320 PATHOGENIC BACTERIA.

cholera excrement; and how many other interesting in-
fections are made possible. The literature upon these
subjects is so vast that in a sketch of this kind it is
scarcely possible to mention even the most instructive
examples. One physician is reported to have been in-
fected with cholera while experimenting with the spirilla
in Koch's laboratory.

The evidence of the specificity of the cholera spirillum
when collected shows that it is present in the choleraic
dejections with great regularity, and that it is as con-
stantly absent from the dejecta of healthy individuals
and those suffering from other diseases; but these facts
do not admit of satisfactory proof by experimentation
upon animals. Animals are never affected by any dis-
ease similar to cholera during the epidemics, nor do foods
mixed with cholera discharges or with pure cultures of
the cholera spirillum affect them. This being true, we
are prepared to receive the further information that sub-
cutaneous injections of the spirilla are often without
serious consequences, though cultures differ very much
in this respect, some always causing a fatal septicemia in
guinea-pigs, others being as constantly harmless.

Intraperitoneal injection of the virulent cultures pro-
duces a fatal peritonitis in guinea-pigs.

One reason that animals and certain men are immune
to the disease seems to be found in the distinct acidity
of the normal gastric juice, and the destruction of the spi-
rilla by it. Supposing that this might be the case, Nicati
and Rietsch, Von Ermengen and Koch, have suggested
methods by which the micro-organisms can be introduced
directly into the intestine. The first-named investigators
ligated the common bile-duct of guinea-pigs, and then in-
jected the spirilla into the duodenum with a hypodermic
needle. The result was that the animals usually died, some-
times with choleraic symptoms ; but the excessively grave
nature of the operation upon such a small and delicately
constituted animal as a guinea-pig greatly lessens the value
of the experiment. Koch's method is much more satisfac-