SUMMARY AND CONCLUSIONS 201 



sion of the brain of a guinea-pig (p. 186) three days after 

 the return to normal temperature, a period when the blood is 

 believed to be no longer infective. 



We have demonstrated rickettsia in human and guinea-pig 

 tissues. In human tissues they have been found in blood ves- 

 sels in the skin, brain, kidneys, muscles, and testes, in fact in 

 all tissues the blood vessels of which become the sites of 

 lesions. As far as we know, physiological knowledge indicates 

 no differences between the blood vessels affected by typhus and 

 those not, by which the selective localization of Rickettsia 

 prowazeki can be explained. 



The demonstration of Rickettsia prowazeki in lesions de- 

 mands careful attention in matters of technic and tissues not 

 materially changed by post-mortem processes, and above all 

 rigid control. As there is no specific stain for Rickettsia 

 prowazeki the following criteria must be satisfied: (1) the size, 

 (2) morphology, (3) staining reactions must correspond with 

 those of Rickettsia prowazeki in sections of lice, and (4) they 

 must be present in vascular endothelial cells in situ in rela- 

 tion to the lesions of typhus. 



We have avoided presenting as evidence observations made 

 upon cells with secretory activities and nerve and neuroglia 

 cells, and cannot accept such observations by others as bear- 

 ing upon the question of the relationship of rickettsia to 

 typhus. 



We have not included in this report the fruitless attempts 

 we have made to cultivate Rickettsia prowazeki from guinea- 

 pigs. It does not seem possible to us that any one of the 

 bacteria cultivated from typhus patients and experimental ani- 

 mals can be Rickettsia prowazeki. The staining reactions and 

 morphology of Rickettsia prowazeki from lice furnish criteria 

 yet to be met by any culture. Our inability to preserve the 

 infectivity of typhus blood under a variety of conditions at 

 incubator temperature for more than four days is strong 

 additional evidence that none of the reported bacteria can be 

 Rickettsia prowazeki. 



