666 REPORT OF COMMISSIONER OF FISH AND FISHERIES. [112] 



developed and the beart and vitelline vessels have been formed. There 

 is also present around the yelk of fish embryos an homogeneous en- 

 velope, yh^ Fig. E, containing free nuclei, which is not present iu the frog's 

 ovum, nor has it any exact homologue iu the latter, because it incloses 

 a large mass of deutoplasmic matter often optically and physiologically 

 different in character from the protoplasm of the germ or the envelope 

 itself. 



Van Bambeke and E. Van Benedeu, who first described some of the 

 most important phenomena presented by what I have called the .vei/c- 

 hypohlast, named it the intermediary layer. Iu consequence of the 

 presence of a large amount of deutoplasm in the yelk of the fish ovum, 

 yelk-segmentation has been retarded, in fact has been wholly inter- 

 rui)ted, so that nuclear multiplication alone, unattended by actual seg- 

 meutation, occurs in the yelk envelope yh, shown in Figs. B and G. 

 The distribution in it of the free nuclei //^ is shown in Fig. E. 



In the frog the liver develops as a ventral diverticulum of the fore 

 part of the mesenteron, which grows into and at once appropriates a part 

 of the segmented yelk; the latter in fact becomes fused with and forms 

 the exceedingly thick ventral wall of the mesenteron. In the embryo 

 fish this fusion of the yelk mass with the mesenteron or primitive gut 

 never takes place ; the liver arises as an independent diverticulum or 

 thickening of the ventral wall of the mesenteron. There is no evidence 

 of continuity between the intestine and yelk of the osseous fish at any 

 stage of development. The appropriation of the yelk hy the budding of 

 leucocytes from the yelk hypoblast is effected in some types by direct 

 gemmation into the segmentation cavity, from which the white blood- 

 cells are sucked up by the heart [Alosa, Pomolobns), or where the peri- 

 cardiac septum between the heart chamber and body cavity is so ac- 

 celerated iu development as to extend over the yelk as a veil in which 

 the vitelline vessels are developed {Sabno) and which are concerned in 

 breaking down the yelk. In the fircst case the body and segmentation 

 cavities remain connected ; in the latter they are soon separated by the 

 development of the vascular veil which grows over and around the vi- 

 tellus. In all forms of osseous fishes at an early stage the two seem to 

 be continuous. This is shown in the cross-section of the body of a young 

 fish, younger than E, in Fig. II, where the epiblast e}) and sep and the 

 hypoblast /ty include the muscle and splanchuopleural i^late which ends 

 bluntly at mp, and to the right of which the segmentation cavity extends, 

 so that in the event of the splitting of mj) to form the splanchnopleure 

 and muscle plate projjer there would be a continuity established be- 

 tween the two; none of the lower layers in fact extend far out on either 

 side of the body between the epiblast and yelk hypoblast. That this 

 view is the true one is proven by the mode in which AmpMoxus develops 

 its muscl(3 plates as outgrowths into the blastocoel or segmentation 

 cavity of the blastula after the invagination an<l development of the gas- 

 trula, as seen in the cross-section J of an embryo of Amphioxns, modified 



