1884.] 71 [Hyatt. 



The ovum is isolated and the cyst in the mesoderm is entirely- 

 filled by it often even in specimens prepared in alcohol and 

 glycerine. A membrane with double contour makes its appear- 

 ance at an early stage, but the cells in this were not seen until 

 the multicellular morula was formed. In Halichondria the cells 

 of this envelope became large and a chorion was formed, 

 tearing with difficulty at this stage notwithstanding its del- 

 icate hyaline aspect. The outer surface, however, is roughened 

 by the projecting cells which form a close-set epithelium-like sur- 

 face at this stage in Halicondria and Chalinula oculata, and 

 arbuscula, Reniera and Suberites. In some specimens of Sube- 

 rites more than one ovum was sometimes included in the same 

 cyst and might lead to serious errors in studying segmentation 

 since they are liable to be closely compressed, and irregular in 

 number. 



We did not see the aggregation of amoeboid cells to form 

 the chorion in Halichondria or Suberites as shown by Keller 

 in his clear delineations of this stage in Chalinula, but as these 

 cells are at fii st very small and transparent they are easily over- 

 looked. 



The eggs which had undergone segmentation had cells filled 

 with granules, similar to nucleoli in aspect, which obscured the 

 nuclei. The earliest stage observed in Hymeniacidon caruncula, 

 Halichondria incrustans, Reniera socialis, and Suberites suberea, 

 consisted of three cells, an azygos, and two bilateral cells which 

 occupied the lower half of the ovum. In a morula of H. incrus- 

 tans the egg had three segments, as if an orange were split 

 longitudinally into three sections, the azygos cell being angular 

 and reaching to the centre, but smaller than either of the other 

 two, which came up on either side to the periphery. The 

 base of the azygos cells occupied only about a fifth of the whole 

 circumference of the ovum and each of the other cells two-fifths. 

 It is, however, evident from our drawings that great variations 

 in the form and size of the segmentation cells exist in the same 

 species. 



The azygos cell in Suberites suberea was in most cases fusi- 

 form in section, and extended somewhat beyond the centre 

 of the egg, while the two lateral cells reached up on either 

 side forming the halves of a thick divided crescent, the base of 



