I.—PHYSIOLOGY. 219 
of African sleeping-sickuess a hopeful possibility, Brown and Pearce tind 
it necessary to introduce yet other considerations to explain its effects. 
Tested by Ehrlich’s therapeutic index—the ratio between the lowest 
curative and the highest non-toxic dose—it gives a relatively unfavourable 
figure. Brown and Pearce practically abandon the attempt to account for 
its action on the supposition that it directly kills the parasites, and attri- 
bute its value largely to its power of penetrating easily into the tissues and 
reinforcing there the processes of natural resistance. 
ili. ACTION oF BISMUTH. 
Another conception of the mode of action of these arsenical remedies, 
also involving a direct participation in the host’s tissues, was put forward 
by Levaditi. He found that from atoxyl a directly parasiticidal prepara- 
tion could be obtained, by incubating it with an emulsion of fresh liver 
substance. As the first step, therefore, in the curative action of atoxyl, 
he postulated a combination of its reduction product with some constituent 
of the liver or other tissue, giving rise to the essential curative complex, 
which he named ‘ trypanotoxyl.’ Levaditi’s observations were explained 
by Ehrlich and Roehl as due simply to the reducing action of the liver 
substance on atoxyl ; but it would be difficult to apply this explanation to 
the quite recently published observations by Levaditi and his colleagues, 
on the mode of action of bismuth in curing spirochetal infections. A 
sodium potassium bismuthyl tartrate—a bismuth analogue of tartar 
emetic—had been found to have valuable curative properties in syphilis 
and other spirochetal infections. Later, various other bismuth salts, 
bismuth suboxide, and even finely divided metallic bismuth, were found 
to produce similar effects. According to Levaditi and Nicolau, these 
preparations have, by themselves, a relatively weak action, or none at all, 
on the spirochets outside the body. If they are mixed, however, with a 
cell-free extract of liver, which is itself harmless to spirochets, the mixture, 
after incubation, acquires a potent spirocheticidal action. The possibility 
of a mere reducing action of the liver extract seems here to be excluded, 
since bismuthous oxide, or metallic bismuth itself, yields a spirocheeticidal 
mixture, containing Levaditi’s hypothetical ‘ bismoxyl,’ when incubated 
with the liver extract. If these observations are confirmed, there will be a 
strong indication that some cell-constituent enters into the composition of, 
or is essential to the formation of, the directly active substance from any 
of the derivatives of arsenic, antimony, or bismuth, as a preliminary to its 
action on an infection due to a trypanosome or a spirocheet. Again we have 
evidence of an organotropic property of the remedy, as an essential 
condition of its activity. 
iv. RESISTANT STRAINS OF TRYPANOSOMES. 
In the phenomena of the acquisition of resistance, by a strain of infect- 
ing trypanosomes to a particular curative drug, discovered and largely 
worked out in Ehrlich’s laboratory, we meet again with facts which can 
only with the greatest difficulty be reconciled with the assumption that 
the drug directly attacks the parasites. It was found, for example, that 
if a mouse infected with trypanosomes received an incompletely effective 
series of doses of atoxyl, the trypanosomes appearing in the blood at each 
relapse were more and more resistant to the drug, until they could not 
