14S PROCEEDINGS 0^ UNITED STATES NATIONAL MUSEUM. [1685. 



membraues and fluids, oxygen being constantly supplied and carbonic 

 dioxide carried oft" by means of a specialized portion of the blood-sys- 

 tem of the maternal organism- 

 There is still another character which distinguishes the development 

 of Oamhusia from that of the mammal. The body of the former is built 

 up by a gradual transformation or conversion of the substance of the 

 yelk into the various structures which make up its organization. In 

 other words, the young Gambusia obtains no nutrition from its parent ; 

 there is merely an incorporation bj' the embryo of the stored proto- 

 plasm of the yelk- sac. In the Eutherian mammal, on the other hand, 

 haijxl, the embryo receives nourishment through the placental apparatus; 

 by far the greatest proportion of the embryo being built up from the 

 protoplasm supplied and conveyed to it from the blood-system of the 

 parent. Judging from the large size of the young of some viviparous 

 fishes, such as those of Anableps and the Emhiotocidw, it is obvious that 

 there are exceptions to the method of development characteristic of 

 G(WiI)usia. 



Besides the very intricate net-workof capillary vessels which covers the 

 follicles of the ovary of Gambusia, a large opening. Figs. 9, 10, 11, and 12 

 mp, Plate VIII, and Fig. 13, Plate IX, of a circular, though usually 

 more or less oval, form makes its appearance in the wall of each one at 

 or near the point of attachment of the vascular stalk by which they are 

 supported. This opening appears to increase in size a? the young fish 

 develops ; whether it is present during the earliest stages of the intra- 

 follicular development of the embryo I do not know, as I did not have 

 an opportunity to see those phases. A branch from the main nutritive 

 vessel frequently lies near the margin of the opening, curving around 

 it. Whether this opening serves the same purpose as the micropyle of 

 ova provided with a membrane would appear very probable, as it is 

 difficult to see in what other manner the milt, which is i>robably intro- 

 duced into the abdominal cavity by the male, could reach the ovum 

 through the wall of its follicle. The ovary itself seems to have no ex- 

 terior investment, so that the follicles lie directly within the abdominal 

 cavity, the young fishes uj^on the completion of their development rup- 

 ture them and escape into the latter, and from thence through an 

 abdominal pore into the outer world. The opening into the follicle 

 licle may be named the foUicidar pore. Through it the cavity in which 

 the embryo lies is brought in direct communication with the cavity of 

 the abdomen, because it is extremely doubtful if the ovary as a whole 

 is invested by a distinct membrane. If it is, its tenuity is extreme, 

 since the follicles when gravid are readily separable as if entirely un- 

 covered, or as if there were no external peritoneal investment reflected 

 over them. 



The nearl}^ globular vitellus of the mature Q^g measures about a line 

 in diameter. The germinal protoplasm ])robably occupies a ]ieripheral 

 position covering the nutritive or vitelline portion of the Qgg as a con- 

 tinuous envelope with strands of germinal matter running from it down 



