From the recent topographic maps of the U. S. Geological Survey, we find the drainage area, including the lake, is 131 square miles instead of the 146 that has been published.

Calling total capacity of Indian Lake 4.7 billion cubic feet, this would 4,700,000,000

supply for 100 days

544 cubic feet per second in addition

86,400 x 100

to the inflow meanwhile.

Calling the evaporation loss 5 inches in each of the months July and August, this, on an area of 3,500 acres as partly drawn, would be equivalent 3,500 x 43,560 x 5

to

12 x 30 x 86,400

= 24 cu. ft. sec., but the inflow from 131 sq. miles

at, say, 0.18 cu. ft. sec. per sq. mile in any but the most extreme drought would be 24 cu. ft. sec. and would keep the lake from lowering by evaporation.

The full contents of the reservoir, if 4.7 billion gallons, would be filled by a run-off equivalent to 15.7 inches in depth, from the watershed of 131 sq. miles.

The daily heights in the lake as published in the State Engineer's report have been plotted in a diagram on plate 5 and on the same sheet is given such information as we have been able to work out regarding the times and rates of storage draft in each of the recent years.

On the same sheet, the roughly approximate diagrams that we have used for computing the discharge of sluices and logways are presented for use until something better is available and a capacity curve of the lake for various gage heights is plotted on same sheet, prepared from such few statements of volume at given heights as could be found published.

The computations of outflow have been made in two ways, and the results agree in a general way in showing that a discharge of not far from 600 cu. ft. per second is commonly maintained for about three to four months in all, or from 90 to 130 days, and it is of particularlar interest to note that this storage is as commonly drawn upon in January and February as in mid

summer.

LUMBERMEN'S STORAGE.

It should be remembered in studying records of low water gagings of the Hudson, that there is still more or less control of sundry lakes by lumberman's dams for purpose of log driving, and although according to the topographic map the

4

areas that can possibly be flowed are small and the dams are probably low, the outflow when the dams are “tripped” may exercise an appreciable influence for a few days upon the quantity flowing in the river. Since these dams are in part owned by the owners of mills on the river below, these old reservoirs may perhaps be sometimes utilized outside the log driving season to assist in furnishing water for power. Piseco Lake, for example, is under some storage control. In the work of the coming season, all these matters of the storage control now possible upon the Adirondack streams should be investigated for purpose of completeness of the record. Plainly, they can exert no substantial effect in the monthly averages of flow past the gaging stations.

WATER POWER ESTIMATES.

The five main questions of data required for measuring the magnitude and value of a water power project are:

1. The quantity of water that flows in the river.

2. The extent to which uniform flow can be secured by means of storage reservoirs.

3. The height of fall available for power.

4. The cost per horse power of development. 5. The market for the power.

We will consider these questions in their order and present the data and proofs as briefly as will make the basis of our estimates plain; first in relation to the Sacandaga project and second in reference to the Genesee project.

I have taken much care to make certain about the quantity of water available in time of drought and of the extent to which this can be regulated by storage and to confirm this data from various independent sources.

In order to determine the flow of the Sacandaga, I have relied mainly upon the published records of the Hudson river gagings maintained day by day for twenty years at Mechanicville for the paper mills formerly of the Duncan Co., now owned by the West Virginia Pulp and Paper Co., under the able and painstaking supervision of Mr. E. P. Bloss, C. E., but have allowed for the greater rainfall and run-off of the Sacandaga.

As the first step in these estimates, it was assumed that since the drainage area of the Sacandaga above Conklingville is about 1,050 square miles, while that of the Hudson above Mechanicville is 4,500 square miles, the Sacandaga yield will be 188023.3 per cent. of the Mechanicville yield.

For comparison and confirmation, we have made use of twelve years' daily gaging of the Hudson at Ft. Edward where the flow is from 2,800 square miles, and have further confirmed the Mechanicville data by comparison with several years' record of daily measurements on two other adjacent tributaries of the Hudson, the Schroon, having 563 square miles, and the Hoosic, which has a drainage area of 579 square miles above its gaging station at Buskirk. For each of these four Hudson branches, we have determined the greatest continuous flow possible in various periods of draught and the corresponding volume of storage required, by means of "the mass curve method," which is, I believe, by far the most accurate method for computing storage requirements. A further confirmation of our estimates of river flow and reservoir depletion is found in a comparative study of the records of rainfall and run-off from the Croton with the rainfall upon the Sacandaga.

The rainfall records kept for years at various stations around the head waters of the Sacandaga and the Hudson show plainly that the Sacandaga drainage is certainly more prolific per square mile than the area above Mechanicville as a whole, and so finally an adjustment for this larger rainfall and run-off per square mile of the Sacandaga compared with the other tributaries has been made.

All of these sources of information, the Mechanicville gigings, the Fort Etwari gigng, the rainfall studies and the comparison of yield per square mie with the yield of the Croton and other drainage areas from which the tally fow has long been gaged, are found to enfrm one another remarkably We at will appear in the following pages, and as a result of the work of the past few months in studying the various records, we can now predict the dactarge amallable under various conditions from the proposed Sacandaga reservoir accurately enough for all practical purposes, and the present estimates of quantity of power avaliable rest on a much more full and reliable foundation then is ordinarily available when planning a great power development.

