210 BULLETIN OF THE UNITED STATES FISH COMMISSION. 



walls of the follicles, cells are to be found which have the nucleus so char 

 acteristic of the more mature ovules. These, I am inclined to believe, 

 are the representatives of what will later become ova, and not the rep- 

 resentatives of spermatoblasts. It is a singular fact that the spermato- 

 zoa have a tendency in 0. edulis to cling together in masses of about a 

 uniform size. Though the spermatic particles which comprise these 

 masses are somewhat separated from each other, if compressed together 

 they would evidently form a body about the size of the spermatoblasts 

 from which they were derived. Later they tend to break up and form 

 a more homogeneous, granular mass at the outlet of their parent tubule, 

 where the latter joins the outgoing efferent duct. While it is true that 

 some sections of 0. edulis show little evidence of the presence of any- 

 thing else but the product of one sex, it appears to me that there is 

 sufficient evidence of the hermaphrodite character of the generative 

 glands of the species presented by a pretty large series of sections 

 taken from about fifty individuals from different localities along the 

 coasts of Wales, Scotland, England, France, Holland, and Germany. 

 Sometimes a portion only of a section will be hermaphroditic, showing 

 that different parts of the generative glands of the same animals may be 

 of different sexes. The result of this arrangement is that it is scarcely 

 possible for the eggs to escape impregnation by the milt generated along- 

 side of them, and we may, I believe, fairly assume that Ostrea edulis is a 

 self-fertilizing hermaphrodite. 



The condition of things in the generative tubules of Ostrea virginica 

 and angulata is very different, as may be gathered from the following 

 account. In the first place I have never found any evidence of hermaph- 

 roditism either in the living animal or in sections of the reproductive 

 organs. The mode of pressing out the spawn from the gland and ducts 

 of 0. virginica, and the physical test used to determine the sex of the 

 products in practical work during the last season, afford the most posi- 

 tive demonstrations of the unisexuality of that species. Examining sec- 

 tions, however, we never find either in the reproductive follicles of 0. 

 virginica or of 0. angulata any evidence of the coexistence of ovules 

 and spermatozoa. In fact, the mode of spermatogenesis in the uni- 

 sexual species is very different from that of the hermaphroditic. As 

 indicated in Brooks' figure of a part of a section of a male oyster, the 

 spermatozoa are peculiarly arranged in the follicle or tubule. Upon 

 applying a high power (500 to 800 diameters) I find that the heads of 

 the spermatozoa show a very marked tendency to be arranged in rows 

 like beads and not in oblong clusters, as in the hermaphroditic species. 

 Moreover, the walls of the generative tubules are lined by relatively 

 very much smaller spermatoblasts than those found free in the repro- 

 ductive follicles of the hermaphrodite form. This spermatogenetic 

 layer is often very marked in the males of the unisexual species, and 

 even at an early stage of the functional activity of the testicular or- 

 gans presents much the same structure that it does later. The rows 



