the data available for estimating the Croton flow prior to 1900 have now been pretty nearly exhausted, and that this part of the problem can rest until the record of some more severe cycle of years of small rainfall appears in the future. The safe yield of the Bronx and Byram was also investigated, so far as the data permitted, with the result, as I believe, of showing that the estimate heretofore common (20 million gallons per day) is 33 per cent., or 5 million gallons per day, too high. This is set forth in Appendix No. 2. Consumption:

The question of daily rate of total consumption of water in Manhattan. and Bronx has also been gone into with a fair degree of thoroughness, and it is a relief to find that this is not so large and has not been increasing so fast as was supposed, but it is plainly increasing 15 million gallons per day per year.

Use and Waste:

The proportion of the water wasted in Manhattan and Brooklyn was investigated in a broad, general way by certain measurements of the comparative consumption by day and by night. This investigation of waste would have been extended to the other boroughs had time permitted.

The relative proportion of use and waste for Manhattan and Brooklyn is brought out so clearly by the curves of Diagrams No. 5 and No. 6 as to leave very little doubt that more than half and probably two-thirds of the total supply is wasted through leakage.

Costs:

On some other questions, as, for example, the cost of dams and aqueducts for an additional water supply, there has been no opportunity to go into the matter with great thoroughness, and my figures for the Ten Mile and Housatonic, although very carefully studied and based on careful reconnaissance in the field, are but approximations. Surveys more in detail may show here or there ten or perhaps twenty per cent. difference in the quantity of excavation or of masonry; perhaps in some places more, but more often less, for I have tried to be particularly liberal in anticipating the costs of the structures proposed, and of the damages liable to be claimed, for developing the Housatonic and Ten Mile supplies and have put in a very generous price for margin for contingencies. I have figured costs throughout on the basis of well managed contract work at such labor prices as to-day prevail on the extensive constructions of the Massachusetts Metropolitan Water Board.

I trust that the margin of uncertainty in all these estimates due to lack of complete surveys, test-pits and borings, will be found so clearly stated that

none of my figures can seriously mislead or be misunderstood. For this purpose, also, and as an aid to a critical understanding by the non-expert, I have written into the estimates more notes than common of comparison with cost and yield of similar works, and made fuller note of the facts which I believe must lead others to substantially the same conclusions as stated.

New Sources:

A hasty reconnaissance was also made of all the possible sources of additional supply upon the map of a size worth considering for New York, which appeared to have any promise of future use. It is a matter of much regret to me that I have not had time to continue the explorations of the low level line northwest of the Croton Watershed by a line of accurate instrumental survey; and of further regret that I have not had time to go into details and make closer studies of the possibilities along the Moodna, Wallkill, Neversink and the Delaware (my investigation of the Delaware tributaries was very incomplete); but I have gone far enough on those studies to convince me that they are much less hopeful sources than the Ten Mile and Housatonic.

Memoranda and Explanations:

In the appendices will be found a part of the data that I have collected bearing on these various questions, some of which I have had to dig out laboriously from old reports, or by questioning old employees of the Water Department and by much personal inspection and work of assistants, and which it appears very important to leave recorded in such a way that it shall be conveniently accessible to others who may have occasion to investigate these matters, now or in the future. In these appendices I have also written up the basis of my estimates with more than ordinary fullness that they may be open to criticism and may be readily understood, and since it has appeared desirable that all the facts collected should be quickly available for the use of the engineers of the committee recently appointed by the Merchants' Association to investigate New York's water supply and open to their criticism and extension, advance sheets of the results of the computations upon yield and consumption and such of the other diagrams as appeared of interest, have been furnished them.

Coming now to the results of this work, my answers to your several questions are as follows:

FIRST QUESTION.-CAPACITY AND ADEQUACY OF THE PRESENT WATERSOURCES.

A careful recomputation from the daily measurements made by the aqueduct keepers during the past thirty-two years shows, as already stated, that although the consumption will very soon be uncomfortably close to the safe limit, with prudent management the Croton Watershed, supplemented by the Bronx and Byram, and with a moderate and practicable restriction of waste, can furnish a safe and ample supply for the boroughs of Manhattan and The Bronx for four or five years to come; thus giving a short but adequate time in which to build works for an additional supply.

This is shown in the most convincing manner by a study of Diagram No. 3-B in connection with Diagram No. 2, and will be found explained in detail in Appendix No. 1.

These recomputations and diagrams show that under special care, the Croton Watershed can be prudently relied on, after the New Croton Dam is finished in 1902, to furnish (taking some chances on impairment of quality of water), temporarily, say

The greatest safe yield that can properly be relied on permanently after the new dam is completed and its reservoir filled, is

Prior to the filling of the reservoir above the New Croton Dam, taking the storage reservoir capacity as we find it to-day, and assuming the depletion of the several basins regulated with utmost skill under a repetition of the most severe drought of the past thirty-two years, the safe yield would be....

By means of adding flashboards on all dams to the greatest practicable height, temporarily, until the New Croton Dam is completed, and further increasing the storage by lease of ponds not owned, additional storage can be gained probably sufficient, with the rainfall the same as in 1880 and 1881, to make the safe yield prior to 1902....

300 million gallons per day.

275 million gallons per day.

232 million gallons per day.

250 million gallons per day.

The actual average consumption of Croton water in the year 1899 averaged 226 million gallons per day.

The consumption has been increasing 15 million gallons per day each year for four years past, and four or five years hence will be greater than the Croton can be relied upon to furnish in a series of dry years.

The actual daily consumption as measured by the delivery of the New Croton Aqueduct to New York is found to be less than heretofore supposed by 38 million gallons per day, because of an impairment of the carrying capacity of the aqueduct, probably due to its not having been cleaned for nine years. The velocity of flow in an aqueduct is reduced when organic growths collect on its interior, just as the speed of a ship is reduced when its bottom is foul.

FORMER ESTIMATES OF CROTON YIELD 10 PER CENT. too Large.

The results of the recomputation of the yield of the Croton, day by day and month by month, for thirty-two years past, show the flow to have been less than the estimates previously published, by about 10 per cent. on the average, or 38 million gallons per day.

Thus, curiously it happens that the over-estimate of the Croton yield averages the same as the recent over-estimate of the consumption.

The cause of the excess of the earlier estimates of Croton river yield over the present estimates is fully discussed in Appendix No. 1, but may be briefly summarized here as follows:

The summary of the yield of the watershed is made up of four independent measurements:

1st. The quantity wasting over the Old Croton Dam.

2d. The quantity drawn through the Old Aqueduct.

3d. The quantity drawn through the New Aqueduct.

4th. The quantity taken out of the storage reservoirs to augment the flow, or, in turn, taken from the flow to refill the storage reservoirs.

The lack of precision in the old estimates came from:

Ist. A less accurate formula for computing waste over the dam than that which we now possess, and the use of only a single observation each day upon depth over dam-while there were two actually observed, both of which are now used.

2d. The use of assumed coefficients, which were too large, instead of actual gaugings for determining the flow in the Old Aqueduct.