YELLOW FEVER EXPEDITION 493 
6. The most successful staining reagent is carbol fuchsin solution (Ziehl), 
diluted with 5 per cent, phenol solution (to prevent accidental contamination during 
the long staining period) ; immersion for several hours, followed by differentiation 
in weak acetic acid. Two hours staining period may fail to reveal bacilli, which appear 
after twelve to eighteen hours. The bacilli in the stools are often of greater length 
than those in the tissues, and they may stain rather more easily ; naturally the same 
is true of cultures. 
7. Since the bacilli are small and comparatively few in numbers, they are difficult 
to find. To facilitate matters at our last two autopsies (14th and 15th), a method of 
sedimentation has been adopted. A considerable quantity of organ juice is emulsified 
with antiseptic solutions, minute precautions against contamination, and for control 
being taken ; the emulsion is shaken from time to time and allowed to settle. The 
method is successful and may form a ready means of preserving bacteria — -containing 
material for future study. The best fluid for the purpose has yet to be worked out ; 
hitherto normal saline with about one-fifth per cent, sublimate has been employed. 
8. Pure growths of these bacilli are not obtained in ordinary aerobic and anaerobic 
culture tubes. 
9. Some pure cultures have been obtained by placing whole mesenteric glands 
(cut out by means of the thermocautery) into broth under strict hydrogen atmosphere. 
Investigation into the necessary constitution of culture media for successful cultivation 
is in progress. 
10. Much search was made for parasites of the nature of protozoa. We conclude 
that yellow fever is not due to this class of parasite. Our examinations were made 
on very fresh organ -juices, blood, etc., taken at various stages of the disease, with and 
without centrifugalizations, 1 and on specimens fixed and stained in appropriate ways. 
We may add that we have sometimes examined the organs in the fresh state under 
the microscope within half an hour of death. 
11. The endeavour to prove a man-to-man transference of yellow fever by means 
of a particular kind of gnat by the recent American Commission, is hardly intelligible 
for a bacillary disease. Moreover, it does not seem to be borne out by their experi- 
ments, nor does it appear to satisfy certain endemiological conditions. It is proposed 
to deal more fully with the endemiology and epidemiology of the disease on a later 
occasion. 
12. We think that the evidence in favour of the etiological importance of the 
fine small bacillus is stronger than any that has yet been adduced for any other pre- 
tended ' yellow fever germ.' At the same time there is much further work to be 
done ere its final establishment can be claimed. The acquisition of a new intestinal 
bacterium would explain the immunity of the ' acclimatized.' 
1. We have found this sometimes useful in examining the blood of ague patients. 
L I 
