EARLY DEVELOPMENT OF THE COMMON NEWT. 49 



growth of the invagination plates by a process of cell division, 

 for the mesoblast does not arise at separate points, capping 

 the hypoblast, but in direct continuity with the invagination 

 mesoblast. 



In the Elasmobranch Fishes, in which the origin of the 

 mesoblast has been carefully observed, there is no doubt 

 that this layer arises as two lateral masses, splitting off 

 from the hypoblast at the same time that the latter arises as 

 a distinct stratum from the lower layer cells. Here, however, 

 the lateral plates do not form a continuous layer with the 

 mesoblast which occasionally arises at the reflection of the 

 epiblast at the sides, but are distinct from it. 



Calberla,^ as previously stated, explains the growth of the 

 mesoderm (mesoblast) in the Lamprey, as an early splitting 

 of the outer portion of the primary endoterm. This view 

 fully confirms our interpretation of the lateral growth in 

 Triton. 



In Kowalevsky's earlier researches upon Amphioxus he 

 fell into the error of supposing the mesoblast of double 

 origin, hypoblastic and epiblastic, an error which he cor- 

 rected later^ by attributing this layer to a constriction off 

 from the hypoblast, which occurs subsequent to the forma- 

 tion of the notochord. The simple invagination does not 

 give rise to any but the two primitive layers. There can 

 now be no doubt thai the formation of the mesoblast is in 

 all types a secondary phenomenon which is retarded in the 

 simpler forms, and hastened in the more complex into an 

 earlier period of development. 



To review the features noticed in Stage a. The meso- 

 blast arises by invagination as two lateral plates, and is 

 never found across the median line. Subsequent growth is 

 partly by cell division of these plates ; mostly, however, at 

 the expense of the hypoblast. The most rapid development 

 is posteriorly, both in respect to thickness and downward 

 growth. There is no tendency to split into somatic and 

 splanchnic layers. By Stage b the mesoblast shows a very 

 marked progress. It is now thickest immediately below the 

 medullary plates, and causes that upward curve in the out- 

 line of the epiblast previously mentioned (Plate IV, fig. 5). 

 At the same time the lateral plates have approached each 

 other, bending the hypoblast downwards, so that now it is 

 continuous with the epiblast only in the median line. Tlie 

 section figured is in the anterior part of the embryo near the 

 head region. The cells appear larger than in the la^t stage, 



» E. Calborla, loc. cit. 



' Vide A. Kuwalevbkv, 'Archis'. fur Micros. Auatomie.' Band 13, p. I'Jl. 



4 



