212 Mr. J. T. Cunningham on some 



mental cavity ' to Prof. Ed. E. Prince's less ambiguous term 

 'germinal cavitj.' " The term I used was segmentafion- 

 cavity, not segmental cavity ; and how anyone can maintain 

 that that term, whose meaning has been firmly established 

 by Balfour and every other leading embryologist, is more 

 ambiguous than Prince's use of the term germinal cavity, 

 utterly passes my comprehension. 



But we cannot yet leave the account given by M'Intosh 

 and Prince of the segmentation-cavity. Tliey speak of 

 another cavity observed in some Teleostean ova as repre- 

 senting the true blastocoel ; but it is generally admitted by 

 recent observers that there is but one cavity — that which 

 M'Intosh and Prince call the germinal cavity, and it is this 

 alone which I am discussing. Those authors proceed to 

 argue that this cavity, having been, as they think, proved not 

 to be the blastoccel, is really the enterocoel or cavity of 

 invagination. It would take too much space to summarize 

 their arguments. It will be sufficient to mention one or two 

 facts which entirely disprove their conclusion. They say 

 that the cavity is roofed over by endoderm- and epiblast-cells. 

 It is a simple fact, which admits of no dispute, that the 

 portion of the blastoderm which forms the roof of the cavity 

 does not consist of endoderm at all, but wholly and exclusively 

 of epiblast. The hypoblast or endoderm is represented by the 

 lowest layer of the germinal ring and by certain cells derived 

 at a later stage from the periblast ; the germinal ring all goes 

 to form the dorsal rudiment of the embryo. No part of the 

 outer covering of M'Intosh and Prince's " germinal cavity " 

 ever has anything whatever to do with the formation of the 

 intestine, and therefore has nothing to do with the hypoblast. 

 Now an enterocoel must be entirely surrounded by hypoblast j 

 what, then, becomes of the extraordinary proposition of the 

 St. Andrews embryologists ? 



The Periblast. 



1 have previously referred to the account given by Agassiz 

 and Whitman (6) of the origin of the nucleated periblast. I 

 fully accept their conclusions as to the origin of the first 

 nuclei in that layer ; but I consider that their figures indicate 

 a different subsequent history of the layer from that which 

 they describe in their text. I believe, as I have said before, 

 that as the nuclei of the marginal cells from the sixteen-cell 

 stage onwards continually divide, cell-division also takes place 

 in these cells, but at a slower rate than the nuclear division. 

 In consequence of this new cells are continually being sepa- 



