ORIGIN OP HYPOBLAST IN PELAGIC TELEOSTEAN OVA. 33 
tiated and has grown down over the periblast, as is shown 
in the figure. An hour and a half afterwards (fig. 3) the 
periblast is seen to have pushed its way completely across 
the floor of the segmentation cavity, and now contains quite 
a number of free nuclei and cells. The blastodisc in spread- 
ing over the yolk has thinned out somewhat, but there is still 
no sign of invagination. Half-an-hour later (fig. 4), the cells 
in the periblast have accumulated under the rim of the blas- 
todisc ; on the left hand of the figure these cells are seen to 
be quite round and arranged in a row ready to take their 
places alongside the lowest layer of the blastodisc. The right 
hand of the same figure shows an abnormal form of the peri- 
blast in which the cells, nuclei, and surrounding protoplasm 
have been withdrawn into a pocket in the yolk, possibly caused 
by shrinking in the hardening process. The epidermic layer 
is still seen to reach some way over this mass of periblast. 
Half-an-hour later still (fig. 5), the periblast again covers the 
floor of the segmentation cavity more thickly, and free nuclei 
are to be observed rising from the yolk to help in building up 
this layer. The first row of hypoblastic cells are now seen 
attached to the lowest layer of the disc, but are recognised by 
their round clear outline. A little later again (fig. 6) two 
rows of hypoblast cells are seen in their proper place, and the 
living egg in this stage shows a clearly defined ring with the 
beginning of the prominence to form the embryonic shield. 
The outline of the original blastodisc is well marked off, and 
the new cells are sufficiently different in shape to mark off 
where one layer begins and the other ends. The epidermic 
layer is still seen in its original position and has taken no part 
in the process. The new layer is now formed rapidly, and 
nuclei and free cells are seen crowding up from the yolk to 
help in this work. Fig. 7 represents a little later stage in the 
embryo of Motella mustela seen in optical section where 
much the same process has been at work. Here, however, the 
resulting cells of the new layer are so much larger and clearer 
in outline than is usual in Teleosteans that the line of demar- 
cation between the old and new layers is quite distinct from 
VOL. XXV. NEW SEK. 
c 
