1884.] 73 [Hyatt. 



several times larger than in Halicbondria or the other species of 

 Hymeniacidotn or Suberites. 



There is also among our drawings the ovum of a species of 

 Hyniemacidon with one enormous azygos cell and two compara- 

 tively very small coronals, the azygos being more than twice the 

 size of both coronals combined, in the two specimens figured. 

 This agrees with Keller's observation upon the large size of the 

 primitive azygos cell in the bicellular stage of C. fertilis. 



The presence of an azygos cell at such an early stage, indi- 

 cates, that the unequal segmentation observed by Keller in Chal- 

 inula fertilis is, as surmised by that author, of general importance. 

 Barrois first noted the same facts among Keratose sponges. 

 (Embryol. de quel. Eponges de la Manche. Ann. Sci. Nat. s. 6.) 

 Tom. in. His larva in Aplysina Verongia showed unequal division, 

 and is figured at an older stage than Keller's Chalinula, being 

 already multicellular. Barrois, however, from differences in 

 color was led to conclude that the pigment-bearing cells of one 

 pole were the same as the dark-colored granular cells of the 

 basal area in still older stages, and Keller strengthened this state- 

 ment by following the history of the largest of the two primi- 

 tive segmentation cells of his larva. The primitive segmentation 

 in C. fertilis produces two cells, the upper the largest. These 

 split into four, the lower equally, the upper unequally. The 

 largest portion of the upper cell becomes central and remains for 

 a time undivided, the three cells below, equivalent to our coronals, 

 then divide more quickly and become the "ectoderm " cells. 



Barrois' figure of Halisarca lobularis (Ann. Sci. Nat. ser. 6, vol. 

 in, pi. 15, f. 2-1-25) agrees closely with these stages in Keller's 

 figures, and show, that, though no difference in color at the two 

 poles occur as in Keratosa, there is a difference in the cells. 

 Barrois notes that the three-celled stage is so prevalent in his 

 embryos that it is probably the next stage after the bicellular 

 thus agreeing with our observations in Suberites and Halicbon- 

 dria. He supposed that this differentiation did not take place in 

 Halisarca and the Halichondrida until a comparatively later 

 larval stage, but in this he was probably deceived by the 

 absence of distinct coloration. 



Carter's figure of the ovum of Harlisarca lobularis (Ann. 

 Mag. Nat. History, 1874, pi. 20, f. 4) has two cells with definite 



