it cannot be said that as yet we are even near the complete solution of the problem. One thing, however, is fairly well ascertained, and that is that these germs are often introduced by means of drinking-water. The statistically ascertained improvement in the public health in respect of dysentery in such large towns as Calcutta and Madras, following so closely on the introduction of improved water supplies, and the improvement in the health of the British Navy following the introduction of regulations requiring that in all places in which the water supply is not above suspicion the drinking-water served out to the men shall be distilled, constitute powerful testimony in favour of regarding dysentery as a water-borne disease. This conclusion receives additional support from the occurrence of epidemics of dysentery in the crews of ships which have watered at polluted sources, as well as from the occurrence of similar epidemics in large institutions in which, by some accident, surface water has leaked into the water supply. This does not exclude the possibility of other sources of infection, such as privies, dust, flies (especially), fouled vegetables, and the vessels or instruments used by dysenteries.
The germs of dysentery may be fly-borne. —In certain parts of the world Fiji, for example where water can be definitely excluded as a source of infection and where extensive epidemics of dysentery are of annual occurrence, it seems probable that the germ is conveyed and diffused by house-flies. Bahr has shown by direct observation that flies fed on fæcal matter containing dysentery bacilli ingest the germ, which they may subsequently deposit on human food either in their own fæces, or by regurgitation, or by their soiled feet. This observer associates the annual dysentery epidemics in Fiji with the concurrent enormous multiplication of house-flies, and has isolated dysentery bacilli from the intestinal tract of wild flies caught in the vicinity of patients suffering from the disease. In the case of the dysentery bacillus he has cor firmed Graham-Smith's observation on the persistent vitality and virulence of bacilli of
