524 MR GREGG WILSON ON THE DEVELOPMENT 



the third groove of the head-kidney, then splitting from the segmental tube, and 

 posteriorly growth by cells " passing from the Wolffian duct " to the distinctly separate 

 Mlillerian duct, In later times Janosik (pigeon), Mihalkovics (duck, fowl, etc.), and 

 Nagel (man) have supported an independent growth backwards ; Wiedersheim 

 (reptilia) has described a growth that recals Waldeyer's observations ; and Amann, 

 from a study of the sheep, has concluded that increase in length is chiefly due to a 

 proliferation of the cells of the duct, but that both the Wolffian duct and the epithelium 

 may contribute. 



In Axolotl "/3" the short rod already described ends in a simple epithelial thicken- 

 ing, which finally dwindles till it is a one-cell-deep band (fig. 19). 



i In Axolotl "y" the course of the rod is frequently interrupted as it passes back. It 

 becomes more or less distinct at intervals ; and its posterior is lost in the epithelium. 



Axolotl " S" shows a long open tube passing backwards, but towards the posterior 

 this diminishes greatly in size, and gives place to the usual rod, which again at intervals 

 becomes more or less indistinct, and finally merges in the epithelial thickening that 

 invariably lies outside the Miillerian duct Anlage (fig. 20). The duct of Axolotl " e " 

 likewise dwindles towards the posterior ; the small tube is followed by a rod, and it is 

 succeeded by cells, which are at first dense, and further back are' much less compact 

 (figs. 21 and 22). 



Salamandra atra resembles Axolotl in this respect. In " G " towards the end of 

 the duct there is noticeable variation in the size of the Anlage ; and at intervals the 

 short posterior rod loses its distinctness. At one or tivo places there is a distinct 

 connection betiveen the rod and the epithelium, so it is not only at the extreme posterior 

 end that the epithelium contributes to the formation of the duct (fig. 23). 



In none of the specimens examined by me have I seen any communication between 

 the segmental duct and Anlage of the Miillerian duct. On the other hand, I have 

 always found the duct-Anlage ending in the epithelium. Independent growth back- 

 wards seems to be excluded by this constant relation of the tip : the appearance of 

 independence is given by the fact that almost immediately anterior to the tip the 

 Anlage is found to have become free from the epithelium. The irregularities in the 

 development are also opposed to the view that growth is independent from before back- 

 wards : in such a case one would expect a continuous development. From the relations 

 in " G" it would appear that the posterior end is lost in the epithelium, — that is to say, 

 that the thickened epithelium is the first stage in the development of the posterior 

 part ; that almost at the posterior end the duct frees itself, or separates from the 

 surrounding cells : this implies that the formation of the rod follows rapidly 

 on the occurrence of the thickening, and that this differentiation is not equally rapid 

 all along ; this is shown by the occurrence of undifferentiated parts a few sections 

 anterior to the already differentiated posterior end. 



A backward proliferation of the cells of the rod is not improbable ; but I have never seen 

 an independent free termination ; and the epithelial thickening always precedes the rod. 



