220 SCIENCE PROGRESS. 



posed, of " endodermal " origin. 1 The flagellated epithelium 

 of the larva is entirely used up in forming chambers. 

 The differentiated cells of the inner mass of the larva 

 form (i) all the flattened epithelium of the adult sponge, 

 (2) its contractile cells and sphincters, (3) the connective 

 tissue elements and spongoblasts. There remain only 

 the scleroblasts and the amoeboid and sexual cells, the 

 origin of which from the inner mass we have already 

 traced. 



The canal system thus formed soon acquires a com- 

 munication with the exterior by pores and an osculum, 

 and the organism becomes a young sponge. It is inter- 

 esting that both our authors are agreed in regarding the 

 pores as intracellular ducts, as described by Bidder and 

 myself in Ascons. 



The account given by Yves Delage of the metamor- 

 phosis differs from that of Maas, as has been said, with 

 regard to the details of the process, and he describes a 

 series of events which are without a parallel in the animal 

 kingdom, and which are certainly a tax upon our credulity. 

 We have seen that he describes the larva as consisting of 

 an outer layer of " cellules ciliees," with beneath this a 

 layer of "cellules epidermiques," and an internal nucleus 

 of "cellules intermediaires " and " amceboides," the last- 

 named elements being marked out by their large size, 

 coarsely granular contents, and vesicular nuclei. At the 

 metamorphosis, according to him, the epidermiques come 

 to the surface and form a layer covering the ciliees, this 

 change of position not being effected by any process of 

 epibole or overgrowth, but by each epidermic cell struggling 

 up to the surface independently and passing between the 

 ciliated cells over it, which separate to permit of its 

 migration. 



At the same time a most extraordinary process is 

 going on. The large amoeboid cells of the interior send 



1 In his first paper (1) Maas regarded the epithelium lining the exhalant 

 canals as derived, together with the chambers, from the ciliated layer of the 

 larva, a view which in his later paper (2) he abandons. 



