40 W. B. SCOTT AND HEN HY F, OSBORN, 



Our sections do not wholly accord with the observations 

 of either of the above, for if it is clear that the invagination 

 gives rise in the median line to a single layer of cells, it is 

 equally clear that at the sides it gives rise to a double layer, 

 namely, of mesoblast as well as hypoblast. 



The process in Triton agrees then more closely with that 

 occurring in the Elasmobranch Fishes,^ where the lower layer 

 cells, confluent with the reflected epiblast on either side of 

 the axial line, form a layer of spherical cells above and co- 

 lumnar cells below, and the former is ultimately separated off 

 as the mesoblast proper, while in the axial line the lower 

 layer cells give rise simply to a columnar layer. 



Now, turning to the transverse section of a Triton embryo 

 Stage A (see Plate IV, fig. 4) we find that it adds still 

 further probability to this view, for the relations of the 

 layers fully accord with the above interpretation of the 

 invagination. 



Now, as concerns the further growth of the mesoblast, it 

 results from the foregoing conclusions concerning the hypo- 

 blast that the mesoblast is never present across the axial 

 line in the early stages. In transverse sections of Stage a 

 it appears as two lateral plates extending on either side to 

 a point just above the side limits of the alimentary canal. 

 The layer where it is nearest the alimentary canal consists of 

 small round cells, one or two deep, which can be readily dis- 

 tinguished from the adjacent hypoblast. These are the 

 cells which we have just referred to as having resulted from 

 invagination, and we shall speak of them hereafter as the 

 primary mesoblast cells. 



In conclusion, all the observations we have made favour 

 the above interpretation, while none in any way disprove it. 



Thus, at once three important distinctions are established 

 between the development of the layers at the point of invagina- 

 tion in Triton and Bombinator, if we accept in full Dr. Gotte's 

 investigations of the latter. First ^ in Triton there is a 

 decided thickening of the single layered epiblast as it 

 approaches the point of invagination. In Bombinator there 

 is non**. Second : the resulting hypoblast in the axial line 

 is in direct contact with the epiblast. There is no inter- 

 vening mesoblast as in Bombinator. Third : the mesoblast 

 is found in Triton as two lateral plates, and is not con- 

 tinuous across the middle. 



These observations, coupled with those of Calberla,we think 

 leave little doubt that Gotte has mistaken the upper hypo- 

 blast cells for mesoblast, and thus at the start fallen into- an 

 1 Vide F. M. Balfour, ' Elasmobranch Fishes,' p. 49. 



