140 RICKETTSIA PROWAZEKI IN LICE 



serial sections to determine their shapes and sizes. Some are 

 tapered and some contain clear spaces. They seem to be 

 parasitic in nature. In some sections of experimental lice the 

 rickettsia stain a brilliant red. In such preparations the chains 

 of bacilliform rickettsia bodies may resemble the smaller 

 specimens of the filamentous bodies. Indeed, in sections 

 where both "filaments" and chains of bacilliform rickettsia 

 occur it may be impossible to tell whether rickettsia chains or 

 "filaments" lie in the field. 



During prolonged search for rickettsia in sections of negative 

 lice, bright red stained granules 0.5/x to l.Oju in diameter have 

 been found singly and in pairs in epithelial cells of the gut. 

 These granules are quite distinct from precipitated stain. 

 Their nature is wholly undetermined. In one louse, collec- 

 tions of uniformly red staining granules were seen in the fat 

 body (see louse Box xxxvm). These collections measured 

 about 20/4 by 30ju; and seemed to have no definite limiting 

 membrane. The granules which they contained varied in 

 size but were usually about 2ju to 3/z in diameter. 



In poorly fixed lice, where the fixative has not had free 

 access to the organs owing to insufficient incision of the abdo- 

 men, the rickettsia in many greatly swollen cells appear as 

 poorly-stained, poorly-defined granules and their presence 

 can be confirmed only because of the great distension of the 

 cell and the presence of unmistakable rickettsia in other parts 

 of the same louse. In some experimentally fed lice upon typhus 

 patients which we have recorded as negative moderately 

 swollen cells with granular cytoplasm may actually have con- 

 tained rickettsia. 



It is also possible that a small proportion of the lice noted 

 as uninfected in Table VIII harbored small numbers of rick- 

 ettsia. Occasionally (Box XLVIII, louse 18) an isolated cell 

 packed with rickettsia in its bacillary form was found. A 

 casual search of ten or fifteen minutes would usually fail to 

 detect such a slight infection; and were a smear made of 

 such a gut, the infection would almost certainly escape notice 

 (Box LV, louse 4). 



