Hyatt.] 72 [March 5, 



each cell occupying about a third of the circumference. In 

 another specimen of Suberea at the same stage a similar 

 outline was observed, but the two coronal cells, as we propose 

 to call the ectoblasts at this and in succeeding stages of seg- 

 mentation, formed a sort of crescent, the azygos being rounder 

 and less pointed. In other specimens of the tame species on the 

 other hand the ovum was more elongated, and the azygos cell 

 was in outline like the carapax of a beetle, the coronal cells being 

 like the wing covers. The azygos cell occupied nearly one half of 

 the length of this ovum, and its longest transverse diameter was 

 about one-fourth longer than the longitudinal diameters of either 

 of the other cells. The base of the azygos cell occupied about 

 three-sevenths of the whole circumference, the bases of each of 

 the others two-sevenths. It was found by means of intermediate 

 forms, and by differences observed on the two sides of speci- 

 mens, that the first form could not be separated from the others 

 and one specimen showed an outline like Keller's Chalinula in 

 the bicellular stage, except of course that the lower cell had a 

 vertical line of division and the upper cell was smaller than the 

 .two lower. This indicated that certain doubtful bicellular forms 

 which we also observed in this species were true bicellular 

 stages. The ovum, therefore, has first transverse division as in 

 Keller's Chalinula, and probably other forms of Carneospongiae 

 and then multiplies the lower cells by vertical fission. 



In the next stages the coronal cells multiply by division, while 

 the azygos remains for a time single. Four coronals were pres- 

 ent in some specimens making with the azygos five cells, and in 

 others five and six coronals were observed, making six- and 

 seven-celled stages in all the species named above. 



In H. incrustans, similar stages were observed except that the 

 fifth and sixth coronals were very small in some specimens, 

 and in Hym. sulphureaV a sixth minute coronal arose close to the 

 azygos. The size of these coronals indicates, that they were due 

 to the unequal division of the larger coronals. 



No segmentation cavity appeared at this stage in Hal. incrus- 

 tans, Hym. caruncula, R. socialis, or S. suberea but in sections of 

 the ova of Hym. sulphurea? a large segmentation cavity was 

 distinctly seen, the ovum at this stage being in this species 



