MARY |. ROSS. 



ment of nuclei at different levels also corresponds to the insinking 

 at the surface, shown in all of the series of Amblystoma studied 

 (Figs. 1 8, 19, 20, d\ As the gland fundaments increase in size 

 by cell division the yolk spherules become absorbed ; the epi- 

 thelium lining the stomach becomes narrower and lower, until in 

 a larva of 12 mm. (Fig. 20) a stage comparable to Fig. 3 of 

 Desmogiiatlius is obtained. The yolk disappears from the glan- 

 dular epithelium more rapidly than from the surface epithelium. 

 Thus the origin of the fundus of the gland is more readily traced 

 than the development of the excretory duct. Figs. 21 and 22 in 



.- ; '-, 7 



;^f*..\ .rtfjji'-i I'-./.''--'*' '?*?&"., 



Jl 



/ P?; -#?f f 



>J8.>ir/ 



, _..;Ajjir . ..'.. .... 



..'; -'i' ' : ::'^^-;.-- ; - - .3^ &v-' 



FlG. 19. Same as Fig. IS. The primary round cells now divided to form two. 

 Obj. 7, Oc. I. 



a larva 16 mm. in length show slight insinkings or indentations 

 at the surface. In the same figures the gland fundaments have 

 progressed further in development than in the corresponding stage 

 in Dcsinognathns. The fundus of the gland extends through- 

 out several sections and the cells have already become arranged 

 around a central lumen (Figs. 21 and 22, /). The cells which 

 are to form the excretory duct and neck, although arranged 

 around the gland as a central axis, do not present a marked 

 opening. Instead of a narrow oval slit, as in a corresponding 

 stage of Dcsmognatlius, there is formed a shallow pit. From this 



