E. MAESHALLT. REPRODUCTIVE ELEMENTS. 185 



discerned. Nor could I be absolutely sure that the part contain- 

 ing the nucleus had not been sliced away by the microtome- 

 knife. 



Are the above described cells in E. marshaUi really eggs? 

 Possibly they are, though I know nothing about their subsequent 

 development. While their homology with the large cells tliat I 

 have discovered in Leucopsacus orihodocus is questionable, I hold 

 it fairly j)robable that in them we have the identical cells which 

 were taken by Schulze for the ova of E. aspergilliim. The 

 little discrepancies between his descriptions and my own may be 

 explained by assuming that only comparatively young stages in 

 oogenesis have come under my observations ; whereas, some of 

 those observed by Schulze had a diameter more than ten times 

 greater than that in the largest seen by me. 



Here let me i-eturn to the archœocyte-congeries, in their 

 bearing on Veproduction in the Hexactinellida. I have endeavored 

 to show that, under certain circumstances, they may transform 

 themselves into masses of thesocytes (p. 174). Of the others 

 that do not undergo development in that direction, there may 

 possibly be some that represent an early stage in spermatogenesis 

 (p. 181). At the same time, I believe that a good part of the 

 primitive archœocyte-congeries are directly and actively concerned 

 in the formation of certain reproductive bodies, asexual or sexual 

 but other than spermatozoa. 



From Rhabdocalyptus mirabilis F. E. Scfî., wdiich multiplies, 

 similarly to Lophocalyx philippinensis, by producing small buds 

 on the external surface, F. E. Schulze ('gg, p. 63) has described 

 as the first Anlage of the bud a solid tissue-mass containing 

 small nuclei in tolerably dense and uniform distribution, and 



