ON THE EMBRYOLOGY OF THE TELEOSTS. 195 



Returning to the egg whose segmentation we are discussing we have to note the suc- 

 ceeding features of division and first the changes in going from four to eight cells. In six 

 minutes after the appearance of the nuclei in the four blastomeres they again disappear as 

 before, and fourteen minutes elapsed before any further changes were visible. A 

 depression then appeared at the middle of the inner margin of each cell and gradually 

 extended outward and downward to the outer surface and to the yolk. These furrows 

 were parallel to the first and at right angles to the second segmentation furrows and like 

 them cut completely through the protoplasmic portion of the egg. The whole process of 

 division occupied but a few seconds and the nuclei reappeared two minutes later and the 

 blastoderm of eight cells was before us. 



In the egg which formed the basis of this description the segmentation furrows appeared 

 simultaneously in each of the four cells of the blastoderm and proceeded at nearly regular 

 rates in all. Other eggs agreed with this, but still others (and they formed a large minor- 

 ity) exhibited at this early period a heterochronous division, as in proceeding from four to 

 eight blastomeres an intermediate stage was observed in which for a few moments the 

 blastoderm consisted of six segments, two of the cells having divided slightly in advance 

 of their fellows. In one egg which I studied three blastomeres divided some little time 

 before the fourth, the first three dividing simultaneously and the result was a blastoderm of 

 seven cells. This egg presented also another peculiarity in that the nuclei reappeared at 

 nearly the same moment of time in all of the cells and before the fourth had divided, 

 and thus the seventh cell had two nuclei and only after the lapse of a minute and a half 

 did this cell divide. 



It was at this time, when the egg has eight segments, that we noticed the first traces 

 of Van Bambeke's intermediary layer (figs. 21, 22, 23i). It was a very thin layer of pro- 

 toplasm extending between the blastoderm and the yolk and at this stage was without the 

 thickened margins and the free nuclei which are present in the later stages. No traces of 

 granulations or oil globules could be seen in it. We would not, however, by the forego- 

 ing account, be understood to say that the intermediary layer made its appearance at this 

 time, but merely that we first noticed it then. Of its time and method of origin we can 

 say nothing except that it certainly was not present at the first segmentation of our eggs. 1 



As will be seen further on we do not agree with Van Beneden in regarding this inter- 

 mediary layer as the hypoblast, but I am inclined to believe that he is correct in his idea 

 that the tail-like processes of the cells in Haeckel's figure of an egg with two cells, in 

 reality repi'esent this intermediary layer. I also agree with the Belgian savant in his 

 opinion that both Kupft'er and Lereboullet observed this same intermediary layer, the 

 statement of Kupffer to the contrary notwithstanding. Lereboullet ( '54 p. 250) says 

 " II existe sous le blastoderm une membrane particuliere, distincte, composee de grandes 

 cellules tres pales ; c'est d'elle que se formeront les organes abdominaux ". This agrees per- 

 fectly (except in regard to the last portion, the destination of the layer) with the condi- 

 tion of the layer in many of the fresh water fishes with numerous oil globules the " gran- 

 des cellules " being either oil globules or vacuoles. But of this intermediary layer we 

 will speak more at length further on. 



X I now think that this statement will have to be modified, from the blastomeres, as the first traces of the intermediary 



as I regard the thin portions of protoplasm which are left layer, and the same investigations confirmed me in my beliet 



extending down over the egg at the time of maturation that the layer is largely produced by the change of deuto- 



(fig. 11), and which are also shown in fig. 12 extending out plasm into protoplasm. 



