MUSEUM OF COMPARATIVE ZOOLOGY. 235 



other and the intercellular regions precisely corresponded. In such 

 cases the evidence seems strongly in favor of delamination, but I have 

 never seen a nucleus dividing in that direction. This negative evi- 

 dence, however, should have little weight, since all cell divisions occupy 

 a comparatively short time, and are also obscured by the numerous yolk 

 spherules. The observations just recorded agree very well with the 

 rapid thinning of the somatopleure to which 1 alluded in discussing the 

 histology of the tubules. 1 conclude, then, that the mesenchymatic tissue 

 of the pronephros arises from the adjacent somatopleure, and probably 

 also to some extent from the capsular membrane. 



The glomus (Plate 1. Fig, 10) has attained in this stage nearly its final 

 dimensions. The lateral plate having become wholly detached from the 

 protovertebrre, the glomus has the appearance of being attached to the wall 

 of the body cavity at its dorsal angle (compare Fig. 9, of a younger stage, 

 and Fig. 47, of Bufo). There is some individual variation, but in general 

 it may be stated that the ridge constituting the organ under consideration 

 extends continuously from the first nephrostome backward to a position 

 slightly behind the third. It appears in cross section (Fig. 10) obovate, 

 being attached by the narrower end. In structure it is very compact, 

 so that it is difficult to locate with precision cell boundaries in the dense 

 interior mass. The investing portion consists of a single layer of cells, 

 which is continuous with the peritoneum. These cells are large, and 

 have the form of spheres flattened on their inner surfaces (compare Plate 

 VI. Figs. 49 and 50, la. pVton., which represent this layer in Bufo). They 

 are slightly pigmented, and a distinct row of pigment granules can usually 

 be seen close to the inner surface of the layer. These outer cells are 

 evidently the representatives of the large cells of the splanchnopleure 

 which was folded, at a previous stage (see page 227), to form the earliest 

 fundament of the glomus (Fig. 9, jnd. glm.). In certain favorable re- 

 gions I have seen a thin structureless membrane lying directly within the 

 outer cellular layer (compare Figs. 49, 50, rnb.ba.). When any of the 

 cells of that layer become detached, which frequently happens, this base- 

 ment membrane usually remains in place, and gives a sharp outer con- 

 tour to the glomus in that region. Besides the compact mass of large 

 cells there occur within the organ one cr two cells in each section (com- 

 pare Figs. 49, 50, enUh.) which have an elongated form. They lie close 

 to the basement membrane, with which their long axes are parallel. In 

 sections each cell has a central swollen portion containing the nucleus, 

 from which it tapers in both directions. 1 have not been able to trace 

 the delicate lateral portions to their terminations, but I believe that 



