Disputed Points in Teleostean Emhryologij. 211 



most reputed embiyologists have always maintained, the same 

 thing as the segmentation-cavitj in other ova ; for they do 

 not seem to admit the fact that all ova are homologous, and 

 that the various modes of development, leaving aside those 

 exhibited bj Coelenterata and Crustacea, are modifications of 

 one fundamental plan. But I would point out that the term 

 germinal cavity is synonymous with segmentation-cavity, and 

 that if they wish to maintain that the cavity in Teleostean 

 ova is something else, they ought to give it some other 

 name. These authors admit that Balfour's segmentation- 

 cavity in Elasinobranchs is homologous with the se^-men- 

 tation-cavity of Amphibians, although, as they also admit and 

 as Balfour states, the floor of the cavity in Elasmobranchs is 

 at one stage formed by the yolk with its external protoplasmic 

 layer, as in Teleostei. The basis of the surprising conclusion 

 of ]\l'Intosh and Prince is obvious enough. Tliey speak of 

 the Teleostean germ after segmentation as a morula which 

 flattens out and becomes lifted up and separated by a chamber 

 from the appended tropliic mass. That is to say, they regard 

 the yolk with its envelope of nucleated . protoplasm as some- 

 thing distinct from the germ, and the germ, or, to use the 

 proper term, the blastoderm, as alone homologous with the 

 morula of an &gg with simple equal segmentation, such as 

 that of Amphioxus. But, as is satisfactorily shown by 

 Balfour's Comp. Embryology, chap. xi. vol. ii., and by my 

 paper " On the Relations of the Yolk to the Gastrula '"' (2), 

 not to mention numerous other papers by different embryo- 

 logists, the Teleostean e.gg must be compared whole for whole 

 with any other Qg^. The yolk is not something added on to 

 the outside of the eg^, but is an accumulation of food-materinl 

 within the tgg itself. As development proceeds certain cells 

 are separated, while others, namely the nucleated periblast, 

 contain the whole yolk ; and it is as certain as any other ascer- 

 tained relation in embryology that the periblast and the yolk 

 are homologous with the yolk-cells in the Amphibian ovum, 

 which there form the floor of the segmentation- cavity. That 

 this is so is conclusively proved by the fact, demonstrated in 

 my paper on Ku])fler's vesicle, that the periblast gives rise to 

 cells which form the floor of the intestine, as do the 3'olk-cells 

 in Amphibians. My account of Kupfter's vesicle, excepting 

 that part which refers to the formation of the floor of the gut 

 from the periblast, has been entirely confirmed by a paper 

 published last year (15) by Henry V. Wilson in America. 



In the light of the above considerations it is somewhat 

 extraordinary that Prof. ]\PIntosh, in his review of my work 

 on the sole (1 1), .'^hould write, " lie prefers the term * seg- 



