348 Metschnikoff on Germ-Layers. [April 
spheres, which in some cases would represent swollen-up placulæ, 
in others proplaculæ that have become spherical. In the devel- 
opment of existing animals the placula, according to Bütschli, 
appears as the flattened blastosphere stage found in Cucullanus 
(Bütschli), Rhabdonema (Götte), Lumbricus, Chiton (Kowal- 
evsky), Phoronis, and Ascidia mentula. Among adult animals it 
is represented by Trichoplax adhærens F. E. Sch. But Bütschli 
does not perceive that the flattened blastospheres of the Meta- 
zoa just mentioned agree with his placula in external form alone, 
and not in any essential or morphological respect. The funda- 
mental difference between the two lies in the fact that the two 
layers of the former have not been acquired by cell-division par- 
allel to the surface, which is the essence of the placula. In Pho- 
ronis, Ascidia, and, generally speaking, in the other animals cited 
above, the placula-like stage is attained by the flattening of a 
previously more or less spherical blastosphere, and not con- 
versely as the theory requires. According to the theory- the 
delamination of the Geryonidz, accomplished by a transverse 
division of the blastoderm cells, is a process similar to the for- 
mation of a placula or amphiblastula, such as is supposed to 
occur in other animals. If this be true, we should find in the 
formation of this so-called placula a transverse division of the blas- 
tomeres. But this is not the case. The endoderm cells of the 
flattened blastospheres are not split off from the ectoderm cells 
immediately above them, but arise by the longitudinal division 
of parent cells. We are thus forced to the conclusion that a 
placula stage does not appear in the development of existing 
animals endowed with a regular segmentation. There is, how- 
ever, a degree of similarity between the placula and a certain 
stage in the development of Ctenophores, where the endoderm 
has the form of a plate, and is covered by an interrupted layer 
of ectoderm. But this stage will scarcely be looked on by any 
one as embodying a primitive condition, and cannot, therefore, 
afford any basis for a morphological generalization. 
If, going farther back in the development, we regard the eight- 
celled embryo formed by the transverse division of the first four 
blastomeres as a short-lived placula, we thereby gain nothing; 
for we must bear in mind that the eight-celled stage of the de- 
_ laminating Geryonide is in all respects like the same stage of 
_ the invaginating Acraspeda, and must therefore be regarded as 
