The provision of an adequate supply of pure water is the first essential of municipal sanitation. In New Haven the public supply is furnished by the New Haven Water Company, the city paying a lump sum for its service. The water supply is very generally distributed, except in the outlying districts of the thirteenth, fourteenth and fifteenth wards, and private wells are very rare. Our special investigator, Mr. D. O'Brien, found a few dug wells on Crescent and Fitch and adjoining streets in Ward 13. These should be eliminated as rapidly as possible.

In the central part of the city the various supplies are mixed in the pipes, and the same house may be supplied from one or another source, depending on varying draught in various sections.

The public water supply includes four principal systems which enter the city by the mains shown in Fig. 6, while a fifth supply from the Maltby Lakes serves a small part of the West Chapel Street section as well as the town of West Haven. The first three of the main supplies are drawn from Lakes Saltonstall, Whitney and Wintergreen, respectively, the fourth or Woodbridge supply from Lakes Dawson, Chamberlain and Glen. Up to 1902 all of the water was used without purification. In this year a sand filter was installed at Lake Whitney as a result of a serious [32] typhoid epidemic, which was definitely traced to the Woodbridge supply. More recently (in 1912) a chlorination plant has been installed at the Woodbridge reservoir. The Saltonstall, Wintergreen and Maltby supplies receive no purification except that which they gain by storage.

The Lake Whitney Watershed covers 37.7 square miles; that of Lake Saltonstall 20.2 square miles; Lake Wintergreen 1.1 square miles; the Maltby Lakes 13.3 square miles, and the Woodbridge supply 13.6. The population on the watersheds of Lakes Wintergreen, Maltby, and Saltonstall, the only supplies without purification in addition to storage, is 16,757.

Lake Whitney furnishes 12,000,000 gallons daily; Woodbridge supply, 9,000,000; Lake Saltonstall, 6,000,000; and Lake Wintergreen, 1,000,000--a total of 28,000,000, which means a consumption of 180 gallons per capita per day. This consumption is about three times what it should be, and in the interests of economy a system of metering should be introduced. Any excess over a consumption of 50-60 gallons per capita per day is always found to be due to waste through leaking fixtures and leaking street mains, and the installation of meters soon leads to the detection of such waste and its elimination without any limitation on legitimate use.

The Lake Whitney sand filters have now been in operation for 14 years and have yielded excellent results under the able superintendence of Mr. O'Connor. The system of purification of the Woodbridge supply by chlorination is also entirely satisfactory. The other three supplies, as pointed out above, are improved by storage and are protected by a very thorough system of sanitary patrol and by the ownership by the company of a considerable portion of the watershed. Such precautions are good, but they do not, and cannot, ensure a wholly safe supply. No amount of patrolling can prevent occasional pollution by picnickers, tramps, or by those using the roads near the reservoirs. The factor of safety offered by storage may at any time be nullified by sudden rushes of rain or melting snow or by currents induced by wind, such as caused the epidemic of typhoid at Auburn, N. Y., in 1908. New Haven cannot afford to run even a slight risk of repeating the disastrous experience of 1901.

Our laboratory examinations of the various supplies bear out the conclusion that tile three unpurified supplies are not at all times entirely safe from a sanitary standpoint.

During November and December, 1916, a series of samples was examined on 29 different days for the colon bacillus, the characteristic microbe of the intestine. As indicated in the table below, this organism was found but once in the disinfected Woodbridge water (in two out of five one c.c. portions examined), but twice in the filtered Whitney supply (once in one [33] and once in two out of five one c.c. portions examined) while it was isolated on four days from Saltonstall, on five days from the Maltby supply, and on seven days from the Wintergreen supply. The colon bacillus may occasionally be isolated in one or two out of five one c.c. portions without the result being significant, but the Wintergreen and Maltby results tabulated below indicate that these waters at times

BACTERIOLOGICAL EXAMINATIONS OF NEW HAVEN WATER

Number of Days on Which B. Coli Was Isolated

	Number of one c.c. portions showing positive results on a given day
	0	1	2	3	4	5
Saltonstall	25	4
Wintergreen	21	4	2		1
Woodbridge	27		1
Whitney	25	1	1
Maltby	23	3	1

show evidence of pollution in amounts which are not considered admissible by such standards as those laid down by the U. S. Public Health Service for drinking water served on interstate railroad trains. It is true that most of the colon bacilli in these waters do not come from sewage but from the washings of fertilized land, roadways and the like. Our studies of the typhoid death rate of New Haven do not indicate that the water supply has been recently in any way responsible for its causation. It is apossible future danger, not a present one, we have in mind, but a danger which an unhappy combination of circumstances might at any time make an actual one.

It is the clear consensus of opinion of competent sanitarians that sanitary patrol and storage alone cannot generally be relied upon to furnish a safe water supply in thickly settled regions like those of southern Connecticut, and that additional measures of purification should be provided. Fortunately in chlorination, as now applied to the Woodbridge supply, we have an effective safeguard at a cost so low as to be wholly negligible. We therefore urge:

Recommendation V. That steps be taken to secure at all times the complete safety of the New Haven water by providing for the treatment by chlorination of the Saltonstall, Wintergreen and Maltby supplies.