274 E. V. Cowdry, 



In the case of the nongranular leucocytes (large and small 

 lymphocytes of Weidenreich) the only structures with which these 

 bodies are likely to be confused are the azurophile granules origi- 

 nally described by Michaelis and Wolff ('02, p. 159) and the , Vacuoles 

 et grains de segregation' of Renaut ('07, p. 497) and Dubreuil ('13, 

 p. 78). All the visible granules in these cells stain specifically with 

 both janus green and diethylsafranin. This does not mean that the 

 bodies observed by me are azurophile granules, for the azurophile 

 granules, as seen in heat fixed smears of the human blood stained 

 with methylene blue erythrosinate (Kyes), personal communication 

 are characterized by their irregular number, size and shape, while 

 the granules to which I refer show but little variation in these 

 respects. Neither does it mean that the bodies are segregation 

 granules, because Dubreuil ('13, p. 84) found that mitochondria 

 were distinct from the granules of segregation in the nongranular 

 leucocytes of the dog, rabbit and certain other mammals. Further- 

 more, Dubreuil's observations were made chiefly upon lymph drawn 

 from the thoracic duct. He produced no evidence of the occurrence 

 of granules of segregation in the cells of the circulating blood 

 of man. 



In the finely granular leucocytes (neutrophil es) these granules 

 may be differentiated from the specific granulations of Ehrlich by (1) 

 their affinity for janus green and dietl^lsafranin (vide fig. 5 a and b); 

 (2) their uniform rod-like shape and (3) the fact that they are not 

 so prone to exhibit Brownian movement as the specific granulations. 



5. Comparison with bodies known to be mitochondria. 



The relative amount, morphology and cytoplasmic distribution of 

 these structures is identical with that of mitochondria as described in 

 the fixed and stained blood cells of lower forms by Meves ('10, p. 656), 

 Ciaccio ('11, p. 15), Alagna ('11, p. 32), Dubreuil ('13, p. 74) and 

 others. They vary, from granules, which tend to be arranged in rows, 

 to rods and filaments, sometimes straight, at others curved and bent, 

 but retaining their individuality and never coalescing to form a 

 reticulum. Hence the term , mitochondria' from ft hog a thread and 

 XOvôqoç a grain (Barratt, '13, p. 553). 



