A second illustration may be cited at Butler, Pennsylvania, which occurred in 1903. The water-supply of Butler, a borough of 16,000 people, comes from a reservoir on the creek which flows through the phase. On account of the gross pollution of the water at the pumping-station, a long supply pipe has been laid from the reservoir directly to the pumps. The water also was filtered through a filter of the mechanical type. Through some accident the filter was thrown out of service for eleven days, between October 20 and 31, 1903, and unfortunately, on account of the failure of the reservoir dam, the water was at that time being taken directly from the creek at the pump well, and had been since August 27. Only ten days after the filter was shut down, the epidemic broke out in all parts of the town. Between November 10 and December 19 there were 1270 cases and 56 deaths. In the subsequent investigation it developed that not only was the stream generally polluted by the sewage at various points above the intake, but that there had been several cases of typhoid fever on the watershed, some on a brook that enters the creek within one hundred feet of the filter plant. As at Caterham, the inference is patent that the introduction of some specific infection into the drinking water was the direct cause of the general epidemic. The occasional outbreaks of typhoid fever which occur in single families are not so easy to explain, particularly since the small number of persons affected does not usually call for a widespread interest on the part of those experienced in such epidemics. In the Twenty-seventh Annual Report of the New York State Department of Health, the following description of an outbreak in a small hamlet, where the cause seems to have been the use of a pond for a wash tub by some Italian laborers, thereby transmitting the disease germs from their clothes to the water afterwards used in a creamery, is given. The diagram, Fig. 37, shows that the creamery secured its water for the purpose of washing cans from a small pond by means of a gravity pipe line. The foreman of the creamery, who boarded at the residence marked _A_, first contracted typhoid fever. A week later an employee at the creamery also contracted the fever, the residence of the latter being marked _B_ on the diagram. About six weeks later the railroad station agent, living at the point marked _C_, contracted the fever, and two weeks later his wife was attacked with the same disease. The residences at _B_ and _C_ are only about three hundred feet apart, both families taking their water-supplies from a spring between the two, but nearer _B_. During the summer previous to this outbreak a gang of Italian laborers, engaged in double-tracking the Central New England Railroad, were housed in box cars standing on one track of the railroad. One of the members of the gang was reported to have been taken ill with a fever and was at once removed, it was supposed, to a hospital in New York.