In other cases sewage gains access into leaky water-pipes. Formerly contamination was occasionally due to improper connection between the overflow pipe of the cistern and the soil-pipe, or to the water-closet being flushed by a pipe directly connected with a water-main (as in the Caius College outbreak at Cambridge), or connected with the drinking-water cistern (page 76). Milk may, by the admixture of water, become contaminated with enteric matter, and produce widespread epidemics. Where the water is very impure, the small amount used in washing cans may suffice to cause infection. _Cholera_ was first proved by Dr. Snow, in 1849, to be due to the specific contagium of cholera gaining an entrance into drinking water. This contagium is derived as in enteric fever from the intestinal evacuations, the urine, and the vomit of patients suffering from the same disease. [Illustration: FIG. 9.] The close connection of the spread of cholera with an impure water supply has been repeatedly shown in this country. The cholera epidemic of 1854 was very severe in the southern districts of London. At that period these districts were supplied with water by the Southwark and Vauxhall Company, deriving its water from the Thames at Battersea, and by the Lambeth Company, having its intake at Thames Ditton, where the water was purer. The two companies were acting in rivalry, so that in many streets their mains ran side by side; and houses in the same street similar in all other respects, received a different water supply. An investigation of the distribution of cholera in these districts gave the following results:— POPULATION CHOLERA CHOLERA DEATHS IN 1851. DEATHS IN PER 10,000 14 WEEKS. OF POPULATION. Houses supplied by Southwark Co. 266,516 4,093 153 „ „ Lambeth Co. 173,748 461 26 The facts, when examined in detail, brought out still more strikingly the exemption of the houses supplied by the Lambeth Company; the infection picking out in a given street the houses supplied by the Southwark Company. The great epidemic of cholera at Hamburg in 1892 proves the same point. Hamburg, Wandsbeck and Altona are three towns adjoining each other, and really forming one large community; but while Hamburg suffered terribly, the two other towns had no cases of cholera, except the few that were brought into them. In all respects except water-supply the conditions were alike; but Wandsbeck obtained filtered water from a lake, Altona obtained filtered water from the Elbe below the town, while Hamburg was supplied, previous to the epidemic, by unfiltered water from the Elbe just above the town. =Diphtheria= and scarlet fever have never been traced to polluted water.