1887] Metschnikoff on Germ-Layers. 347 
sphere composed of similar cells, Balfour(28, 29) adopts the 
amphiblastula as the transitional form between the Protozoa and 
Metazoa. I will therefore speak of his view as the Amphiblastula 
theory. Since it must be looked on as a modification of the Gas- 
trea theory, it is open to the same objections as the latter. It is 
quite unable to explain the phenomena which occur when the 
endoderm does not arise as a single continuous structure, but as 
cells separated from one another by intervening ectoderm cells 
(the case of multipolar immigration especially). As regards the 
application of Balfour’s theory to the Sponges, its untenability 
is shown by the fact that in many Sponges (especially Calci- 
spongiz and Halisarcinz) the endoderm is a nutritive layer. 
This fact, already emphasized by early investigators and more 
than once by myself, has lately been confirmed by K. Heider (30) 
on Oscarella lobularis. The objection here raised, moreover, up- 
sets the arguments which Balfour used to prove the isolated 
position of the Sponges among Metazoa. 
Akin to the Amphiblastula theory is the Placula theory of 
Bütschli (22), not only because the latter author also believes in 
the separate descent of the Sponges, but because the placula 
in many respects may be considered as a flattened-out amphi- 
blastula. Bütschli appreciates the weak points of other theories 
which deal with the genealogy of the germinal layers, and at- 
tempts in a purely diagrammatic way to construct the connection 
between invagination and delamination. He deduces both meth- 
ods of forming the endoderm from the modification of a primary 
placula form. Abandoning the starting-point of other views,— 
the spherical colony of Flagellata,—our author adopts as his 
primitive form a Gonium-like one-layered plate, which for con- 
venience I shall style proplacula. “It therefore seems fair to 
assume that the two layers first arose in a protozoan colony, the 
cells of which were arranged in one plane so as to form a one- 
layered plate. All the cells then divided parallel to the surface, 
and there thus arose a two-layered plate, the layers being prob- 
_ ably as yet undifferentiated. To this stage of a two-layered plate 
we will give the name of placula.” Such a placula, by assuming 
a sac-like form, became changed into a gastrula. In other cases ` 
the proplacula, in consequence of a secondarily retarded cell- 
division, gave rise to a delaminate blastosphere. As a result of 
these assumptions there would exist a radical difference in blasto- 
