56 EMBEYOLOGY OF THE LOWEE VEETEBEATES ch. 



mesoderm segment in its earliest stage, in the typical Vertebrate, 

 with that of Amphioxus, the way is found to be blocked by a second- 

 ary adhesion (or absence of separation !) having come about between 

 the mesoderm segment and the endoderm from which it has arisen. 



The young mesoderm pouch of Amphioxus is attached to the 

 endoderm at its base — i.e. its ventral end. Its homologue in the 

 more typical Vertebrate, on the other hand, is continuous with the 

 endoderm in two different regions, one ventral and one dorsal. This 

 is illustrated by such a diagrammatic section as that shown in 

 Fig. 33, B, in which the solid mass of mesoderm on each side, indi- 

 cated by the medium tone, is continuous with the mass of endoderm 

 or yolk-cells at the points a and b. The question is, which of these 

 two points is to be interpreted as representing the root of the meso- 

 derm pocket in Amphioxus ? Clearly only one of them can represent 



Fig. 33. — Diagram illustrating (B) the origin of mesoderm from endoderm in an Amphibian, 

 and (A and C) the two methods of correlating it with the mode of mesoderm formation 

 in Amphioxus. 



a, b, sue text ; ect, ectoderm ; ent, enteric cavity ; A r , notochordal rudiment. 



this and the other region of continuity must represent a secondary 

 fusion of mesoderm with endoderm. 



The majority of embryologists, following 0. Hertwig (1882), 

 believe that the dorsally situated region of continuity marked b is 

 the primary connexion as is illustrated by Fig. 33, C. On this view 

 the mesoderm segment of the Vertebrate springs from the endoderm 

 at a point about the level of the notochord, it grows downwards on 

 each side of the alimentary canal and eventually its tip meets the 

 tip of its fellow of the other side of the body in the mid-ventral line. 

 In this view, again, the continuity which can sometimes be shown 

 to exist between mesoderm and endoderm at the point a would be 

 regarded as secondary and without evolutionary significance. 



If however due weight be accorded to what is observed in the 

 development of the lower holoblastic vertebrates it seems more 

 reasonable to the present writer to conclude that the more ventrally 

 situated connexion, that marked a, is the primitive one and that the 

 more dorsallv situated, b, is the secondary acquirement (Fig. 33, A). 



