1891.] Embryology. 289 
The fact that every young Scyphistoma was attached to a cyst, its stalk 
passing through the opening and spreading out on the lower flat wall, 
first aroused my suspicions, and I finally succeeded in observing the 
embryos leaving the cyst, and have sections through forms in various 
stages of emergence. Encystment has been observed by Kowalewsky 
in Lucernaria, but was supposed to be a precursor of death. No one 
has yet observed what I have mentioned above in any Scyphomedusa, 
but my preparations do not allow of any doubt as to its existence in 
C. arctica. 
Shortly after their emergence from the cyst the mouth forms, placing 
the internal cavity in communication with the exterior, and four ten- 
tacles make their appearance. I could not detect any invagination to 
form the mouth, such as Claus, and especially Goette, have described 
for other Scyphomedusz. My preparations show that the ectoderm 
and endoderm come into contact at the margin of the mouth opening, 
and that there is no stomatodeal invagination of ectoderm such as 
Goette maintains exists in Cotylorhiza and Aurelia. It is to be 
noticed that a similar absence of an ectodermal stomatodeum occurs 
in Lipkea ruspoliana, described by Vogt as the representative of a new 
tribe of sessile Medusze, but which, it seems probable, is simply a Scy- 
phistoma. 
With regard to the formation of the mesenteries of the Scyphistoma, 
my results are quite at variance with those of Goette. The young 
Scyphistomas with four tentacles show no signs of them; in older spe- 
cimens with the same number of tentacles traces of them are occasion- 
ally to be found ; but as a rule they are not formed till the young larva 
has acquired eight tentacles. It is unnecessary to state that in G 
arctica their formation stands in no connection with the formation of 
an ectodermal stomatodzum, since this structure does not exist. 
An account of the structure of the mesenteries, and the formation 
of the “ trichter’’ and of the mesenterial filaments, will be given in 
the complete paper.—J. PLavrair McMurricu, Clark University, 
Worcester, Mass. 
