COUNTING GRANULES 277 



one is apt to count the same granule more than once, 

 and it is easy to lose one's place in which case it 

 becomes necessary to begin all over again. On looking 

 at the cells through the microscope, the granules appear 

 as though they might with care be counted, and it is 

 most inviting to attempt to do this and to rely on it; 

 but on testing this rough-and-ready method we have 

 found that it usually involves an error of nearly 50 per 

 cent. No estimate whatever can be made of the 

 number of granules contained in a cell by merely 

 looking at it through the microscope, no matter what 

 magnification is used. 



Obviously the granules must be stained, and then 

 it is necessary: (1) to distinguish readily between an 

 eosinophile and a basophile leucocyte; (2) to kill the 

 cells, and then to burst them so as to cause their 

 stained granules to rest discretely side by side in one 

 plane and not on top of one another; (3) to magnify 

 the image of the ruptured cell in such a way that one 

 can "tick off" each granule with a pencil on paper 

 as it is counted, so as to avoid counting the same 

 granule twice over. 



By the following procedure the staining, killing, 

 differentiation, and bursting can be readily accom- 

 plished. In order to magnify the image of the ruptured 

 cell so as to count its granules and to "tick them off," 

 it is necessary to obtain a photomicrograph negative of 

 it, and then to project the photographed image on to 

 a paper screen with an optical lantern, when the image 

 of each granule can be marked off on the paper with 

 a pencil. 



