Hyatt.] 76 [March 5, 



blastula is apparent and it is probable that all others are also 

 really amphiblastulae. The conclusion is also sustained by the 

 appearance of a few larger cells around the blastuiapore in some 

 specimens. It is evident, however, that such a form contains in 

 part at least the inclosed elements of a mesoderm and is really a 

 primitive Parenchymula, and not a primitive Planula. 



The walls were either double or single on different sides of 

 the same amphiblastula in Halichondria ; when double composed 

 in one place of large cells and in another perhaps of small 

 cells of irregular shape, or even mere granules. The inner 

 cells alone are replaced by granules which are evidently 

 derived from the bursting or disintegration of these cells. 

 Whenever two or more of the internal cells occurred at the edge 

 of the blastuiapore the aspect was such as to entirely mislead the 

 observer into the belief that he was looking at a true gastrula. 

 This was the case in a blastula of Tethya hispida where my 

 drawing shows, however, that only one side of the blastuiapore 

 really had a double row of cells, the other side being composed 

 of a single row. Another drawing of this same species had in 

 optical section from the side precisely the appearance of a true 

 double-walled gastrula, and yet I have no doubt that it was 

 simply a case in which delamination had occurred more exten- 

 sively than is usual at this stage. The blastula was not a 

 closed sphere in these specimens, and each undoubtedly had a 

 distinct large opening at the smaller end surrounded in Hal. pan- 

 icea by four large cells. A similar blastula and opening was 

 seen in a species of Reniera, but having a single layer of nucle- 

 ated cells of regular shape and size. In another specimen of the 

 same species, however, about one hundred morulae were exam- 

 ined, many by sections. All were composed of large cells, the 

 interstices filled with granules and small cells. Among these no 

 two were exactly similar on account of great variations in size of 

 the cells, though the ova were all of about the same size, and 

 found in the same sponge. All of these were solid to the centre, 

 except one, which was hollow and had only a single layer of cells 

 of variable size but no granular protoplasm. This stage may 

 have been a true blastula. It had an exceedingly small aper- 

 ture but this was not like a true blastuiapore, and may have been 

 produced by accidental rupture. These ova were at least twice 



