Guinea pigs and monkeys were placed in plague houses in pairs, both protected from soil contact infection and both equally exposed to aerial infection, but one surrounded with a layer of tangle-foot paper and the other surrounded with a layer of sand. The following observations were made: (_a_) Many fleas were caught in the tangle-foot, a certain proportion of which were found on dissection to contain in their stomachs abundant bacilli microscopically identical with plague bacilli. Out of eighty-five human fleas dissected only one contained these bacilli, while out of seventy-seven rat fleas twenty-three were found thus infected. (_b_) The animals surrounded with tangle-foot in no instance developed plague, while several (24 per cent) of the non-protected animals died of the disease. Thus, the experimental evidence that fleas transmit the plague from rat to rat, from rats to guinea pigs, and from rats to monkeys is indisputable. There is lacking direct experimental proof of its transfer from rodents to man but the whole chain of indirect evidence is so complete that there can be no doubt that such a transfer does occur so commonly that in the case of bubonic plague it must be regarded as the normal method. Rats are not the only animals naturally attacked by the plague but as already suggested, it occurs in various other rodents. In California the disease has spread from rats to ground squirrels (_Otospermophilus beecheyi_), a condition readily arising from the frequency of association of rats with the squirrels in the neighborhood of towns, and from the fact that the two species of fleas found on them are also found on rats. While the danger of the disease being conveyed from squirrels to man is comparatively slight, the menace in the situation is that the squirrels may become a more or less permanent reservoir of the disease and infect rats, which may come into more frequent contact with man. The tarbagan (_Arctomys bobac_), is a rodent found in North Manchuria, which is much prized for its fur. It is claimed that this animal is extremely susceptible to the plague and there is evidence to indicate that it was the primary source of the great outbreak of pneumonic plague which occurred in Manchuria and North China during the winter of 1910-11. Of fleas, any species which attacks both rodents and man may be an agent in the transmission of the plague. We have seen that in India the species most commonly implicated is the rat flea, _Xenopsylla cheopis_, (= _Lœmopsylla_ or _Pulex cheopis_) (fig. 89). This species has also been found commonly on rats in San Francisco. The cat flea, _Ctenocephalus felis_, the dog flea, _Ctenocephalus canis_, the human flea, _Pulex irritans_, the rat fleas, _Ceratophyllus fasciatus_ and _Ctenopsyllus musculi_ have all been shown to meet the conditions.