had indeed suffered from enteric fever, and, laborers having laid bare the drains, these were all seen to communicate with the sewer above mentioned, this being further of such faulty construction as to be little better than an elongated cesspool.
In view of the danger of direct communication between a sewer and our dwellings, "What," you may fairly ask, "is the remedy?" I answer that the remedy is simply breaking the direct connection which has been referred to. In the case of a waste-pipe from a sink, the pipe should be brought through the wall into the outer air, and there be cut off, its contents flowing to a trapped drain-inlet outside the dwelling. (This point was explained by means of diagrams.) This principle of disconnection is, however, of much wider application than I have as yet indicated. All waste-pipes coming from lavatories, baths, water-closets, etc., as also the overflow-pipes from cisterns, and the rain-pipes, especially such as have their heads anywhere near windows, or beneath overhanging eaves, should, like the sink-pipes, have an air-space intervening between them and the drain-inlets into which they empty.
There is exceptional danger in the direct connection which is often maintained between houses and the sewers by means of the overflow-pipes of cisterns. These pipes are very generally provided with a "siphon-bend," but the water constituting the trapping is often absent. The ball-cock of the cistern is intentionally so contrived as to prevent overflow, and hence, when once evaporation of the water in the trap has taken place, sewer-air passes through it without let or hindrance.
Adapting the principle of disconnection to the house-drain itself, I would further urge that an air-break should always be contrived between the end of the drain and a trapped inlet leading to the public sewer; the more so as when this is effected a further safe-guard can be insured, namely, two ventilating apertures to the drain, and the maintenance of a constant current of air through its entire length.
(The conveying of infection by means of an "intermittent water-supply" was next described.)
I feel sure that many other methods by which water can act as a vehicle for conveying infection will occur to you. Milk, also, must be regarded as at least an equally important medium for the transmission of infection. I shall, however, ask your further consideration only of certain distributions of ice and cream as forming channels by which disease may be conveyed to households.
I believe that the first instance in which the consumption of ice was shown to have been followed by an outbreak of disease is that recorded in the "Seventh Annual Report of the State Board of Health of Massachusetts." The occurrence took place in one of the large hotels at Rye Beach, New Hampshire. At the beginning of the season
