becomes acclimated in the pigeon, makes it sick and
somnolent more quickly, kills it sooner, and the blood
of the last pigeon injected into the pig, manifests there
a virulence superior to that of the most infectious mate-
rial from a pig which has died of erysipelas, even if the
pig was naturally infected. Here we have then aug-
mentation of virulence for the pig by passing the or-
ganism through the pigeon. The maximum to which a
virus can attain by passage through a race is, therefore,
not always, the maximum for that race.

There we have a case of augmentation, here is a case of
attenuation to which I wish especially to call attention.
Let us substitute the rabbit for the pigeon in this series
of experiments. The microbe becomes accustomed to
the rabbit; all the animals die, but if we inoculate pigs
with the blood of the last rabbits for comparison with
that taken from the first rabbits in the series, we find a
progressive diminution of virulence. Soon the blood
of rabbits, inoculated into pigs, no longer kills them;
it only makes them sick and leaves them vaccinated
against the fatal erysipelas. Entirely parallel facts
have been worked out with other microbes. They
furnish a method of attenuation of viruses by passages
through living species, and increase our means of action
in a field of studies the future of which will show its
astonishing fruitfulness. ¹

We have now come back, apparently, to a conclusion
already stated: Virulence is a state of perpetual becoming.
But how much we have developed this idea, and what
precision the new facts have given to it, and to the bond
of theory which has enabled us to unite them! In the

-311-