228 SCIENCE PROGRESS. 



Metschnikoff interpreted these appearances as a gradual 

 flattening of the flagellated cells to form the outer cell 

 layer of the adult. A precisely analogous misinterpretation 

 of such appearances was made by Maas himself, who, in his 

 former work 1 on the development of Spongilla, described 

 the ciliated cells as flattening to form the ectoderm of the 

 adult, though he now recognises that the appearances seen 

 were due to the cells of the inner mass growing over the 

 flagellated cells. Moreover, O. Schmidt, in deriving the 

 adult ectoderm of Ascetta from the flagellated epithelium 

 of the larva, was obliged to assume a fusion of several 

 flagellated cells into the large amoeboid cells of the young 

 sponge, which is, to say the least, improbable. 



For all these reasons we are justified in saying that the 

 derivation of the ectoderm of Ascetta from the flagellated 

 cells of its larva is neither proved, as a matter of fact, nor 

 probable in theory, since only on the contrary supposition 

 would it be possible to compare the larva of Ascetta with 

 the larva of Sycon and siliceous sponges. A study of the 

 literature of sponge embryology shows that nearly all investi- 

 gators have been dominated by the idea that ciliated cells 

 in the larva must become ectoderm in the adult, perhaps 

 from a false analogy with Ccelenterate development. This 

 superstition, as one might call it, was first overthrown by 

 Metschnikoff and Schulze in Sycon twenty years ago ; it 

 still remained, however, to confuse sponge embryology, 

 until Delage and Maas showed its falsity in the case of 

 siliceous sponges. It now remains for some fortunate 

 investigator to establish the origin of collar cells from 

 flagellated cells in the larva in the few cases where asser- 

 tions to the contrary still hold the field. 



After what has been said with reference to Ascetta it 

 is hardly necessary to discuss the developments of Hali- 

 sarca and Plakina, where the appearances seen seem to 

 have been similarly misinterpreted. The development of 

 Oscarella, however, requires a brief discussion. Here the 

 larva is hatched as an oval blastula composed of a single 



1 Op. cit. 



