Omitting the sinking fund provisions and thus performing precisely the service proposed by the Ramapo Company with works built by the City itself, it would probably cost the City only about $20 per million gallons for this Esopus and adjacent water for which $70 per million was to be charged.

Water supply developed on the large scale, with present low rates of interest, from any favorable source is wonderfully cheap.

(M) While the Legislature of the State of New York cannot give to your City the right to divert the water of the Upper Housatonic, and perhaps not the full rights to divert the Ten Mile river, the ease with which similar questions have been settled in the case of the diversion of a portion of waters of the Nashua river from New Hampshire for an addition to the supply of Boston, and the promptness with which rights were secured and settlement made for the diversion of a portion of the waters of the Blackstone river from many Rhode Island mill sites, for the public water supply of Worcester, Mass., gives confidence that these rights can be secured on the basis of ordinary business procedure, by paying generously, and, on the whole, come out nearly as well as in seizure under eminent domain. It is also believed that constitutional authority can be found for this, and that in a neighborly spirit the State of Connecticut will grant any power needed for New York to complete and perfect its title without hindrance from any few adverse interests, so soon as full and generous payments for all injuries to the property or prospects of its own citizens are made secure. Possibly, also, terms could be arranged, fair and advantageous to both States, by which a transfer of the State boundary line could be made, such that the proposed Housatonic Reservoir would lie within New York, much as Rhode Island and Massachusetts readjusted their boundaries near Providence and Fall river. Eminent legal authority well versed in the laws and decisions regarding water rights assures me that the legal obstacles to Housatonic and Ten Mile water are such as can be readily overcome, providing the Connecticut Legislature will act in the neighborly spirit which we have a right to expect.

(N) Should negotiations for the Ten Mile river fail, there is a watershed lying north of the Croton, which is free from all interstate questions, and which can apparently be made to deliver 150 million gallons of water per day into the New Croton Reservoir, but does not have the advantage of elevation possessed by the Ten Mile and certain other sources.

A high-pressure supply from the watersheds adjoining the Croton on the north and northwest appears from my reconnaissance to be impracticable.

(0) An ample supply of excellent water up to even 1,000 million gallons per day, or so much of this as wanted, can be brought from the Adirondacks by gravity and delivered at the same elevation of 300 feet above sea level, at a cost, including all interest, depreciation, maintenance and sinking fund to pay off the whole cost in 40 years, for from $50 per million gallons to $25 per million gallons, according as a supply is taken of 200 million gallons per day, or of 800 million gallons per day, but while cost per million gallons is low, the total bond issue that would be required for this would be so extremely large as to make this project probably impracticable.

(P) By taking water from the Hudson river at Poughkeepsie and filtering it, and by constructing storage reservoirs in the Adirondacks, to let down water whenever, in extreme drought, there was any sign of brackishness in the Hudson water at Poughkeepsie, and then by filtering this Hudson water, just as Poughkeepsie has done for its own supply from the Hudson for 27 years, or as Albany is doing to-day, New York can obtain any desired quantity, even up to 800 million gallons per day, at a cost, including interest, maintenance, depreciation and sinking fund to pay off all costs of construction in 40 years, for water delivered at city limits at 300 feet above sea level, less than $50 per million gallons.

For filtered Hudson water delivered at city limits at same level as the Croton at Jerome Park, the cost will be about $7 per million gallons less than for water delivered at elevation 300.

(2) If the Ten Mile and Upper Housatonic water can be secured, these sources will supply Brooklyn much more cheaply than water from Suffolk County.

If these sources are not obtainable, and if the New York Legislature will remove the restriction forbidding Brooklyn taking water from Suffolk County, then the surface waters and ground water from the saturated gravels above the blue clay in the region east of Massapequa can probably furnish from 75* to 125 million gallons per day of excellent water, at a cost a little less than $40 per million gallons.

*(A geological study of this watershed from Massapequa to Brookhaven raises serious question of it being so extensive as assumed in the Brooklyn investigation of 1896. On the other hand means can probably be devised that will probably yield a larger proportion of the total annual rainfall.)

(R) A deep, artesian well supply from the deep gravels of Long Island lying beneath the blue clay, of any sufficient quantity to be of material value in solving the problem of Brooklyn's future supply, appears from a geological study, set forth in this report, Appendix No. 16, to be utterly hopeless.

(S) For Staten Island. A simple inspection of the ground or of the topographic map quickly shows that no adequate supply of surface water can be obtained on the island. A geological study shows the obtaining of any adequate future supply from subterranean strata to be utterly hopeless. The best means of supply appears to be by submerged pipes from Brooklyn, carried across a little north from the Narrows.

(T) West of the Hudson, the sources south of the Esopus are not very promising, and beyond all question are less advantageous than the Upper Housatonic.

The RAMAPO RIVER, which (with the Mahwah) would for $15,000,000, or somewhat more, furnish about 60 to 75 million gallons per day, probably cannot be diverted from flowing into New Jersey, because of its availability for the future supply of certain New Jersey cities. The Brooklyn report of 1896 appears to have underestimated the cost and overestimated the quantity of water available from this source.

The POPOLOPEN appears to be favorably located for furnishing about 20 million gallons per day of good water, at such elevation that it would flow by gravity, without pumping, to a level at the city limits of 300 feet above the sea.

