1894.] Embryology. 441 
ed as invaginations of the “ectoderm,” but both sorts of canals arise as 
independent lacunae, subsequently acquiring an epithelium and con- 
necting together, and the chambers are formed from groups of the 
immigrated cells. The development of Spongilla is thus brought into 
accord with that of the marine cornacuspongiae. 
In a comparative review of the various types of sponge development, 
the author points out the fundamental similarity between the develop- 
ment of the cornacuspongiae and that of the caleareous sponges, as 
exemplified in Sycandra. The ciliated cells are homologous in the two 
kinds of larva, as are the granular cells of the amphiblastula and the 
inner mass of the other larva. The difference in the character of the 
metamorphosis arises from the fact that in the amphiblastula there is a 
large cavity, while in the larva of cornacuspongiae there is none. In 
this comparison Wass and Delage agree. 
The author thinks the development of those sponges (Ascetta, Osco- 
rella, Plaxira, etc.), which apparently differ from the plan of develop- 
ment described in this paper, needs to be worked over. A fundamen- 
tal harmony with the development of cornacuspongidae and Sycandra 
will be revealed. 
Touching the relationship between sponges and the other metazoa 
the author, without dogmatizing, is inclined to believe that they had a 
common ancestor above the protozoa. This ancestor is represented in 
the two-layered larva of both. But the community of origin goes no 
higher than this simple form—the sponges are not coelenterates. In 
the two-layered ancestor of the sponges, the superficial ciliated cells 
migrated into the interior, resigning their function of locomotive 
organs in order to generate internal currents of water, made necessary 
by the adoption of a fixed habit of life with subsequent increase of 
bulk. In other metazoa, the ciliated cells continue to form the super- 
ficial covering of the body. The immigration of the ciliated cells in the 
larva of cornacuspongiae, and the invagination of the ciliated cells in 
the Sycandra amphiblastula, are the ontogenetic expression of this 
chanye of position of the ciliated cells in the early ancestors of spon- 
ges, and have nothing to do with a process of gastrulation—the two- 
layered embryo being already formed before the occurrence of this 
immigration or invagination. 
H. V. Witson. 
