138 s. HATTA. 



it is found ; Ijiit in different regions, it presents some modifications. 

 In the ceplialo-cervicnl region, the process has retained best its primi- 

 tive characteristics, while in the region of the trunk, it has become 

 altered to a great degree, owing no doubt to tlie presence of the yolk- 

 cells. On this account, we sliall begin with the description of the 

 mesoblast formation in the first named region, although in point of 

 time it is formed latest in this part, as the formation of the gastral 

 mesoblast proceeds from Ijehind forwards. 



As has been shown by Scott, (/. c, Taf. A^III, fig. 10 b) the head 

 is formed Ijy the Ijlind end of the invaginated tube protruding itself 

 above the oeneral surface of the eo'i>'. The head when first formed is 



O CO 



thus a double- walled tube — the wall of the inva^-inated tube bein«: 

 enclosed by the epiblast. Fig. 19 is a section from near the blind end 

 of the invaginated tube at the time when the head is ready to protrude. 

 The lumen of the invaginated tulje is here more expanded than further 

 behind. Its roof is formed by a regular columnar epithelium in 

 contact with the epiblast. Adapting itself to the slightly thickened 

 medullary ])late, the columnar epithelium is slightly bulged out down- 

 wards in the median line. This median part becomes afterwards the 

 chorda dorsah's while the part immediately adjacent to it on each side 

 forms the gastral mesoblast. J.aterally, in continuation with the 

 epithelium of the invaginated tulje, the outermost row of the yolk-cells 

 is arranging themselves into a regular epithelium. This, as the sub- 

 sequent lnst(jry shows, Ijecomes the wall of the definitive gut, all the 

 yolk-cells enclosed within this layer being eventually entirely absorbed 

 (cf. tig. 27). This, in my opinion, corresponds with the la^'er marked 

 z in fio-. lib. or that marked Dm in tiii's. 11a and 14 in Taf. A'lII of 

 »Scott's paper (/oc. cit.), although this writer regards the layer as giving 

 rise to what he calls the " Dottcr-mesodcrrn.'" 



J^'ig. 20 shows a transverse or rather obli(jue section through the 



