’s ‘Comparative Embryology,’ vol. ii, p. 280, and my paper 
on the early stages of some polycheetous annelides in ‘Studies from the Biological 
Laboratory of the Johns Hopkins University,’ vol. ii, No. 2, 1882 ; compare also the 
very interesting observations of Ciaus on “ Die Entwickelung des Aequoriden-Eies,” 
Zool. Anzeiger, No. 112, June, 1882.) 
In fig. 94 one of the cells (a) is in the act of cleavage, and the direction of the 
amphiaster and the form of the cell indicate that the cleavage is in a horizontal 
plane—.e., isa delamination cleavage. Fig. 96 represents a section (osmic acid) through 
an egg, a little later, in which the inner portions of several of the spheres are separating, 
or have recently separated, as entoderm cells. Unluckily, the nuclei do not appear 
in the section, which is furthermore somewhat disfigured by shrinkage cavities. 
In a few cases I have observed at a much earlier period divisions of the nuclei, 
which may possibly represent delamination cleavages. Such a case is shown in fig. 
98. The egg is about to divide into sixteen spheres, but contains two amphiasters, 
which have the same position as the true delamination amphiasters already described. 
It seems possible that the inner star of each amphiaster is destined to form the 
nucleus of an entoderm cell, and the outer star that of an ectoderm cell when division 
of the vitellus takes place. Ihave not traced this out, however, and the appearance 
may be open to a quite different interpretation. It is, however, certain that there is 
a good deal of variation in the delamination process, and the embryos do not display 
the beautiful regularity in this respect which has been described in some Ccelenterate 
eggs. As already mentioned, Leptogorgia appears to differ from Renilla in this 
respect, since the period of delamination is sufficiently marked to produce a special 
“active stage,” represented externally by the simultaneous swelling of all the spheres. 
At the close of the delamination process, the egg consists of a solid mass of cells 
in which every trace of the segmentation cavity has disappeared. As shown in figs. 
99 and 100, the ectoderm does not at first form a distinct layer, the cells dovetailing 
with those of the central entodermic mass. As the egg passes into the resting stage, 
