made in India in 1906 show conclusively that close contact
of plague-infected animals with healthy animals does not give rise to
any epidemic, so long as the passage of fleas from infected to healthy
animals is prevented. When opportunity, however, was given for fleas to
pass from one animal to another, the bacillus and the disease was
generally carried over. It has also been found that while this species
of fleas have their normal residence on the body of rats, they will also
desert a rat for man, if the infected rat is dying and no healthy rat is
in the vicinity to receive them. It is, then, obvious that to eliminate
the disease, the most direct and positive course is to destroy the rats
which are the home of the disease.

In India, where the plague appeared in 1896, causing about 300 deaths,
it rapidly increased in virulence until in 1907 it caused 1,200,000
deaths. The ports of the Pacific coast became much alarmed, and when
cases of the disease were actually found in San Francisco in 1906, the
matter was so terrifying that the United States Marine Hospital Service
was at once instructed to stamp out the disease if possible. This
procedure was directed almost entirely against rats. Deposits of garbage
on which rats might feed were removed, rat runs and burrows were
destroyed and filled in, and stables, granaries, markets, and cellars
where rats might abound were made ratproof by means of concrete. Rats
were trapped and poisoned by the thousand, nearly a million being thus
disposed of. As a result of such thorough work, the plague was stayed,
and in 1909 not a single case of the disease among human beings was
found, and although 93,558 rats captured were examined, only four cases
of rat plague were found.

In southern California, however, the fleas deserted the rats for ground
squirrels, and one county in particular, Contra Costa County, had an
epidemic which caused the squirrels to die by the thousands. The
attention of the scientists was thus turned to the squirrel as a host of
the flea, and a warfare similar to that against the rat has been for a
year past carried on against the infected squirrels. Between September
24, 1908, and April 12, 1909, 4722 ground squirrels were killed and
examined for plague infection, and from June 4 to August 13, 1909, the
work being continued, 178 squirrels were found to have the plague.

Now that the relation between fleas and their hosts and the transmission
of the disease is kno