J. A. Arkwright, E. E. Atkin and A. Bacot 29 
observed and this organism lies on the surface of the epithelial cells. It has 
been cultivated on blood agar by Noeller and Jungmann. 
(8) R. ctenocephali was found by Sikora ( loc . cit.) in 20 out of 100 cat fleas 
examined. It is said to be very like R. quintana. 
(9) The last-named writer found similar forms in smears from the Mal¬ 
pighian tubes of a mouse flea. 
Rickettsia in C. lectularius. 
Our knowledge of the form we are about to describe, which has hitherto 
apparently escaped attention, resulted from an attempt to infect bugs with 
the virus of trench fever by feeding them on patients suffering from this 
disease. Examinations of smears made from the guts of these insects showed 
in nearly every case thread-like “bacteria” with some admixture of shorter, 
rod-like forms (Plate II, fig. 2). 
Stained with Giemsa’s stain there appeared to be an outer sheath which 
took the eosin rather lightly while in the interior were granules or groups of 
granules which stained more deeply and of a purplish hue. 
In some cases these smears also showed numbers of small deeply staining 
coccal or diplococcal bodies which, although slightly larger than Rickettsia 
found in gut and excreta smears of lice that had been fed on trench fever 
patients, were still sufficiently like them in size and general appearance to 
suggest that they might be the same species modified by development in the 
body of an unusual host. The fact that these bodies were not detected in any 
of the earlier smears from control bugs, although the thread-like forms above 
described were present, led to unsuccessful attempts to infect two volunteers 
with the emulsified guts of infected bugs in which the Rickettsia-like bodies 
were present. The examination of further control smears showed, however, 
that these minute bodies were also present in bugs that had fed only on 
normal men. 
Suspicion of a relation between the rod and thread forms and these 
minute bodies arose owing to the occurrence of darkly stained granules in the 
rod and thread forms. 
That these minute bodies frequently escaped our notice in smears which 
showed the long bacterial forms, is due partly to their small size, which makes 
it difficult to recognise them unless a considerable number are present in a 
single field; the same difficulty has been found in the case of the Kickettsia 
of trench fever. The fact that their distribution is often very localised is, 
however, the chief cause of difficulty in their detection. Whereas the bacillary 
forms are nearly always generally distributed in smaller or larger numbers 
in smears of well teased guts or Malpighian tubules, the minute forms, even 
when present in large numbers, are frequently only to be found in proximity 
to fragments of the gut or tubules. In some smears they were only seen 
where they chanced to be escaping from the ruptured end of the gut or a 
