THE POSITION OF SPONGES, ETC. 449 



of the larva and adult correspond, and a terminology which 

 implies any such correspondence is to be avoided. Both 

 for these reasons, and on account of the homologies implied 

 by the words " ectoderm," " mesoderm " and " endoderm," 

 I have proposed to revive for the layers of the adult sponge 

 the purely descriptive terms originally coined by Haeckel 

 (1872, vol. i., p. 130), namely, the term "gastral layer" to 

 denote the internal collared epithelium and the term "dermal 

 layer," to denote the remaining tissues. Thus the gastral 

 layer corresponds in Ascons to the wholeof the "endoderm" of 

 authors and in other sponges to a part of it ; the dermal layer 

 is the "exoderm" of Haeckel, the "ectoderm" with the "meso- 

 derm," and sometimes part of the endoderm of other authors. 



The larva, as we have seen, consists of an external 

 ciliated layer and an inner cell mass. How do these layers 

 correspond with the layers of the adult ? Which of the two 

 layers of the larva is " gastral " and which is " dermal " ? 



In a former number of this journal (1894) I discussed 

 the question of the relation of the layers of the sponge 

 larva to those of the adult at some length, especially with 

 references to the works of Maas (1892 and 1893) and 

 Delage (1892). I showed that the works of these two 

 authors, differing much in respect to details, but agreeing 

 on the points of fundamental importance, rendered it 

 possible for the first time to treat the embryonic develop- 

 ment of sponges generally from a uniform and reasonable 

 stand-point. Both authors, in fact, agreed in affirming that, 

 contrary to the usual belief, the ciliated layer of the larva 

 of Cornacuspongise passed into the interior at the meta- 

 morphosis and became the gastral layer of the adult, that is 

 to say, the collar cells of the chambers, while the inner mass 

 of the larva trave rise to the remaining tissues of the sponge, 

 namely, to the flat epithelium lining the exterior and the in- 

 current and excurrent canals, and the whole of the so-called 

 mesoderm, in other words to the dermal layer as here de- 

 fined. Thus it became at last possible to reconcile the 

 apparent contradictions resulting from a comparison be- 

 tween the well-known development of Sycoit and that of 

 other sponges. There remained however certain cases 



