84 
BULLETIN OF THE BUREAU OF FISHERIES 
preferably after Bensley’s corrosive bichromate fixative. The iron haematoxylin, 
differentiated to just the right degree, stains the granules an intense bluish black, while 
the rest of the cytoplasm is clear or may be counterstain ed with eosin. The cell granules 
are relatively uniform in diameter. In many preparations one can distinguish only a 
slight variation in the size of the granules. In certain preparations, however, there is 
considerable variation in the diameters of the granules within a single cell, a variation 
that is uniform throughout the cells of a given stomach. The size of the granules in a 
series of specimens measured varied around 0.8 to i in diameter. The extremes noted 
in these preparations were 0.4 and 1.2 fi, respectively. In several samples where the 
granules were especially uniform in diameter and in which the granules crowded the cell 
cytoplasm for its full extent the size ran from 0.4 to 0.5 ix in diameter. 
The salmon pancreatic cells and the gastric gland cells also contain granules, zymogen 
granules. But the number and arrangement of granules in each of these two types of 
cells is characteristically different from that of the granule cells proper. Measurements 
show that the zymogen granules of the pancreatic cells also_ vary through a rather wide 
range. In sections from the same salmon from which the measurements of granule cells 
given above were taken the pancreatic zymogen granules measure from 0.6 to 1.5 in 
diameter. The number of granules in the individual cells was small and the granules 
were located chiefly on one side of the cell, presumably the side next the lumen. In 
another fish the granules of the cells of the pancreas measured from 0.4 to i /i in diameter. 
The loading of granules was somewhat greater in number than in the cells with the 
larger granules. 
The granules of the granule cells take stains somewhat differently from the granules 
of either of the two types of secreting cells under comparison. The pancreatic cells and 
the granule cells of the pyloric cjeca present in the same microscopic field are always 
in sharp contrast. The stain of the granules of the granule cells is sharper, more distinct 
and brilliant, and of a different shade of color when both are stained with the same dye. 
Evidence of solution of the granules of these peculiar cells has been observed in the 
preparation of tissues by certain fixatives. Unfortunately this factor has not been 
investigated sufficiently for final report. 
The character of the distribution of the cells, the staining properties of the granules, 
and the persistence of these cells during marked degenerative changes of much of the 
structure of the stomach and other portions of the alimentary canal all point to some 
function of special nature. This function, I believe, is that of internal secretion. The 
granules present are, therefore, true zymogen granules. In another connection I have 
presented the view that this function is one of lipase production. The absence of any 
tubular arrangement or ducts, and the apparent lack of an adequate capillary net would 
suggest that the granule cells must discharge their products directly into the surrounding 
tissue spaces from which distribution takes place. 
The granule cells were observed by Gulland.® He figures a bit of the stratum 
compactum and adherent granule cells in its meshes. Gulland calls these cells “large 
“ Paton, D. Noel, op. cit., pi. 4, fig. 10. 
