EAELY DEVELOPMENT OF GRANTIA COMPEESSA. 285 



mesoglea and get, into amcebocytes, we are then forced to believe that each 

 sperm-bearing amoebocyte, whatever its primary position, always carries the 

 sperm to one side of the egg ; there would then undoubtedly be cases where 

 the amoebocyte would have to perform a comparatively long journey with the 

 opposite pole of the egg as its goal. Against this amoebocyte view there is a 

 fourth reason : cells carrying sperms can never to my knowledge be found 

 on the short " side " of the egg, as marked by the star on the right of text- 

 fig. 3. Had the sperm-bearing cell been a mesogleal amoebocyte I would 

 surety have found one with a sperm, in positions other than the gastral side 

 of the oocyte (text-fig. 4). 



Reference to Professor Dendy's pi. 24. figs. 50 and 51, and comparison with 

 fig. 52, will show that the cell which carries food to the oocyte is not only 

 bigger than that which carries the sperm, but has different cytoplasmic 

 contents. In Prof. Dendy's fig. 52 the sperm-carrying cell (NO) is about 

 the same size as a collar-cell. 



To explain my text-fig. 4 satisfactorily on the basis of the amcebocyte 

 view, one would need to assume that there was a row- of waiting amcebocytes 

 ready in the position in which the sperm is later found. No author who has 

 drawn sponge-oocytes has figured such amcebocytes in this region, and I 

 have never seen an amoebocyte waiting in this part of the gastral epithelium. 

 To assume that the sperm pushes its way into the mesoglea till it finds an 

 amoebocyte to carry it to the egg is preposterous, for if the sperm can push 

 its way, " under its own power," into the mesoglea, there would be no need 

 for another cell to carry it to the egg. 



I think it will be obvious to all that the sperm does penetrate into a collar- 

 cell, cannot itself pass through to the oocyte, but is carried by the collar-cell, 

 which, during the process, may occasionally become somewhat altered in 

 appearance. 



(e) Germ-cells and Sex in Sponges. 



My observations lead me to consider that all Grantia individuals are 

 probably potential hermaphrodites, but that the occurrence of sponges with 

 sperm-stages is somewhat rare. Tn only one case out of twenty-five that I 

 know was the sponge a positive hermaphrodite, and in all the other examples 

 I examined only oogenesis stages were discovered. 1 am led to consider that 

 Grantia compressa is either a simultaneous hermaphrodite, or female, both 

 sorts of individuals being found. The sponges which I found to contain 

 oogenesis stages alone might have been either protandrous or protogynous 

 hermaphrodites, but this seems rather improbable, for at the time I fixed my 

 material the bulk of the oocytes were ripe and being fertilized, while there 

 was a general absence of sperm-stages, except in one individual, which 

 contained both spermatids and oocytes being fertilized. 



