DEVELOPMENT OF TKACHINUS VIPEEA. 279 



invaginated hypoblast arises in tlie Trout by an involution of tbe 

 nervous layer of the epiblast, and that the epidermic layer ends 

 on the surface of th.e yolk, taking no part in the process. On 

 the other hand, Messrs. Kingsley and Conn maintain that in 

 th.e Gunner the " epidermic layer " only of the epiblast takes part 

 in tlie invagination. According to Balfour, in smaller teleostean 

 eggs th.e nucleated cells of the intermediary layer form the hypo- 

 blast. Without the aid of sections it is impossible to tell posi- 

 tively wliat really does take place ; and the difficulty of preserving 

 and hardening pelagic eggs in this stage is vrell known. I am, 

 however, inclined to think that Henneguy is right, and that the 

 invagination observed in optic section in the living egg is an 

 inward folding of the lower layer cells of the epiblast. The cells 

 thus formed are, to begin with., in a single layer only, the outline 

 of which, top and bottom, is very well marked, but the cells 

 themselves are not columnar as described by Kingsley. 



Intermediary ifflyer = Parablast of Klein. — My observations 

 confirm Yan Bambeke's and Klein's figures and descriptions as 

 regards the presence of a thickened peripheral layer or welt, the 

 " nuclear zone " of Kupffer. This thins ofi" under the blastodisk, 

 and also gradually becomes indistinguishable as it passes away 

 from the blastodisk round the yolk. Very early in the segmen- 

 tation of the disk, certainly as early as the eight-cell stage, and 

 probably earlier, there is visible around the disk a granular zone, 

 the largest granules being nearest to the disk, and becoming 

 finer and finer further away. This is a prominent zone in the 

 sixteen-cell. stage (fig. 7), and it is in this zone that the free cell- 

 formation already described takes place. A cellular structure 

 has certainly been traced under the rim of the blastoderm, but 

 not far, probably only so far as the thickened portion of the 

 blastodermic rim extends. There is, however, nothing in the 

 intermediary layer under the disk corresponding to the median 

 lens ilentille) described by Yan Beneden. 



'Retardation of this Stage. — The development described up to 

 the present was observed in April at a mean temperature of 

 49° E. ; whereas the later stages were observed in July at a 

 mean temperature of 60° F. Thus the rate of development in 

 this section is altogether out of proportion to the remainder ; but 

 I have had to avail myself of these earlier observations, because 

 in the hot summer weather the eggs laid during the night were 

 well advanced in the segmentation stage by 7 a.m. ; and thus I 



