4 88 
THOMPSON YATES LABORATORIES ; REPORT 
For stained specimens careful fixation with formaldehyde and osmic vapour must 
be practised, and a variety of staining methods — ' Romanowsky,' haematoxylin, acidi- 
fied basic stains, etc, — tried. Lastly, any odd or unfamiliar object must be sketched 
and its size determined. At autopsies it will be necessary to examine the probable 
paths of infection ; that is to say, the superficial lymph glands in various regions. 
Should events guide satisfactorily it will be necessary to examine possible intermediary 
agents such as gnats. 
Several considerations have to be borne in mind in enquiring into a ' new ' 
disease from the bacterial point of view. In the first place the gross lesions, especially 
those ot the liver in yellow fever, give no probability that the bacterium is especially 
localized in these parts ; for instance, a very small amount of the botulismus toxin 
is sufficient to produce most intense fatty change in the liver, provided the animal 
lives long enough. This time factor is a consideration which apparently has not entered 
the minds of experimenters who have endeavoured to reproduce yellow fever in animals, 
it is unlikely that the lesions could be produced when a period of twenty-four to 
forty-eight hours only elapses between inoculation and death. In making experiments 
on animals it will be necessary to gauge the doses to allow the animal to survive 
several days. 
It is unlikely that a highly septicaemic condition or that large numbers of 
bacteria become localised in the tissues (as in plague) otherwise it is difficult to under- 
stand why such bacterium has hitherto defied recognition. The two bacterial diseases 
which seem to be most apt tor analogy, or the pathology of which might be kept to the 
fore, are botulismus and tetanus. In the former van Ermengen only succeeded in 
getting very few colonies of living bacilli from the organs of the fatal cases ; but 
whilst he thus proved that actual infection had taken place he was able to reproduce 
the disease by absorption ot toxin through the alimentary canal. Tetanus is also 
not particularly encouraging, for, with a scarcely discoverable number of individual 
bacilli, fearful havoc is wrought by intoxication. What has been already written 
in regard to paths of infection in a protozoal disease can only be reiterated for a 
bacillary one ; it may be that in or about the site of primary infection it may be more 
easy to discover an organism, or, by examining the bottom of centrifugalized material. 
In the symptoms as described there does not appear to be much that suggests 
localizing — the vomiting, kidney and liver mischief might be purely toxic. On the 
other hand, pain in the epigastrium, upon which some authors lay much stress, is 
suggestive of some lesion possibly due to special localization. 
It is not likely that the supposed bacterium will thrive on ordinary media, it 
might be, however, that insufficiently large proportions of tissue have been taken. 
This will apply to anaerobic as well as aerobic trials. 
As general routine it will be sufficient to use liquid broth cultures in the first 
instance ; any organisms that should grow freely must be found in pure condition 
(by observation and plating) ere it can be granted an etiological position. 
Lastly, it must be remembered that a bacterium may produce very different 
effects according to the mode by which it is introduced into the system. (Compare 
experimental inoculation of animals with certain bacteria). 