The MOODNA apparently can be developed to furnish about 100 million gallons per day, at 300 feet elevation, but there is some question whether its water might not require filtration, or possibly prove unsatisfactory, even with filtration, because of peat and black muck in the bottom of some of the important reservoir sites.

The WALLKILL is attractive from its location on the map and because of lying at an elevation sufficient to deliver water by gravity at 300 feet above sea level at the city limits of New York. From its 417 square miles of watershed above New Hampton, by converting the Drowned Lands into an enormous storage basin of about 50 square miles area and about 25 feet depth, this watershed apparently could be developed so as to yield a safe flow of about 315 to 365* million gallons per day, if the estimate

* From personally observing the surprisingly small size of the stream, in comparison with that from other watersheds with which I am familiar, and from also briefly studying the flood marks, and from a hasty view of the geological conditions, I am led to question if the yield per square mile at the outlet of this basin would not be uncommonly small, because of percolation through deep strata, but I did not have time to investigate this matter of flow thoroughly.

A brief study of the geology of this dam site, which, from brief inspection taken in connection with the report of the State Geologist on Orange County, appears to be in a secondary valley eroded in a deep deposit of drift, and where formerly a dam of glacial drift converted the site of the Drowned Lands into a Lake, lends additional color to this suspicion of a small collectible yield, and without a year or more of gaugings of the actual stream flow, at some favorable spot, as near the railroad bridge at New Hamburg, I should hesitate to believe that nearly so great a yeild can be had from the Wallkill as that ordinarily obtainable from a watershed of this size with this proportion of storage. A thorough geological study directed toward this particular question is desirable, also a series of gaugings.

is made according to ordinary rules derived from experience on watersheds where there is an impervious substratum and an impervious dam at the outlet.

Six thousand acres, or 10 square miles, of the proposed reservoir bottom appears to be filled with swamp muck and peat deposits from 5 feet to 50 feet in depth, about 90 per cent. of which is organic matter, which, with shallow flowage under the summer sun would probably grow algæ as luxuriantly as they now grow onions, and there are doubts if its water could be satisfactorily purified and made potable by filtration.

There are important interstate flowage questions connected with flowing 10 or more square miles of land in New Jersey by this reservoir that would probably be almost as difficult to settle as any interstate question regarding the Housatonic.

Such examination as I have had opportunity to make on the Wallkill gives very small encouragement of this ever becoming an available source for the water supply of New York City.

The SHAWANGUNK RIVER would probably afford 35 million gallons per day of good water, at an elevation sufficient to flow by gravity to elevation 300 in New York, but at a high cost for storage and aqueduct, if the Wallkill storage basin proves impracticable.

Future Filtration.

(U) I venture to express the belief that by 20 years hence the public will have become educated to demand a higher standard of purity in public water supplies and that all future work should be laid out with a view to filtration 10 or 20 years hence of all water entering the distribution system. The conduits should run past land suitable for filter beds and head or fall be reserved suitable for working the filters without any expense for pumping. Personally, I believe that with complete meters and proper waste restriction filters could be properly advised at once for the Croton supply, and I find as detailed in the report that to filter the present supply with all its waste would cost only about 35 cents per capita per year.

Results of Certain Other Investigations Made.

Turning back now from consideration of possible new sources to the present water supply it may be added that:

A careful recomputation has been made of the actual run off of the Croton river day by day for the past 32 years, applying more accurate data to computing the flow from the old records of depth; whereby it plainly.

appears that the computations heretofore relied upon gave too large a flow by about 10 per cent.

An investigation by experiments on full-size models of the Croton damı crests-made in the new hydraulic laboratory of Cornell University-proves that the formula used heretofore for computing the water wasted exaggerated this flow about 9 per cent.

A gauging of the flow in the New Aqueduct made at my request by the same assistant engineer, with the same current meter by which the New Aqueduct was originally gauged, proves that this has suffered an impairment of about 40 million gallons per day in carrying capacity, due probably to it not being kept clean.

The corrected results of the Croton river gaugings covering this period of 32 years, and of the Croton Watershed rainfall records have been condensed into tables and diagrams of great value for estimating the safe yield of such other watersheds as may be selected for new sources of supply.

The present rate of consumption of Croton water, in gallons per day, has been accurately measured with the result of showing that the estimates. of consumption and of growth in consumption of water made by the Department of Water Supply at the time the Ramapo contract was urged were about 40 million gallons per day too large.

The rate of consumption of water in the Boroughs of Manhattan and Brooklyn at the various hours of the day and night has been measured with a degree of precision not heretofore attempted, thereby learning that in the quietest hour of the night or at 3 A. M. the draft in Manhattan is at the rate of 94 gallons per capita per day, and in Brooklyn, at the rate of 57 gallons per capita per day, thereby obtaining a better measure of the waste of water than has been heretofore available.

Experiments on the night consumption in the thoroughly metered cities of Fall River and Woonsocket have been made for comparison.

(W) A very full geological study of the deep subterranean strata of Long Island and of Staten Island has been made with a view to learning of the possibility of a supply from deep wells and of the best means of obtaining water from the saturated gravel. This, as already stated, shows that a deep artesian well supply, adequate for any general public supply is practically hopeless.

While the above facts and many others have been collected by the reconnaissances made by me in person and by others under my supervision, it is to be understood that the shortness of the time and the broad field to be