ENTRY OF ZOOXANTHELLE INTO OVUM OF MILLEPORA. 699 
if in some manner the vacuolated portion of the manubrium 
with its contained zooxanthellae were drawn in. 
That the medusz are liberated at this stage can be con- 
cluded from their mature appearance within the ampulle, and 
from the fact that one free medusa (fig. 10) was met with 
which exhibited an essentially similar structure. 
The other free medusz examined, possessed ova, more com- 
pletely vacuolated, and with more numerous zooxanthelle. 
The egg-cells showed various phases of an encroachment 
upon the manubrial substance, which eventually was almost 
entirely reduced (fig. 12), the ova at the same time becoming 
rounded off and similar to the extruded ova that were 
examined. 
During ali this period the zooxanthelle exhibit their 
normal appearance, and it can be observed that they divide 
fairly frequently within the ovum. ‘These cells have been 
figured by Moseley (’81) who was able to examine fresh mate- 
rial. He remarks that they closely resembled those of other 
Hydroids. ‘They contained irregular granules of a bright 
gamboge-yellow colour, the cell-contents frequently dividing 
into two, and sometimes, more rarely, into four. In the older 
portions of the colony the pigment was of dark-brown hue. 
IT show their structure as displayed in stained specimens 
(fig, 18). The spherical nucleus exhibited a mass of closely- 
packed chromatin granules. A pyrenoid was always present, 
the clear space around which, in most cases, gave the reaction 
for starch. ‘lhe pigment-bearing granules varied in number 
and size, did not always stain to the same degree, and in 
some cases had a little starch associated with them. The cell 
membrane did not respond to cellulose tests. I observed in 
a few cases division of a cell into four. Their average 
diameter was somewhat over 9 . 
That zooxanthelle pass into the ovum from the parental 
tissues appears to be undoubtedly the case from the foregoing 
evidence; but the part these play in the future economy of 
the animal remains to be solved. It may be that their en- 
closure is more accidental than physiologically necessary, for 
