248 na. g. thin. 



nuclei by its round form and well-marked small multiple 

 nuclei, is seen in the figure. In short, what was evident 

 enough in the carmine preparations, and confirmed by the 

 osmic-acid and logwood preparations, is still further confirmed 

 by the purpurine, namely, that the ** cellule vaso-formative " 

 of Ranvier is not a cell, but a ramifying space of varying 

 extent. 



There is nothing of a special nature in these ramifying 

 spaces, either in the milky patches or in other parts of the 

 membrane, except their breadth. They exist to a less 

 developed extent over the whole omentum. In the omentum 

 of a new-born kitten they may be observed scattered over it 

 in a discrete form. They are simply spaces between the 

 bundles of connective tissue, such as occur in all forms of 

 that tissue, the special appearance they present in a given 

 instance being determined by the arrangement of the 

 bundles. 



This view is further supported by treatment with logwood- 

 glycerine. In order to obviate the destructive effects of 

 water on delicate tissues, I conceived the idea of colouring 

 glycerine by dropping into it a sufficient quantity of a con- 

 centrated solution of the dye, and placing the fresh tissue 

 which I wished to stain in the coloured glycerine. The 

 effects obtained by its use differ in many respects from those 

 observed when an aqueous solution of the dye is used. 

 When the mica square is placed in logwood-glycerine for 

 twenty-four hours, then washed in pure glycerine and 

 examined, the membrane is seen to be permeated by an 

 anastomosing blue network, of which the distended spaces 

 ("cellules vaso-formatives ") are seen to form a part. Few 

 nuclei are stained, even the lymph-cells are not much 

 coloured, and the connective tissue is colourless, but all the 

 wider spaces of the tissue are indicated by a faint blue 

 colour. 



In fig. 3 the effect of purpurine on the spaces is illus- 

 trated. The space, a spindle nucleus and part of the proto- 

 plasm of the spindle-cell, are to be distinguished, and a well- 

 marked lymph-corpuscle is seen at the wide end of the 

 space. 



What I have been able to make out regarding the con- 

 nection which these spaces have with the formation of blood- 

 vessels is exactly in accord with Ranvier's observations, if we 

 substitute for his idea of a cell that of a space. But 

 I have further observed that there is a peculiar development 

 of spindle-cells in connection with this new formation, in 

 regard to which that author makes no remarks, and which 



