In Diagram XI the typhoid death rates of Columbus, Ohio, and New Orleans are shown to exemplify conditions that have not been improved by chlorination. The endemic condition of typhoid in Columbus was brought to an abrupt conclusion by the installation and operation of the softening and filter plant in September, 1908, and no further reduction followed the introduction of chlorination in December, 1909.
In New Orleans the typhoid rate decreased on the inception of the new water works system in 1909 and again after the installation of the Carrollton filters in 1912. The product of the filtration plants has always been above suspicion but aftergrowths occasionally developed and the bacterial count then exceeded the United States Treasury standard. To overcome this difficulty, hypochlorite was used in 1915, but, as was anticipated, it had no effect on the typhoid rate. The high rate in New Orleans is largely due to outside cases received for hospital treatment and to other circumstances beyond the control of the water and sewerage department.
In all the examples previously cited, the evidence as to the effect of chlorination on typhoid mortality rates is circumstantial but, taken as a whole, it is fairly conclusive. In the examples to be considered next the evidence is more direct.
One of the most conclusive experiments as to the beneficial effect of chlorination is that reported by Young[8] of Chicago. The water supply of Chicago was obtained from Lake Michigan by means of intake pipes and pumped to various parts of the city. The distribution system was divided into four districts and, although there was a certain amount of mixing along the borders, the water supplied to each district was substantially separate. The rapid and progressive decline in the typhoid rate of Chicago (from 19 in 1900 to 10.8 in 1911) subsequent to the diversion of the city sewage from the lake, led to the assumption that water-borne typhoid had ceased to be of any moment. Early in 1912, however, permission was secured to chlorinate the supply of one district (No. 1) and the treatment was continued until December when the solutions commenced to freeze. Diagram XII shows the effect of the treatment on the autumnal increase in District No. 1 as compared with the other three districts. The autumnal increase was calculated from the excess of typhoid incidence for July to November inclusive, over that for February to June inclusive.