Gland-Cells of Internal Secretion in Spinal Cord of Skates. 19 
A few experiments were tried using other vital stains, some being 
composed of particles not so finely divided as in the case of trypan blue. 
Injections were made of janus green, methylene blue, carmine, and 
neutral red. In no case did the large cells in the spinal cord show any 
traces of the dye. 
The evidence from these experiments with vital dyes does not by any 
means prove conclusively that the cells may not have a phagocytic 
function. It is conceivable that they might ingest only the granules 
that are normally present in the spinal cord, the reaction being a more 
or less specific one, and reject the finely particulate matter of the vital 
dyes. This, however, is certainly not probable. At any rate, these 
results show that the cells are not of the same nature as the macrophage 
group of cells. 
CHEMICAL NATURE OF THE GRANULES. 
1. Tests for fat.—Pieces of spinal cord were fixed in acetic sublimate, 
absolute alcohol, Flemming’s fluid, and Perenyi’s fluid. Sections pre- 
pared from each of these fixations were immersed in (a) xylol and (6) 
ether. After 24 hours the sections were examined to see whether or not 
the granules had been dissolved by these fat solvents. They were 
present in all except the Perenyi material. (As was found out later, 
the granules are dissolved by strong nitric acid. Since Perenyi’s fluid 
contains nitric acid, this fact probably accounts for the absence of the 
granules.) Treatment with osmic acid failed to blacken the granules. 
Pieces of the spinal cord fixed in 95 per cent alcohol were sectioned 
and treated with Scharlach R dissolved in acetone alcohol. The gran- 
ules remained quite unstained, although the medullary nerve-sheaths 
stained a pinkish red. In fact, in all these fat tests the reactions of the 
granules could be very conveniently compared to the reactions of the 
fat of the medullary sheaths. In all cases the reactions of the two 
were quite different. It is evident, therefore, that the granules are not 
fatty or lipoid in character. 
2. Tests for protein.—The granules gave a positive Millon reaction. 
This result, of course, indicates their protein nature. The biuret, 
xanthoproteic, and Adamkiewicz tests were tried, but it was found 
impossible to complete any of these because of dissolution of the gran- 
ules by the reagents used. 
3. Tests for zymogen granules.—Mathews has shown that zymogen 
granules are speedily dissolved out by water and alcohol; that they are 
not preserved by alcohol, sometimes not by sublimate, nor by acetic 
sublimate. These granules are not dissolved out by water or alcohol, 
and are preserved by alcohol and by acetic sublimate. 
The granules are not easily soluble. Fresh pieces of spinal cord 
were placed in salt solutions of various strengths: 5, 10, 20, and 30 
per cent NaCl. Some pieces were placed in ordinary sea-water, some 
