192 PATHOLOGY OF TYPHUS IN MAN 



Situation is of first importance the localization Jn en- 

 dothelial cells, which do not contain granules of any sort except 

 in stages preliminary to division. We find that the granules 

 stainable by Giemsa's stain in endothelial cells about to undergo 

 or in mitosis are of varying size and irregular shapes, but it 

 has been our practice to disregard such cells in the search for 

 rickettsia. In questionable instances, moving to the section 

 next in series settles the question of nuclear activity. The 

 granules in mast cells stain differently and more deeply than 

 rickettsia, a deep purple or red in contrast to the pale blue of 

 the latter (Plate XXIX, figs. 73 and 74). They are larger than 

 rickettsia but are often paired and may be removed from the 

 nucleus as this cell is extraordinarily amoeboid. Mast cells 

 rarely penetrate vessels. The invariable use of a series of sec- 

 tions mounted upon one slide always enabled us to trace out 

 the delicate processes of amoeboid mast cells to their nuclei. 

 Pigment cells are so distinctive that the possibility of mistak- 

 ing their granules for rickettsia does not exist for the experi- 

 enced histologist. 1 



Other inclusions phagocytized by endothelial cells are 

 larger, of irregular shapes and staining reaction. 



Our control material includes a large variety of pathological 

 skin conditions, infarctions, burns, X-ray radiation, petechiae 

 in septicaemias, measles, scarlet fever, chicken-pox, and Rocky 



1 A paper published by Stevenson and Balfour, since the completion of our manu- 

 script, in the Journal of Pathology and Bacteriology, July, 1921, describes and illustrates 

 granules in human tissues which the authors regard as consistent with rickettsia. They 

 do not, however, express an opinion as to the "true nature" of the granules. 



The distribution of the granules, noted by Stevenson and Balfour (ganglion cells of 

 the brain, larger vessels of the lungs, phagocytic cells of the spleen, the parenchyma and 

 Kupfer cells of the liver, pancreatic cells, kidney cells, and intestinal gland epithelium), 

 does not coincide with the distribution of rickettsia as described by us in this report. 



The fixative used in preserving the tissues of the five cases studied makes impossible 

 an exact comparison of their findings with ours. We have briefly discussed on page 191 

 our criteria for the recognition of rickettsia in tissues. Since Stevenson and Balfour 

 found rickettsia-like (?) bodies in the parenchyma cells of several organs where we did 

 not find them, and since they failed to find them in the endothelium of blood vessels of 

 the skin where in our experience rickettsia are most numerous, we are forced to regard 

 their technic and criteria for identification of rickettsia in tissues as unreliable. 



They failed to note that in a preliminary report (1920) we recorded the presence of 

 Rickettsia prowazeki in the lesions of typhus in human and animal tissues. 



