INTERRELATIONS OF THE MESONEPHROS, ETC. 185 



cuboidal portions of the cells are far apart, of the lobulation of 

 the glomerulus to increase its free surface, and of the apparent 

 fusion of the plates with the underlying endothelium of the 

 capillaries, which thus seem to project uncovered into the inter- 

 capsular space (fig. 1). 



Long ago Drasch investigated the glomeruli of the kidney 

 and was able to distinguish two types, differing in size, lobulation, 

 and position, and also, more interesting in the present study, 

 differing in the kind of sheath (Hlille) which, by gentle shaking, 

 he was able to detach from the knot of blood vessels. In one 

 type of glomerulus this sheath contained nuclei, in the other 

 none, or only a very few, in both the imprint of the capillaries 

 was plainly visible. Von Ebner, in the 6th edition of Kolliker's 

 Gewebelehre, offers, I think, the correct interpretation of these 

 facts in supposing that in the one type of glomerulus the sur- 

 face of the cuboidal parts of the epithelium had become trans- 

 formed into an exoplasm similar in structure to the plates, 

 and that this surface layer with the plates could be detached, 

 leaving the nuclear portion connected with the capillaries. This 

 would account for the non-nucleated sheath; the whole epithelial 

 layer, plates and cuboidal portions detached together, would 

 furnish the other picture. Drasch has established two facts 

 of interest to us, first, that the blood vessels of the glomerulus 

 are actually covered by thin plates, which can be separated from 

 the endothelium by certain artificial means, though in sections 

 it is usually impossible to distinguish more than a single layer, 

 so closely are the two applied; and secondly, that these plates 

 are of a highlj^ differentiated protoplasm, non-granular and rather 

 stiff, in that they hold their shape even after being removed from 

 the capillaries about which they have been molded. The first 

 of these facts is important as explaining a very natural mistake 

 made by Duval in his description of the placenta of the rabbit; 

 the second shows that these modified plates have apparently 

 become inactive membranes, through which a purely physical 

 osmosis may take place, but which themselves may be sup- 

 posed to be physiologically inert. A physiological activity is 



