tion; second, mountain water from almost uninhabited and uncultivated territory, and third, filtered water.

There will be no hesitancy in accepting spring and mountain waters, provided they are not polluted. A few words should, however, be said regarding filtration, as the art of properly filtering water is of recent origin.

At present there is no longer a mystery regarding this process of purification. The conditions which allow water to free itself not only of turbidity but of microscopic bacteria and pathogenic germs suspended therein, are now fairly well known. Artificial filtration is in fact but an imitation of nature's process, operating under the most favorable conditions of converting surface water into spring water.

Highly polluted water can, by this method, be rendered free from the dangerous poisons and organisms. The reduction of typhoid fever and diarrhoeal diseases with the introduction of filtered water has consequently occurred in a large number of cities of Europe and America, and the necessary conditions are now so well known that the result can in most cases be safely foretold.

The oldest filtration plant in this neighborhood is located at Poughkeepsie, and since the works have been properly operated the sanitary effect has been satisfactory. The latest and the largest filter plant in the United States is at Albany. Since its recent introduction the typhoid fever death rate has been lowered very materially. The cities of Philadelphia, Pittsburgh, Cincinnati and Louisville are now engaged in preparing for works to filter their water supply.

The expense of filtration, including interest on the cost of the necessary works, ranges in the projects here presented from $4.23 to $6.91 per million gallons. The great sanitary benefit is therefore not expensively secured.

Quantity of Water to be Provided.

We have already discussed the question of quantity, so far as it relates to the water heretofore consumed, including the question of waste. Whether or not it will be practicable to reduce this waste as much as it should be reduced concerns us at

this place only so far as it will prolong the period at the end of which an increased supply must be available.

The indications are strong that an additional supply will be needed quite as soon as it can be provided, even if preliminaries are begun at once. We believe, therefore, and earnestly recommend that no time should be lost in starting them.

We did not fix a per capita rate of water consumption nor a time limit to be used as a basis for investigating an additional supply, because it is uncertain to what extent an average per capita rate of reduction of waste can be effected, and therefore how long a given supply will last; and it is otherwise unnecessary for our purpose to consider these items.

We have taken the broad view that whatever new source is selected, it should more than double the present supply and we, therefore, decided to investigate a daily addition of 400,000,000 to 500,000,000 gallons.

To decide upon a smaller quantity, irrespective of the fact as to whether or not it could be readily increased thereafter, would be, within a few years, to open anew the question of a future supply, with possibly increased legal difficulties and possibly demanding the acquisition of water rights and the purchase of property for aqueducts at greatly advanced rates.

It is, however, in no case necessary at once to provide works to supply the entire quantity. It is practicable in all of the projects to provide first for about 250,000,000 gallons and later to provide for the remaining quantity. The filter plants for the entire estimated quantities also do not require immediate installation, but can be gradually extended as the consumption increases. Nor is it necessary to make a distinction at this time between the amounts of water later required in each of the several boroughs. The distributing pipes are laid as the new districts develop and their location and sizes can then be adjusted to the timely needs.

Pressure to be Provided.

In addition to the qualifications given above regarding the water to be furnished, we have further to determine the elevation at which it should be delivered. This elevation depends partly upon the service in the city and partly upon the source.

Plate I shows the areas that can be supplied from the present reservoirs and the areas requiring pumping. In the several projects considered those which bring the water into the city by gravity can, as well as not, deliver it for all the territory at a high elevation, which we have fixed at 310 feet above tide.

The project requiring pumping from the Hudson River should deliver the water at no greater height than necessary, and it was decided to deliver a portion of the Hudson River supply at 131.5 feet and another at 260 feet above tide. A small portion will require a further lift into the highest districts of a limited area, to nearly the same extent as required at present.

We are of the opinion that it would be unwise to materially increase the present pressure in the lower part of Manhattan, where the high buildings are increasing in number. It is impracticable to deliver water to the highest floors of these by gravity. Therefor some of it must in any event be pumped.

A material increase of pressure would not only require a remodeling of much of the plumbing, but would cause a large increase in the unavoidable waste of water and a correspondingly large increase in its consumption. After the new aqueduct was brought into service with a better pressure the consumption at once increased about 30.8 per cent., which increase was almost wholly due to the greater pressure.

