56 K. MITSUKURI; 



median Jine of the floor of the posterior umniotic tube. 

 They are suspiciously close at two points, so much so at 

 one of them (marked by an arrow) that the layers cnn 

 not be clearly distinguished. 

 n is the third section behind m. The yolk-plug is distinct 

 and unmistakable and projects out above the epiblast on 

 the floor of tlie amniotic tube. Here is a connection 

 between the yolk-plug and the mesoblast in the median 

 line. The yolk-plug l^egins in the section next behind 

 7)?, and is found in five sections including that one. 

 n is therefore the third or the middle one of the five. 

 For some sections after //, the median line of the epiblast 

 is arched upwards, marking the posterior prolongation 

 of the yolk-plug on tlie epiblast. Moreover, tliere seems 

 to be a fusion, in the median line, of the epiblast with the 

 two layers of the mesoblast, and thus the coelem does not 

 go across from side to side, being intei-rupted by the 

 median partition. 



Although the primitive streak can not be clearly made out in the 

 reo-ion of / and ?;/, I think, we may consider this as due to the 

 fiict that the whole structure is move or less rudimentär}'. The 

 primitive streak and the ]:)rimitive groove in this specimen may there- 

 fore be considered to begin at the 



Woodcut VII. ^''^*^ 1^'"'^ '^^^ *^'^ medullary canal 



0), then to run backwards along the 

 Med%W \ dorsal surface of the tail-swelling 



o»ns\.M -p^ii.s^eUuig. j| (f/-/^'), around and down its posterior 



tip to the ventral surface (Ji-Ï)^ and 

 then backwards along the median line 

 of the epiblast below the tail-swelling 

 (Ic-m), until it reaches the yolk-plug 

 at its posterior end (n). Diagram- 



