No. 2:] LE: ORIGIN OF THE SEX=CELLS. 219 
I have carefully observed about fifty young hydroids of 
Hydractinia, all of which were nutritive polyps. I, therefore, 
infer that this polyp is always developed before the protective 
and reproductive forms. In one instance, I saw a hydroid 
about nine days old with its hydrorhiza, from which a re- 
productive polyp (Fig. 67, R), with its knob-like tentacles, 
had developed. Unfortunately, it was mutilated in being 
removed from the dish, to which it closely adhered. There- 
fore it became impossible to examine it thoroughly, and thus 
to give absolute proof of the fact that the reproductive forms 
are developed in this way, and at such an early period. 
ENDODERM FORMATION. 
Since the form of cleavage is centrifugal to such a marked 
degree, there is a tendency in the very early stages of seg- 
mentation for the cells to arrange themselves as a layer around 
a central cavity ; which has also been noted by Metschnikoff 
for Clytia, Liriope and Nausethoé (p. 41). In Hydractinia, 
when the embryo consists of about sixteen cells, the blastula 
is formed. By a careful study of a number of serial sections 
karyokinetic figures are observed at this time, which seem to 
indicate a delamination to form the endoderm (PI. III, Fig. 48). 
In Figs. 49, 50, are shown two sections of a series in which 
the spindles are seen in different cells. These show that the 
endoderm formation takes place by multipolar delamination, 
even at this early period in the development. The inner 
portion of these cells are cut off, and pass into the blastula 
cavity. During this period,by karyokinetic division, the cells of 
the blastula wall are also increased in number (Fig. 51). Karyo- 
kinetic figures are still to be observed in later stages in cells 
situated at different points of the blastula (Figs. 52, 54), showing 
that multipolar delamination is continued for some time, until 
in cross section, the embryo has its outside wall consisting of 
about twenty cells. 
In some sections we see that a cell will appear as though 
about to migrate into the cleavage cavity (Fig. 54, 4). If this 
took place we would have, according to Metschnikoff, a mixed 
