440 The American Naturalist. [May, 
the ciliated epithelium of the larva, the latter the rer mass. Butin 
the larva of the horny sponges, as in that of Spongilla, the ciliated 
epithelium is continuous over the whole surface. This is explained by 
supposing that in these types the ciliated epithelium (micromere layer) 
completes its growth around the inner mass, which in the other spon- 
ges is left bare at the posterior pole. 
In the metamorphosis of the two-layered larva of the cornacuspon- 
giae, a complete inversion of the layers take place. The ciliated cells 
draw in their cilia, and migrate into the interior of the sponge where 
they form a compact mass, surrounded by the former inner layer. 
Certain of the differentiated cells of the latter layer unite to form the 
thin epidermis of the adult. The boundary line between the rest of 
this layer and the inner mass of once ciliated cells gradually disap- 
pears, elements belonging to both layers becoming distributed irregu- 
larly throughout the sponge body (process of “ durchwachsung ”). 
Groups of the ciliated cells now begin to develop into flagellated cham- 
bers. Independent spaces or lacunae appear and become gradually 
lined with an epithelium formed by the differentiated cells of the lar- 
val inner layer. These spaces are the canals. Connection between 
them and chambers is subsequently established. In two points Wass 
differs from Delage, in his account of the metamorphosis. Delage 
believes a special layer of cells, the epidermic cells, can be distinguish- 
ed in the larva, which during the metamorphosis, take the place of the 
ciliated cells as a superficial covering. Wass finds no ground for dis- 
tinguishing the cells which thus form the adult epidermis, from the 
other differentiated cells of the larval inner mass. Again Delage be 
lieves that during the metamorphosis the undifferentiated cells en 
amorba like, the smaller ciliated cells, subsequently letting them E 
free to form the flagellated chambers. Wass disbelieves 1m this 
remarkable process, though he grants the possibility of amoebol 
occasionally engulfing ciliated cells, which however are never after 
liberated, but undergo degenerative changes (i. e. are digested). his 
On going over the Spongilla development the author, aided by m 
discoveries in marine sponges, finds that a different interpretation 
many particulars is to be put upon his earlier observations. bore 
mentation is not equal, but unequal. A true morula is not "e 
but instead the smaller blastomeres surround the larger. The aa 
epithelium of the larva is not transformed into the adult paea 
the inversion of layers described above takes place in Sp m 
also. The exhalent canals and flagellated chambers are not formed 
diverticula from a single main cavity, nor are the inhalent canals form 
