1894.] Embryology. 75 
chambers). The remaining ciliated cells forming the syncytium, unite 
in a similar manner and build up flagellated chambers. 
The chambers are thus formed independently of any central space. 
The canals are likewise formed independently of each other, as so many 
irregular spaces, which gradually become lined with an epithelium con- 
sisting of “intermediary” cells, the remaining “intermediary ” cells 
becoming the stationary elements of the adult mesoderm. The union 
into a connected system, of chambers and canals, with the formation of 
pores and oscula complete the development. 
The conclusions enunciated in this paper as to the origin of the canal 
epithelium and the collared cells of the chambers differ from those pre- 
sented in the author’s previous notes on sponge embryology (Comptes 
Rendus, 1890, 1891), though the account of the formation of the adult 
epidermis remains thesame. On these points the paper is nearly in 
harmony with the recent contribution of Maas on the metamorphosis 
of Esperia (1892), though Maas derives both canal epithelium and 
collared cells from immigrated ciliated cells. The weight attaching to 
this harmony of observation on the development of Esperella, is how- 
ever, lessened by the direct contradiction in the accounts, given by these 
two authors, of the Spongilla development. In the latter sponge, which 
according to Delage agrees with Esperella, Maas (1890) has described 
and figured in a most detailed way the transformation of the larval 
ciliated cells into the flattened definitive epidermis. 
Germs Layers.—As the author points out, his discoveries make it 
extremely difficult to draw a comparison between the germ layers of the 
sponges and those of other Metazoa. The case of Sycandra he thinks 
already constituted a serious difficulty. In this he adopts the point of 
view of Balfour, and is accordingly perplexed to find granular cells, 
such as in other larvæ constitute the entoderm, here forming the adult 
epidermis, while the ciliated cells invaginate to form the epithelium of 
the paragastric cavity. But the difficulty reaches its height when he 
attempts to compare the larva of Spongilla with that of other animals. 
This larva is covered on the outside by cells, which eventually form the 
epithelium of the flagellated chambers. While the cells which will 
constitute the adult epidermis are in the larva situated in the interior, 
i. e. beneath the surface layer of ciliated cells. These facts place us in a 
dilemma. Accepting the ordinary views on the structure of the sponge 
body, we would call the epidermis of the adult, ectoderm, and the 
epithelium of the flagellated chambers, entoderm. Adopting this posi- 
tion, we reach on turning to the larva of Spongilla, the strange con- 
clusion that the swimming larva is covered with a layer of entoderm, 
