THE EMBRYOLOGY OF THE PORIFERA. 231 



ciliated cells now develop into the collar cells of the 

 adult, reverting probably to their primitive condition in 

 the blastula. The inner mass of the larva becomes the 

 outer layer (ectomesoderm) of the sponge. Ascetta is 

 hatched in the true blastula stage, and on this account, 

 and also from the fact that its inner mass arises by 

 immigration, is perhaps more primitive in its mode of 

 development than any other form. Spongilla and the 

 horny sponges appear to be set free when the inner mass 

 is still quite internal ; from this condition we find tran- 

 sitions through the Halickondrina to a complete amphi- 

 blastula larva as in Sycon and Oscarella. No sponge is 

 hatched at a stage later than this. 1 



If we wish to compare the development of sponges 

 with that of other animals, and to discuss their relation 

 to the germ-layer theory and their position in the animal 

 kingdom, two courses are open to us. We may either 

 (1) start from a comparison of an adult sponge, such as 

 Ascetta, with an adult Ccelenterate, such as Hydra, in 

 which case we should term the ciliated layer endoderm, 

 the outer layer ectoderm; or we may (2) compare the 

 typical life-histories of a sponge and a Ccelenterate, and 

 then we should term the ciliated layer of the sponge larva 

 ectoderm, its inner mass endoderm. The former view was 

 Schulze's, the latter Balfour's. 



Maas is of opinion that, if the germ layers of sponges 

 and other Metazoa are to be compared at all, the ciliated 

 cells of the larva are ectoderm, the inner mass endoderm. 

 Only in this way is an explanation possible for the 

 unequal segmentation and the invagination of the macro- 

 meres into the micromeres, i.e., the wrongly termed 

 pseudogastrulation. If the^ opposite view be taken, not 

 only does the pseudogastrula remain inexplicable, but the 

 displacement of the ciliated cells into the former inner 



1 C/iona, the boring sponge, extrudes its ova, which develop outside 

 the sponge, but in other respects the development appears to be similar to 

 that of Cornacuspongue. The larva is completely ciliated. See Nassonow, 

 Zeitschr.f. wiss. Zoo/., bd. xxxix., 1883, pp. 298, 299. 



