10 



taining many granules are usually smaller and more oval. Those with 

 fewer granules are usually larger, more polygonal and in close relation 

 with the tissue reticulum. 



Many other stains on Flemming material show the granules. 

 Eosin and orange G give brilliant stains in their respective colors. 

 The latter acts very slowly, it being necessary to leave the sections 

 in a weak aqueous solution of the dye for several days. The granule 

 cells then can be easily recognized under a magnification of 85 dia- 

 meters. Biondi-Ehrlich-Heidenhain's tricolor stain, and Ehrlich's 

 triacid stain show no effect of the stains composing the mixtures ex- 

 cept that of orange G (somewhat modified in its tinge). Ehrlich's 

 amphophile stain shows many granules, but I cannot distinguish more 

 than one kind, the eosinophiles, the color being somewhat modified and 

 not quite so bright as eosin alone gives. There are no indulinophiles. 

 Gentian violet gives a sharp definition of the granules. Polychrome 

 methylene blue at 56 "^ C shows many granules sharply defined 

 but none colored distinctly blue. Several other basic dyes also show the 

 same granules. I believe the granules thus stained by the acid and 

 basic stains are identical, judging from their similarity in number and 

 distribution within the cells, and from the number, distribution, shape, 

 size, and nuclei of the cells themselves ^). As further tests I used 

 successive stains and combinations of stains, such as orange G and 

 gentian violet, on material fixed in Flemming's fluid and otherwise, 

 in order to discover whether it is possible to stain more than one kind 

 of granule cells in the same preparation, or more than one kind of 

 granules in the same cell. I was not able to get two classes of granules 

 stained at the same time ; one stain washed out or modified the other. 

 This method might at least be expected to show the two general classes, 

 basophiles and acidophiles, since both are fixed by absolute alcohol 

 at boiling; but I was unable to get such a result, probably for the 

 reason that alcoholic stains act less strongly than aqueous. 



Cells similar to the acidophiles were described by Heidenhain (4) 

 as one of the four classes of cells in the intestinal mucosa of the 

 dog. The acidophiles bear no relation to the plasma cells of Unna (10), 

 nor to those found in the pathological tissues by Krompegher (6). 

 Although I did not get them fixed in alcohol at ordinary temperatures, 

 the cells seem closely related to the coarse oxyphile granule cells de- 

 scribed by Kanthack and Hardy (5), and by Hardy and West- 

 brook (3). 



1) If this be true, the so-called acidophile granules would perhaps 

 more properly be classed as amphophiles. They are at any rate entirely 

 distinct from the basophile granules previously described. 



