Hyatt.] 74 [March 5, 



nuclei seen under compression, f. 5 has four cells and appar- 

 ently this is a truly four-celled stage, f. 6 on the other hand 

 shows six coronals and one azygos. The remaining stages ex- 

 hibit the multicellular morula.j 



In MetschinkofPs figures of the stages of segmentation of 

 Halisarca Dujardinii (Zeitschr. Wissen, Zool. vol. 32, pi. 20, f. 5-7) 

 bicellular, four-celled, and seven-celled stages are delineated. 

 All of these have in each cell a nucleus and in the seven-celled 

 stage nucleoli appear. The four-celled stages f. 6, and the seven- 

 celled f. 7 are both similar to the lower side of ova respectively 

 in the five- and seven-celled stages which we have seen as 

 described above. His f. 8, a section with twelve coronal cells 

 and f . 9 a view from the outer side of the same egg with ten 

 coronal cells show, that the egg had twelve coronals, and f. 9 

 exhibits five small cells above, which are probably due to the 

 division of the azygos cell or at any rate evidently represent the 

 endoblastic layer. 



Schultze (Zeitsch. Wissen. Zool. vol. xxviii, 1877, p. 31) denies 

 that he has seen in Halisarca lobularis the eight-celled stage of 

 Barrois with six coronal cells around the centre and the two others 

 at the top and bottom of the morula. This would be a direct 

 reversion to the Calcispongian embryo and is probably rare since 

 other authors have not remarked it. iSchultze also observed the 

 three-celled stage and attributes it to the division of the larger 

 cell of the bicellular stage, and a still further probable division 

 of the two coronals into four corresponding to our five-celled 

 stages and then adds, " Aehnliche Unterschiede lassen sich auch 

 bei den weiteren Furchungsstadien nachweisen." 



Sollas (Quart. Journ. Micr. Sci., 1884, no. 96, pi. 37, f. 1-7,) 

 gives in section stages of H. lobularis, which may also be com- 

 pared with the above, and his abnormal stage (f. 3 J with three 

 cells is evidently a true normal stage somewhat distorted. The 

 bicellular stage (f. 2) is figured as having one very large and 

 one small segment. This stage is described as occurring some- 

 times, and the five-celled stage as having the fifth cell central 

 and in the same plane with the four coronals, and sometimes in 

 the same circle with them. The whole of these prepara- 

 tions have the aspect of being abnormal, and similar to some 

 abnormal ova which we describe farther on. His structureless 



