the rat, and the rat-flea, Xenopsylla cheopis, and other fleas, such as Ctenocephalus canis (Fig. 68), which is known to bite dog, man, and rat indifferently. This important conclusion is completely supported by the elaborate epidemiological observations of Ashburton Thompson in Sydney, who has shown that the rat-flea theory best explains the behaviour of epidemic plague—at all events in that city.
The flea communicates the plague bacillus either on its fouled mandibles, or by regurgitation of the contents of its stomach during the act of sucking, or by provoking scratching and consequent inoculation of the bacilli deposited in its fæces on the skin.
Fig. 68.—Ctenocephalus canis (female).
C. J. Martin's observations on plague-infected fleas are of great interest. He found that a proportion of the fleas fed on plague-infected rats develop a peculiar pathological condition of stomach and œsophagus, these organs becoming blocked with blood-clot containing a pure culture of B. pestis. When such a flea feeds on a normal rat, part of the culture regurgitates and communicates infection; at the same time bacilli are passed in the fæces and may infect through any existing abrasion. He further observed that the "blocked" fleas died very rapidly if placed in a warm, dry atmosphere, apparently of thirst. (Figs. 69, 70.)
In temperate climates fleas are most numerous during the warmer seasons of the year; hence summer and autumn is the bubonic plague season in such climates. In warm climates bubonic plague is most prone to become epidemic at those times of the year in which temperature ranges between 10° and 30° C. temperatures favourable to the multiplication and activity of the flea. Temperatures over 30° C. are unfavourable to the flea, especially if the atmosphere is
