[7] EMBRYOGRAPHY OF OSSEOUS FISHES. 461 



the ovaiy, in Teleosts in wliicb this organ is lobulated, I have never 

 seen auj' evidence of the tnbulation met with in some vertebrates. With 

 the male organs it is different; here seminal tnbnles are developed in a 

 very distinct manner. The lobnles in an immatnre state in reality ap- 

 pear to represent folds of the germinal epithelium, on the exterior sur- 

 face of which the ova develop in their sacks, which rupture when the 

 ova are mature, allowing the latter to fall into the intraovarian space, 

 where the ovary is a closed saccular organ, or into the abdominal cavity, 

 as in Salmonoids and lampreys. The ovarian leaflets or lobules in an. 

 immature state show a very distinct median vascular stem, from which 

 the blood supply for the individual follicles is derived, as may be seen 

 in the immature or undeveloped ovary of Alosa sapidissima. The fol- 

 licles themselves serve at once to contain the growing egg, and by mean* 

 of a net-work of fine capillary vessels, which traverses its substance, to 

 supply it with proteids, an accumulation of which the egg reall}'' repre- 

 sents. The follicle may be greatly modified, as in Gqmbusia imtrueliSy 

 a viviparous form. Here it is apparently structureless, as far as I have 

 been able to make out; but in reality it is probably covered by a 

 layer of much flattened cells, which, together with an extremely thin 

 vitelline membrane underneath, form the walls of the follicular capillary 

 net-work, the blood cells within which, with their nuclei, are clearly 

 shown in hardened and stained preparations. But these are the only 

 histological elements which can certainly be made out. The capillary 

 net- work is distributed from a thickened annulus at one pole of the fol- 

 licle, where the afferent nutrient vessel enters and the vein passes off. 

 Through the annulus above alluded to there is an opening, the follicu- 

 lar pore, which is variable in size, and answers to the micropyle of the 

 eggs of other fishes. Through this pore the milt of the male apparently 

 finds access during the act of copulation. Impregnation is thus accom- 

 plished within the ovary of the female, and development of the embrya 

 proceeds as in oviparous species. The follicle, however, now acquires 

 a new function, in that it not only serves to develop the egg until a 

 mobile embryo is produced, but also functionates partly as the embry- 

 onic envelope, and partly' as a respiratory structure, by means of which 

 the exchange of gases necessary to the life of the embryo is accomi)lished. 

 Whether any actual conveyance of nutriment from the maternal organ- 

 ism during this intrafollicular development of the embryo takes place 

 is extremely doubtful, in that we find the yelk-sack with its vessels 

 developed just as in many species which develope oviparously. Res- 

 piration is undoubtedly effected, however, by this quasi-placental appa- 

 ratus of Gambusia, for we find the vascular apparatus of the embryos 

 very highly developed long before their escape from the follicle; in 

 fact, the branchial leaflets are already so far developed as to be pinnate 

 in structure, with vascular loops formed in the pinnte, a condition of 

 affairs not usually attained by the embryos of oviparous forms until 



