MUSEUM OF COMPARATIVE ZOOLOGY. 257 



them they were closely associated with the yolk entoderm, which lies 

 medio-ventral to the regiuu of the glomus. It is probable that they 

 arise from the eutoderm and migrate into the interior of the splanchno- 

 pleural fold. I have been unable to lind in either liana or Bufo any 

 cells similar to these large cells in the glomus of Amblystoma, and I 

 have at present no suggestion to offer respecting their significance. The 

 glomus, as I have already indicated, is a highly vascular organ, and even 

 in the younger stages it is possible to find vessels which connect it with 

 the aorta. These vessels usually follow the splanchnic layer quite 

 closely, and appear to lie external to the large cells to which reference 

 has been made. 



In the younger larvae of this stage the body cavity in the pronephric 

 region has the form of separate chambers, from each of which a single 

 nephrostomal tubule arises ; but elsewhere the cavity is wanting on ac- 

 count of the contact of the peritoneal surfaces. In the older individuals 

 it is expanded over a much larger area, but by the development of the 

 lung bud a dorsal portion of the cavity is partially separated from the 

 rest as a pronephric chamber. 



Stage VI. 



Plate VII. Figs. 53, 54. Plate VIII. Figs. 61 65. 



The larvae included in this stage were taken from several different 

 killings made in the course of three or four days. They measure about 

 9 mm. from the anterior end to the tip of the tail. An anterior limb 

 bud is plainly visible upon surface view, and the tail is provided with a 

 distinct membranous fin. 



The great complication in the structure of the pronephros which is 

 attained in this stage is accomplished by a continuation of the same pro- 

 cess of forming convolutions that has been described for the preceding 

 stage. Indeed, the separation of the two stages is at best quite arbi- 

 trary. Figures 61-65 represent various pronephridia of the present 

 stage. It is to be noticed that the portion of the common trunk of 

 which the segmental duct is the direct continuation can be traced from 

 the anterior limit of the pronephros backwards without convolution, or 

 after having formed a few insignificant loops. The common trunk from 

 its junction with the nephrostomal tubules to this most anterior bend is 

 thrown into a series of complicated convolutions, which may be so arranged 

 as to present a gradation of considerable regularity (Fig. 62), or may 

 be quite irregular (Fig. 65). In most cases, however, it is to be noticed 

 that the arrangement of the loops is in general favorable for a compact 

 vol. xxi. — no. 5. 17 



