284 M. M. Mercaur 
Lecer & Dusoscg (1904a) describe a second type of “endo- 
genous cysts” for O. ranarum. In an ordinary multinucleated indi- 
vidual, a bit of protoplasm containing one to four nuclei is said to 
isolate itself from the rest of the protoplasm and acquire a cyst wall. 
The cyst so formed is extruded from the body of the parent. This 
description, which is very brief and unillustrated, needs confirmation 
before it can be accepted. No other student of Opalina has seen 
anything of this sort. 
The formation of the gametes. 
Longitudinal division is observed almost as soon as one finds 
the animals hatching from the cysts, (Figs. 146, 147, 153, Pl. XXII). 
I have found no way of determining how many divisions take place 
before the gametes are formed. The minute animals do not live long 
enough outside the body of the tadpole to allow one to directly 
follow the phenomena from the time of hatching from the cyst to 
conjugation. Size is not a safe criterion, for the cysts and the ani- 
mals that hatch from them are of various sizes; so also are both 
the micro- and macrogametes. Time relations have failed to deter- 
mine the point, for one cannot be certain how long a time is re- 
quired for one division, and the time when one observes copulation 
is different in different infections. 
In animals removed from the tadpole and placed in salt solution 
with a bit of the rectal wall and some of the rectal contents, di- 
visions just begun before removal from the host require generally 
from two to twelve hours to complete even under the most favo- 
rable conditions; in many other cases division is not completed at 
all, even pairs which upon removal from the host were almost di- 
stinct remaining unseparated after more than twelve hours. It is 
not improbable that division under natural conditions in the rectum 
may be more rapid than in even the most favorable artificial cul- 
tures. In one series of infections both micro- and macrogametes 
were abundant after forty-two hours and frequent instances of co- 
pulation were observed. Generally copulating pairs are abundant 
fifty to eighty hours after ingestion of the cysts. In material five 
and a half days after feeding the cysts I have found many zygotes 
but not copulating pairs. The average time required between in- 
gestion of the cysts and copulation is therefore uncertain, as is also 
the number of divisions that intervene, if indeed the number be 
constant. In the multinucleated Opalinae the number of nuclei in the 
