450 W. B. SCOTT AND HENRY F. OSBORN. 



with the greatest care ; and as the albumen is permeated by 

 several membranes, it was found necessary to cut these 

 with fine scissors before the embryo could be with safety 

 extracted. Many hardening reagents Avere experimented 

 with — osmic acid, bichromate of potash, Miiller's fluid, &c., 

 but the most satisfactory one proved to be Kleinenberg's 

 picric acid, with which nearly all the embryos described in 

 the following pages were prepared. In those cases where the 

 entire egg was hardened without previously removing the 

 albumen, the results were most unsatisfactory. Kleinen- 

 berg's haematoxylin was the staining fluid employed for the 

 sections. 



A, 

 This includes embryos intermediate in age between 

 Gotte^s figs. 39 and 40, taf. iii. The blastospore is quite 

 small, a narrow groove, the " Riickenrinne,^' running forward 

 some distance from its anterior edge. The medullary folds 

 do not as yet appear in surface views. The ovum is still 

 almost perfectly spherical in shape. 



B (Unke, Taf. iii, figs. 40 and 41). 



At this stage the medullary folds become well developed 

 and very plainly marked. As yet they are widely separated. 

 The medullary plate is formed, but the groove which divides 

 it into two parts does not reach far forwards of the middle ; 

 or, any rate, if present anteriorly, is extremely faint. The 

 ovum has elongated very slightly, but still appreciably. 



C (Taf. iii, fig. 42). 



The medullary folds now become still more pronounced, 

 and begin to approach each other. The point of closest 

 approximation is in the region which will eventually become 

 the neck, and here is the first point of contact, just as it is 

 in the Batrachia. The medullary plate is plainly divided 

 throughout. The elongation of the embryo is not much 

 more marked than it was in the previous stage. 



D (see PI. XXI, fig. 16). 



Up to this stage no important external differences between 

 Triton and Bombinator are apparent, but now a number of 

 points of divergence begin to be noticeable. The medullary 

 folds have closed throughout the region of the trunk, but 

 still remain open in the head. Posteriorly they separate to 

 form a si?ms rhomboidalis ; this does not seem to be merely a 

 part of the canal which has not yet closed, but a genuine 



