ON THE EMBRYOLOGY OF THE TELEOSTS. 197 



After this stage the segmentation is so irregular that it cannot be timed. 



From the above table, which is based on eggs of apparently the same species, 1 it will be 

 seen that there a considerable variation in the times which were required for the same 

 changes in different eggs but nevertheless in many, well marked periods of rest alternat- 

 ing with stages of activity, may be noticed. These periods of rest and activity have 

 recently been commented upon by Dr. W. K. Brooks ('81) and have also been noticed by 

 many of the older embryologists in the eggs of other vertebrates and also in those of 

 many invertebrata. I am of the opinion that these periods of (apparent) rest are thus to 

 be explained, that at each one of them the deutoplasm, which I believe to have been taken 

 up by the germinal area, is connected with protoplasm, and that while there is an interval 

 of physical rest, the same time is one of chemical activity. There are several reasons for 

 this belief, but before stating them I wish to obtain further evidence and make additional 

 observations not only on the eggs of fishes but also on those of other animals. 



The phenomena of segmentation in the eggs of Teleosts have been several times 

 described, and the accounts which we have presented to us agree in the main with what 

 has been given above, though there are several points of more or less importance in which 

 differences arc to be noted. The first fact which we would discuss is that the planes of 

 segmentation even at first pass through the germinal area, cutting it completely. This is 

 in strone; contrast with the observations of most writers and so far as we are aware occurs 

 in the eggs of all marine teleosts. Every form which we studied presented this peculiarity 

 and the description and figures of Ilaeckel * and Van Beneden of the development of 

 European marine teleosts. (It might here be remarked that the figures of Haeckel are 

 highly idealistic and show many features which certainly do not exist in nature). On the 

 other hand all writers describing the segmentation of the eggs of fresh water fishes agree 

 in that the first cleavage planes pass but partly through the germinative disc, there 

 remaining a portion next the deutoplasm (vitelline globe) which does not segment until 

 much later. These facts are in strict accordance with the ideas of Balfour that eggs 

 undergo total or partial segmentation according to the relative proportions of protoplasm 

 and deutoplasm. In the eggs of fresh-water fishes besides the vitelline globe there is a 

 large amount of deutoplasmic material scattered through the germinal area ; in the eggs of 

 Merlucius there is a very slight amount in the same region, while in the eggs which we 

 studied the protoplasm and deutoplasm appeared to be entirely distinct. 



According to Vogt ('42 p. 30) the segmentation furrows do not entirely cut through 

 the germinal area of the eggs of Coregonus until a stage with eight blastomeres is reached. 

 To Oellacher we must refer for the most detailed account of the segmentation of the eggs 

 of fresh water fishes which has yet been published. His observations on the segmenta- 

 tion of the eggs of Trutta fario (p. 395 et seq. pi. xxxiii figs. 18-20) agree essentially 

 with those of Vogt but from their later date are much more valuable. He lays especial 

 stress upon the fact that the cleavage furrows do not pass at first completely through the 

 germinal portion, and in the later figures (1. c. figs. 22-26) he shows a layer of unsegmented 

 protoplasm underlying the central cells of the blastoderm and continuous with the margin- 



1 These eggs as far as the microscope would show were This would explain the great difference in time in certain 

 identical, but there is a bare possibility that Xo. 4 which was changes which here are very much accellerated. 

 slightly larger than the rest belonged to a different species. 



