Q2 i^'- MITSÜKÜRE 



m 18 tliree sections behind /, and is very inucli like that 

 section : only tlie floor of the posterior aniinotic tube is 

 rather irregularly warped and the coelom has extended 

 across from side to side, dividing the mesoblast into the 

 parietal and the visceral layers. 



n is three sections behind in. 



is six sections behind n. From the third section in front 

 of (inclusive), a median knob begins to show itself on 

 the floor of the posterior amniotic tube and assumes in 

 one or two sections the shape represented in this one. 

 There is a slis^ht connection between it and the mesoblast. 

 I believe we ought to regard this as the yolk-plug. The 

 remarkable fact is that this elevation continues for scventij 

 sections or more; in fact it extends a little beyond the ter- 

 mination of the posterior amniotic tube and is visible in 

 the surface-view as a white streak between the limbs of 

 the horse-shoe shaped opening of that tulie (Fig. 15). 

 The elevation is, in the middle ])()rtion of the tube, even 

 higher than in o, occupying almost the whole of the some- 

 what small lumen of the tube, but becomes broad and flat, 

 as the tube flattens near the posterior end. The connec- 

 tion with the mesoblast seems to be given up rather early, 

 although I can not make out exactly where it takes 

 place. 



A comparative study of this and Series X throws light on 

 the process of the formation of the tail-swelling. It is often stated in 

 Avorks on Embryology that the tail is formed by the tucking in of the 

 blastoderm under what is going to be the tail, i.e., by a process of 

 folding. Such is, however, not the case in Chelonia. In Series X, 

 the tail-swelling is hardly marked, while it is fully formed in this 

 series (XI). A comparison of this series e-j with Series X, e~i 

 shows that a large mass which is not present in the latter series has 

 been added to the dorsal surface of the sections in the former series, 



