Rat breeding, as well as flea breeding, is influenced by climate, but as the reproductive activity of the rat is most retarded by cold weather--an unknown condition in the Philippines--and as the climate of Manila is fairly equable so far as heat and cold are concerned, the only factor which needs to be considered is that of rainfall. As already mentioned, rainfall doubtless serves to drive rats above ground and so, to a certain extent, away from their nests in burrows and underground. Their well-known adaptability to changing conditions, however, permits them to house themselves comfortably above ground when driven out of these burrows and holes. JAVAN OBSERVATIONS.--The following conclusions were reached by Dr. J. J. van Loghem in a report upon "Some Epidemiological Facts Concerning the Plague in Java" (published by Civil Medical Service in Netherlands India-Batavia, 1912): 1. In plague-infected villages, as distinguished from plague-free villages, there exists a considerable mortality among house rats. 2. Rats in plague houses and plague quarters have repeatedly died from plague. Fresh plague rats appear more often in the houses adjoining plague houses than in the houses themselves. 3. The house rat exists even in the immediate vicinity of man. 4. The ordinary parasite of the house rat is _Xenopsylla cheopis_, which experimentally is known to choose man as a host when starving. 5. Fresh plague rats have repeatedly been found to harbor a great number of fleas. 6. Virulent plague bacilli have been demonstrated in the stomachs of such fleas. Concerning the prevention of plague by improving the native dwellings, the same observer says: "Obviously an increase in the distance between man and rat becomes an important factor as a means of preventing the disease." CONDITIONS OF MANILA HABITATIONS FAVORABLE TO RATS AND PLAGUE.--As shown by our own experiences in Manila, this end, the separation of rats and men, is not obtainable by destruction of rats by poison, traps and rat catchers. Rats dying of plague in their nests furnish the greatest danger to man. The plague problem, therefore, where rats are already infected, from the stand-point of direct prophylaxis, is the problem of dwellings. It was from this stand-point that we attacked the problem in the Tondo (Manila) campaign in 1913. MANILA VERIFICATION OF JAVAN OBSERVATIONS.--Having in mind the experiences of the plague investigators in Java during the recent epidemics there (1911-1912), we sought, from the time the Manila outbreak occurred, to verify some of the findings of the Java investigators, at least with special reference to the nesting of rats in close proximity to human beings and the consequent exposure of these persons to the infected fleas which desert the rats dying from plague in these nests.