Hyatt.] # 92 [March 5, 



pseudopodia extending out of the aperture above and forming 

 the usual sarcodic net-work. All of the Physemaria have to be 

 left for the present in the category of doubtful forms in which we 

 must also place the extraordinary Cameraphysema lately described 

 by John A. Ryder. (Smithson. Misc. Coll. Proc. Nat. Mus., vol. 

 xxn, 1883, p. 270.) 



Haeckel describes an apical cell as arising by vertical fission in 

 the eight-celled stage in Asculmis armata (Calcispongiae, pi. 13), 

 in Leuculmis echinus (ibid. pi. 30), and Sycyssa Huxleyi (ibid, 

 pi. 44.). Haeckel also considers the basal and apical cells as all 

 developed in the same plane, and this may have been the case, 

 though the figures of other investigators suggest that when the 

 apical cells were formed there was a change from the monoplacula 

 to the diploplacula. Haeckel's observations also show that the 

 sudden appearance of the eight apicals in Sycandra occurs 

 at a stage when only one apical is usually formed," and this 

 author even describes the apical as not divided into more than 

 four even in the sixteen-celled stage of the three si ecies named 

 above. No three-celled stage has been recorded in any Calci- 

 spongian, so far as we know, except in a form with irregular seg- 

 mentation of Sycyssa Huxleyi, figured by Haeckel (pi. 44, f. 

 9-13) and described by him as abnormal. 



The three-celled stnge is, therefore, not prevalent among Cal- 

 cispongiae as it is among Carneospongiae. The bicellular stage 

 in Calcispongiae is followed, as a rule by a quadricellular, and an 

 eight-celled stage before any change in the mode of segmentation 

 takes place. Haeckel also gives figures of apparently true gas- 

 trulae in Asculmis armata (pi. 13) Sencumis echinus (pi. 30), 

 and Sycyssa Huxleyi (pi. 44), but in all of these the ectoblast 

 cells are invaginated and the whole aspect of the larva identifies 

 it as the transient stage of gastrulation. In Ascetta clathrus 

 (pi. 4, f. 6-8, var. mirabilis), a closed hollow planula is figured 

 with only two layers of cells. This stage has not been observed 

 by others, and the aspect of the embryos we have mentioned 

 above leads to the suggestion that Haeckel's planula in these 

 cases is explicable as a transient gastrula, in which as we have 

 suggested the temporary blastopore has been obliterated by 

 contraction. 



Barrois' description of tiie vertical segmentation in the first 



