FATE OF BLASTOPORE, ETC., IX CÏÏELONIA. 35 



for this point. 



In the next section, the epiblast is separate from the sub- 

 jacent mesobiast in the middle of the yolk-plug. In 

 three or four sections more, all the three layers are in- 

 dependent of one another. 



Woodcut IV., F (p. 12) may be taken as the diagrammatic 

 lonofitudinal section of this stasfe. 



This series shows that in the embryo given in Figs. 1 & la, 

 there is in tbe posterior part a blastoporic passage which tjikes an 

 oblique forward course from its dorsal to its ventral surface. The dor- 

 sal opening of this passage is horse-shoe shaped (Fig. 1) and the outer 

 edges of this opening are proliferating the peristomal mesoblast in all 

 its parts, excepting the anterior lip where the epiblast is reflected and 

 continuous with the chorda-hypoblast. The mass of irregularly 

 arranged cells, which forms the floor of the blastoporic passage and is 

 also continuous with the peristomal mesol^last, sticks out to the level 

 of the dorsal surface between the horse-shoe shaped lips of the 

 blastopore, and is bounded anteriorly and laterally by them. This is 

 the mass which I have called the ijolk-plwj. Posteriorly I have not 

 seen this mass bounded by a groove in any specimen. Ijut as the 

 yolk-plug passes posteriorly gradually into the epiblast, and as the 

 region of this transition is continuous with the peristomal mesoblast, 

 (Woodcut IV. F)^ we may suppose the blastopore lip to be present 

 here in a rudimentary form. I wish also to call attention to the fact 

 that all the three layers are fused on the floor or posterior wall of the 

 blastoporic passage (Cf. Woodcut IV. F). 



As Section/ of this series shows, there is no epiblast on the dorsal 

 surfiice of the yolk-plug. I have reconstructed from the series a sur- 

 face view of the posterior portion of the embryonic shield for the 

 purpose of ascertaining the extent of the area thus free from the 



