40 '■ ■■ ■ K. MITSÜKURI. ' 



the median line in the form of a regular columnar epithelium from cells 

 forming the loose network of the primitive hypoblast. The process 

 proceeds from behind forward. This embryo thus closely agrees with 

 the first stage of Trionyx, which we described in our former 

 paper. 



Formation of the Mesoblast. 



How does the embryo described above change into one in which 

 the three germinal layers are completely and definitely established? 

 As stated in the account of the formation of the mesoblast given in 

 our paper on Trionyx, the process is briefly this : as the forward 

 continuations of the peristomal mesoblast whose commencement we 

 saw in the above stage, a pair of mesoblastic bands become gradually 

 established, one on each side of the median chorda-hypoblast, between 

 it and the gut-hypoblast (= Darm-Enioblast). This is the gastral 

 mesoblast of Rabl. From my study of CI em my s , I am now in a 

 position to give more exact details in regard to the first stages in the 

 formation of this gastral mesoblast, and as will now appear I have to 

 qualify somewhat my views of these earliest stages. This is necessitated 

 by the discovery I have made concerning the chorda-hypoblast. I find 

 that in Chelonia at least, only the median part of what I 

 have hitherto called the chorda-hypoblast gives rise to 

 the chorda dorsalis, and that each side of it becomes 

 eventually incorporated in the mesoblast. Thus, in Fig. 

 11, from an embryo somewhat older than that represented in Fig. 1, 

 the median unshaded portion alone becomes the chorda dorsalis, while 

 the shaded lateral parts form in the sequel part of the mesoblast. 



As this is a very important point, I shall try to prove it before 



