448 SCIENCE PROGRESS. 



cells which therefore protrude posteriorly (Minchin, 1896). 

 In Oscarclla the two-layered condition is formed by a simple 

 invagination. Finally in the highly specialised Cornacu- 

 spongiae the segmentation results in a solid morula or ster- 

 roblastula, and the two embryonic layers are differentiated 

 in situ. In whatever way the two layers arise, the result is 

 a two-layered larva presenting in most cases striking re- 

 semblances to the coelenterate planula. It consists of an 

 external ciliated layer more or less completely surround- 

 ing an internal mass of amoeboid, granular, non-ciliated 

 cells. When food yolk is present in any quantity it is con- 

 tained in the cells of the inner mass. Hence in this point 

 as well as in their relative position and general character- 

 istics, the outer and inner layers of the sponge larva are 

 quite similar to the ectoderm and endoderm of Coelenterata 

 and other Metazoa. 



So far the way is clear, and the facts are scarcely dis- 

 puted, even by advocates of the most opposed views. It is 

 the metamorphosis of the simple sponge larva, that is to say, 

 the fate of its component layers, and the part they take 

 in building up the adult sponge, which is the important 

 question to decide. 



Before proceeding, however, to discuss this question it 

 is necessary to be clear as to the terminology we employ for 

 the cell layers under consideration. The words ectoderm, 

 mesoderm, and endoderm, have been universally employed 

 in sponges, sometimes in a purely descriptive sense, but 

 more often as implying a homology with the similarly 

 named layers in coelenterates. Not only is such a homol- 

 ogy more than doubtful, but even in a descriptive sense the 

 use of these terms is in the highest degree confusing and 

 misleading. For in the larva the ciliated layer has been 

 termed the ectoderm, and the inner mass the endoderm or 

 endo-mesoderm, while in the adult the term ectoderm has 

 been applied to the external, and more or less of the internal 

 epithelium, the term endoderm to the rest of the internal 

 epithelium, and the term mesoderm to all the tissue enclosed 

 between the epithelia. It is, however, far from certain, and 

 indeed highly improbable, that the similarly named layers 