On the other hand, the only advantage from an increased pressure is the better fire service. Yet, our examinations point to the fact that a still greater advantage can be obtained at less cost by a separate salt water fire service in the important districts.

We have, therefore, assumed that of the water brought in by gravity at an elevation of 310 feet, as much of it as may be required by the low service would flow into the present reservoirs, and that the high-service districts would be supplied as far as practicable directly from the aqueducts.

AVAILABLE SOURCES FOR AN ADDITIONAL WATER SUPPLY.

Ground and Filtered Water from Parts of Nassau and

Suffolk Counties.

As has before been implied, we are of the opinion that on Long Island the surface waters should be filtered or the supply confined to ground-water. From Mr. Ward's tables we gather

that the yield of ground-water per square mile last year was 369,581 gallons per day, and that the total yield, including surface-water, was 745,983 gallons per square mile per day.

Including both the ground-water and filtered surface-water, the watersheds of Long Island available for a future supply are probably insufficient to yield more than about 100,000,000 gallons per day. This source could, therefore, only be considered as a partial supply for the City of New York and confined to the boroughs of Brooklyn and Queens, which it might serve for a long time.

The question then presents itself as to the relative economy of the Long Island source and a source from the main land. The decision depends on the particular one selected. If either the Housatonic or Wallkill were obtained, such water would be less expensive than the Long Island water. But, if the water is pumped from the Hudson River above Poughkeepsie, and also filtered, it would be cheaper to pump the near sources of Long Island ground-water. In the latter case, legal difficulties stand in the way, which prevent the use of Suffolk county water. And without this, the quantity available in Nassau county will not be sufficient very long.

For the present and until a much larger supply is required by the Borough of Brooklyn, this can be supplied with greater economy by extending the present system. In making the extensions, so far as they pertain to surface-waters, although they are apparently not greatly polluted, filter plants for purification should, in our opinion, be made an integral part thereof.

In extending the ground-water supplies, it would be well to gather it along a line further from the shore than it is gathered now. Under present conditions its removal is noticeable to agricultural interests. When ground-water is removed from a plane more than 10 or 12 feet below the surface of the ground, these interests could not be seriously affected thereby. If it is not obtained too near the surface, where it may receive surface pollution, nor is taken from too great a depth where it may have a mineral pollution, or be affected by the percolating ocean water, the ground-water on Long Island, filtering naturally as it does through deep beds of sand and gravel, and being

thus transformed into spring-water, has a high degree of purity and should be preferred whenever it is economical.

In view of the uncertainties existing as to the source to be selected for a new general supply and as to the time when the Borough of Brooklyn will be served by it, all of which governs the proper special treatment of the provisional extensions of the local supply, we have not made any further detailed examinations nor estimates of cost than what was necessary for recommendations to satisfy immediate needs.

Mr. De Varona, engineer of the water supply, who kindly furnished most of the information regarding the Brooklyn works, made an elaborate report in 1896 for increasing the supply 100,000,000 gallons per day. His estimate of cost for extending the ground-water supplies is $30 to $35 per million gallons delivered in the reservoirs, excluding payments to sinking fund.

Ten Mile and Housatonic Rivers.

Immediately after entering upon our duties, and knowing that the Housatonic watershed had been heretofore considered as the probable source from which New York would be supplied, after the Croton watershed had become exhausted, we inquired. of our Legal Committee regarding the availability of this watershed, as it was located chiefly in the State of Connecticut, and as all of the water finally passed into that State.

The report of that Committee (q. v.) advised against its availability. We therefore devoted no further time to an investigation of this project, and a further study would, in any event, have been unnecessary, as Mr. Freeman in his recent report to the City Comptroller recommends it for adoption and describes it in detail.

From an engineering point of view, and in the matter of cost, it is certainly a very attractive source. We will, therefore, at this place, record at least its salient features and its cost substantially as it was given to us by Mr. Freeman, for the purpose of comparison with the other projects.

A dam built across the Housatonic River about a mile below Merwinsville, Conn., would control the flow from a watershed of about 1,000 square miles. The reservoir thus formed would have a water surface of about 33 square miles and would