The sources of data and the results concerning this available Sacandaga fow will next be described, with only such fullness as is needed to make clear the reliability of the evidence and the methods of estimating.

GAGINGS OF HUDSON AT MECHANICVILLE.

The records which I have used for the monthly flow of the Hudson and its tributaries at Mechanicville, Ft. Edward and Warrensburg and Buskirk, are found in the annual reports of the State Engineer and in various water supply papers of the U. S. Geological Survey and in Rafter's Hydrology of New York and in advance copies of the records of the past year kindly furnished me by the U. S. Geological Survey, by the State Engineer and by Mr. E. P. Bloss, the Engineer in charge of the water power of the West Virginia Pulp and Paper Co. at Mechanicville.

From October 1, 1887, to November, 1897, the estimates of Hudson flow at Mechanicville were computed under the supervision of Mr. George W. Rafter from notes placed at his disposal by the mill owners. For later years, the records and estimates have been made under supervision of Mr. Bloss, Engineer of the West Virginia Pulp and Paper Co., the present owner of this power site and paper mill.

The daily gaging comprises an estimate of the flow passing over the crest of the stone dam from gage readings said to be taken twice each day and an estimate of the quantity passing through each of the numerous turbines of the adjacent paper mill, made up from a record of the number of hours run by each, the head acting on the turbine for that day and a statement of the discharge of a turbine of this size under this head, given by the manufacturer of the turbines. Most, if not all, of these turbines are fortunately of models and sizes which have been accurately tested in the Holyoke testing flume.

CHECK MEASUREMENTS ON TURBINE FLOW AT MECHANICVILLE.

As confirmation of accuracy of these turbine discharge tables for wheels long in use, it is stated that a test made at the mill by Mr. Bloss, Engineer of the company, of a twenty-nine inch Hercules wheel in use eight years, showed the actual discharge, when run at speed of greatest efficiency, to be substantially as given in the manufacturer's tables.

NOTE. See report New York State Engineer for 1905, page 577; see Rafter Hydrology of New York, 1905, page 365; see also reports of U. S. Geological Survey Water Supply Papers No. 202, page 19; No. 166, page 19; No. 125, page 23; No. 97, page 222; No. 82 page 102; No. 65, page 49; No. 35, page 24.

For further confirmation of the accuracy of the turbine discharge tables, several check gagings by current meter have been made at the Toll Bridge a short distance down stream from the paper mill at times when no water was passing over the dam, as follows:

On page 581, New York State Engineer's report for 1905, it is stated that October 10, 1905, the discharge by regular mill record was 5,181 cubic feet per second, while discharge by current meter gaging was 4,900 cubic feet per second. This comparison gives an excess by turbine record of about 5 per cent.

On October 20, 1900, another comparison was made, which is reported in U. S. Geological Survey Water Supply Paper No. 65 for 1901. The check measurement by current meter at Toll Bridge showed 1,871 cubic feet, but this measurement was not regarded as very reliable because of slack water. Meanwhile the discharge computed from the turbine records was 1,977 cubic feet, or about 5 per cent. greater.

It is desirable that further test measurements of this character be made in time of drought, for on certain occasions the minimum flow in time of drought given in the Mechanicville records is remarkably large and much larger than the estimates already quoted from the Census report on page 65. Moreover, on trying to analyze the sources of the large flow recorded at Mechanicville in the last of August and the first of September, 1907, in view of our gagings on the tributaries and our estimates of Indian lake discharge, I am unable to find where so large a flow all came from. Certain records made at the electric power stations at Spier Falls and Mechanicville also show lower flows in the drought of 1907 than are given by the standard records from the paper mill.

ACCURACY. OF MEASUREMENTS OVER MECHANICVILLE DAM.

The present conditions for accuracy in measuring the flow over the dam at Mechanicville are excellent. The dam is of masonry on a ledge foundation and has a rounded, level concrete crest, added in the summer of 1904, and new discharge tables for this crest have been prepared based upon experiments made at the Hydraulic Laboratory of Cornell University, (it is not stated if the model was precisely similar).

Mr. C. E. Parsons, Chief Engineer, Hudson River Electric Power Co., has stated to me that he has repeatedly looked into the matter of the gagings at Mechanicville and regards them as very accurate. He further states his belief that the crests of the river dams at Mechanicville and at Fort Edward are seldom obstructed by driftwood so as to interfere with accuracy of gaging, and that he has found the Mechanicville crest particularly free from obstruction of that kind.

Prior to the summer of 1904, the Mechanicville dam had a flat masonry crest about seven of eight feet in width, inclining downward upstream on a slope of about one in eight, as shown in the sketch in N. Y. Hydrology 1905, page 366.

It is stated in the published records referred to above, that for the years from 1899 to 1904 the discharge was computed by the Francis dam formula, This formula would apply accurately only at those times when the flash boards were removed, which presumably is the case only in time of flood, but it is stated that the same formula was used at times when flash boards