MUSEUM OF COMPARATIVE ZOOLOGY. 287 



liminary notice, however, does not describe the process of convolution in 

 detail. An interesting feature is the statement that blind diverticula do 

 not appear until the tubes of the gland have become very much convo- 

 luted. In the pronephridia which I have studied, I have never seen a 

 blind diverticulum. My observations do not extend to sufficiently old 

 stages to allow me to deny that such diverticula appear anywhere in the 

 developmental history of the gland, but the organ can reach at least the 

 high degree of complexity shown in Figures 41 and 65, and yet be com- 

 posed of the windings of the nephrostomal canals, the collecting trunk, 

 and the common trunk without possessing any blind diverticula. 



It is needless for me to discuss in this place the histology of the tubu- 

 lar portion of the pronephros. These details have little general interest, 

 and they have furthermore been accurately given by Fiirbringer and 

 Hoffmann ('86). 



The dilated chamber which I have described (page 240) was also ob- 

 served by Hoffmann, but he was unable to determine what portion of the 

 system was concerned in its formation. Similar dilated chambers are 

 likewise described by Marshall and Bles, who regard them as steps in 

 the degeneration of the tubules. The early appearance of these dilated 

 regions in Rana (see page 232) seems to me to render this interpretation 

 improbable. 



According to the usual account, the capsule arises as a differentiation 

 of the connective-tissue stroma, which lies between the pronephros and 

 the ectoderm. Duval ('82, pp. 25, 27) alone has claimed an origin from 

 the overlying protovertebrse ; but, singularly, his statement has been 

 wholly neglected by subsequent writers. His observations on this point 

 agree in all essential features with my own. 



The glomus was discovered by Job. Muller ('30, p. 12), but the sig- 

 nificance of the structure was wholly problematical until Bidder ('46, p. 

 58) suggested its glomerular nature, which has since received general 

 acceptance. This view has, however, been opposed by Semper ('75, pp. 

 439 et seq.), and more recently by Hoffmann ('86, pp. 572, 573). Accord- 

 ing to Goette and Fiirbringer, the glomus arises as an outfolding of the 

 splanchnopleure opposite the pronephric nephrostomes. The interior of 

 the fold becomes occupied by mesenchymatic cells and with blood tracts, 

 which communicate with the aorta. According to Hoffmann, the inte- 

 rior is largely occupied by " columns " of large cells, which would seem 

 foreign to the nature of a glomerular structure. These ' columns of cells," 

 he says, may be seen to arise, in Bufo at least, by the invagination of the 

 superficial covering of the glomus. I have myself seen continuous cylin- 



