Two points already mentioned are essential if well water is to be kept pure. One is to line the well with a water-tight masonry lining, and the other point is to have the cover of the well made with a thoroughly water-tight coating. This does not always give full protection, since in some cases polluting matter may pass through even ten feet of soil. This would be particularly true if the well was in a fissured or seamed rock, and very recently the writer found a well dug in a laminated granite, where a near-by sewer, leaking at the joints, contaminated the water of the well, although the well was cased with an iron casing twenty-five feet deep. The sewage escaped into a crack in the rock and followed the crack down vertically and horizontally into the well. Limestone is even more dangerous if any pollution exists in the vicinity. In cases where a well goes down to a horizontal layer of limestone and where a privy vault is dug to the same rock, it is found that pollution will follow the surface of the rock horizontally a long distance, and this condition of things always makes a well water suspicious. In sand or fine gravel, on the other hand, the danger of contamination is almost negligible; on Long Island, for example, the cesspools and well are both dug ten or fifteen feet deep and only fifty feet apart without any trace of contamination being detected. _Milk infection by typhoid._ Milk is responsible for perhaps 5 per cent of the cases of infection. Although the infection is always foreign to the milk itself,--that is, enters the milk only after the milk is drawn from the cow,--milk frequently becomes infected because infected water has been added to it or because the cans have been washed in infected water, or because some persons in contact with a typhoid patient have had their hands infected and then handled the milk or the milk utensils. There are a number of epidemics which have been clearly traced to milk polluted in one of these ways. In Somerville, Massachusetts, for example, in 1892, 32 cases occurred, 30 of which were on the route of a single milkman. It was found that the milkman had two sons, one of whom had typhoid fever just before the outbreak. This son washed the milk cans and mixed the milk in a milk house in the city, and the inference was that in some way this man infected the milk, probably in one of the mixing cans.