792 
variety of perivascular cells exists, the second variety rather corre- 
sponding to the plasma-cells of WALDEYER than to the vacuoles of 
KUPFFER. 
In the literature on the liver we have found descriptions of two 
other cells, one by Ponrick1) in the frog, who describes small cells 
with fine undulating contours, and with homogeneous, or at the most, 
finely granular protoplasm, of rounded or oval form, containing one or 
two nuclei; the other is by Born ?), also in the frog’s liver, who 
working with osmic acid as a fixation agent, found moderately large 
cells with rounded granular nuclei and strong processes, with around 
the nucleus a light granulation of the protoplasm with strongly re- 
fractile granules situated in it, which are continued into the proto- 
plasmic rami, and are either of pigmentary or fatty nature, for they 
assume a dark appearance after osmic acid treatment. 
Our cells take on no dark staining after exposure to an osmic 
acid solution, and hence are probably not fatty in nature; but other- 
wise BoLL’s description corresponds in some measure to our granular 
bodies. 
All the perivascular cells undoubtedly belong to the connective 
tissue system, and from their position along the course of the capil- 
laries must have some function in assimilating from the blood-vessels 
certain of their serous or cellular contents, and transmitting them, 
probably after alteration, to the more highly organized tissues sur- 
rounding, and in contact with them. 
The presence of the same granular refractile particles in the walls 
of the blood-capillaries lends aid to the view that certain products are 
absorbed by the endothelial sheath of the vessel and passed on to 
the granular cells, where it is modified and perhaps used for purposes 
of nutrition of some of the adjacent cellular bodies. 
In short, our idea of these bodies is that the protoplasmic sub- 
Stance, even to the extremities, is charged with granules absorbed from 
the blood-vessel’s contents, and that these granules are only rendered 
apparent by the method of treatment, and while not usually visible 
become so by this method. 
May, 10, 1893. 
Eingegangen: Teil I (ohne Figuren) am 28. Mai, II und III nebst 
Figuren am 21. Juli. 
1) Ponricx, Virchow’s Arch., Bd. XLVIII, p. 1. 
2) Bout, Arch. f. mik. Anat., Bd. V, p. 349. 
m a 
