they were obliged to take under foreclosure, and assessed themselves $40,000, of which $32,000 was used in the construction of 14 miles of new canal at the heading to improve the original faulty location. * It will be observed that while the water taken out of the Gila by the Florence canal in 1896 was a large quantity, it was really only 11 per cent of the total flow of the river during that period, which in 1896 was yielding an abundant supply. * About the time Mr. Davis's report was made (1896) the Indians began to complain of shortage of water caused by the inroads made upon their supply by the Florence canal, and the latter in turn began to experience a diminution of its normal supply by the diversions made by irrigation canals on the upper Gila River, above San Carlos, in Graham County. An extensive development has been made in that region, and several prosperous towns have been built on the upper Gila as the result of the irrigation of so large a territory, but this has been largely at the expense of the farmers of the Florence region, who, though deprived of their water and impoverished thereby, have had no recourse. * It is now far cheaper to build a new one than to attempt to clean out the old canal and restore its original capacity. * After the appearance of these reports (W. S. P. 33 and S. Doc. 152, 56th Cong., 1st sess.), the people of the Gila Valley naturally expected the Government to build the San Carlos Dam in the near future. And even after the passage of the reclamation act they had every reason to hope and expect that the San Carlos Dam would be chosen as the object of the first development work in Arizona by reason of the double purpose which its construction would serve, viz, the provision of water for the Indians and the reclamation of a large area of Government land. It was a keen disappointment, therefore, when the Roosevelt site was given preference, and the last hope of the valley for irrigation development under Government auspices apparently vanished, at least within any reasonably definite time. As time had gone on and the Arizona apportionment of public land funds available for construction has been exhausted and exceeded on the Roosevelt project, the construction of the San Carlos Dam and the valley canal system by the Reclamation Service has become more and more remote. In fact, the Government officials appear latterly to have sought a pretext for withdrawing their support from San Carlos and have even made a report favorable to the surrender of the canyon to the Southern Pacific to build a road through it on a low level, which would make the dam impossible to build. The present status of affairs on the Gila is that the people of the valley are now seeking from the Interior Department of the Government the right to take the lands of the Apache Reservation at San Carlos for reservoir purposes, and the right to build a dam at the head of the Box Canyon high enough to store all the flood waters of the river, or as much of them as it may be feasible to impound. This question has been pending for several years, but it is likely to be decided in the near future, and I believe favorably to the irrigators. The landowners who have organized the Pinal Mutual Irrigation Co. are fully in accord with the movement to organize an irrigation district and build the San Carlos Dam, and will participate in the cost of it; but while waiting for it have decided to attempt the irrigation of at least 12,000 and if possible 25,000 acres by building a new canal from the river, which, by future enlargement and extension can be made to serve for the entire valley irrigable under the larger San Carlos storage project. They are unable to secure water from the old Florence Canal, as the receiver has surrendered his charge and given up the struggle to keep even with the expenses of maintenance, and therefore in self-defense are compelled to build independently. OUTLINE OF THE PROJECT. The project of the Pinal Mutual Irrigation Co. as outlined by their chief engineer, Mr. Alfred C. Sieboth, consists primarily in building a low diversion dam across the river at a point where granite bedrock outcrops on both sides of the channel, at a location about 4,000 feet upstream from the head gate of the old Florence Canal. This dam is intended to be so substantially built as to withstand the floods of the Gila, and high enough to answer the purpose of a permanent diversion weir for the larger project. From this dam site a canal has been surveyed on a uniform grade of 2.64 feet per mile, or 5 feet in 10,000, for a distance of 18.38 miles, following practically parallel with the old Florence Canal and a little way above it. 55. Mr. Schuyler proposed a dam of the Indian type with a lift of 4 feet and with sand sluice close to the head gate. The head gates recommended were of the radial type. Estimated cost, including headworks and sluice, $78,303. * As before stated it (the canal) will have a grade of 2.64 feet per mile, and carry a depth of 3.5 feet of water. With a bottom width of 16 feet and side slopes of 1 on 1 the grade will give it a mean velocity of about 2.7 feet per second, and a discharge capacity of 180 cusecs. The new canal should keep itself scoured clean and clear of sediment, because it will have a higher mean velocity than the Florence Canal ever had, even when carrying 50 per cent more water. The canal was to be 20 miles long; estimated cost, $113,270. 56. Mr. Schuyler figured on a duty of 5 acre-feet per acre at the heading. * Whenever it may be possible (as it doubtless will be in many years of full flow) to keep the new canal filled to its extreme capacity for 300 days of the year, it would be able to divert a total of 108,000 acre-feet per annum, and this would suffice for 21,600 acres net actually irrigated, or 29,000 acres gross supplied with water, the difference of 25 per cent representing the area which in almost every irrigated region is that occupied by roads, buildings, corrals, fallow land, dry pasturage, and unused corners here and there. While such flush seasons may often occur, the years when the river is low or nearly dry during April, May, and June are of such frequent occurrence as to place a serious limitation on irrigation operations and practically compel the farmers under the system to confine their crops to grain alone rather than to general farming, including the growing of alfalfa, which a constant supply of water would permit. Wheat and barley if planted during the late fall or early winter are not usually affected by this shortage if they can be irrigated before it sets in. For example, Mr. Sieboth states that as the sum of several years' experience in farming under the Florence Canal, he thinks that grain crops can be fully matured on four irrigations of 0.4 feet each; the first application to be in November, the second in February, the third in March, and the fourth the latter part of April. This would be equivalent to 19 inches of rainfall, and that amount in the San Joaquin Valley would produce bumper crops. The estimate therefore seems reasonable. If we assume that the ditches are so well cared for that the loss by seepage and evaporation during these four months does not exceed an amount that would cover the land 5 inches deep, the total volume of water required would be 24,000 acre-feet for 12,000 acres, requiring an average flow of 100 cusecs for 120 days. The record of discharge of the Gila, given in Table II, page 9, indicates that in 7 years out of 10 much more than this quantity was flowing and could be depended upon. In such years, therefore, the returns to be anticipated from grain crops on 12,000 acres alone would be sufficient to justify the proposed investment. 57. Mr. Schuyler devotes a considerable part of his report to a consideration of the probability of developing an auxiliary water supply by pumping. He figured on 10 pumping stations located along the south bank of the Gila at intervals of 500 feet, beginning at the Florence Bridge. Each would consist of two 15-inch wells with centrifugal pump run by electric motors. The estimated cost was $130,000. Cost of operation was taken at about $22,000 a year, including interest and depreciation. The delivery would be 10,000 acre-feet in 100 days, and the cost per acre-foot $2.20. * I therefore feel that this means of securing an auxiliary supply should be thoroughly investigated by boring one or more wells and making such tests as will determine their probable yield. SUMMARY OF CONCLUSIONS. In the foregoing I have endeavored to present the principal facts bearing upon the feasibility of the project, and I feel that I am justified in the following summarized statement of general conclusions: * 3. The lands of the district outlined on the accompanying map cover an area of about 54,600 acres, of which not less than 12,000 nor more than 25,000 acres are to be taken into the project. 4. The lands in the entire district are smooth, gently sloping, and composed of soil of great depth, generally a mellow sandy loam, well adapted to irrigation. In quality these lands can not be excelled. They are capable of raising five or six crops per annum under irrigation. 5. The value of the lands with a sufficient water supply to enable them to be irrigated throughout the year is conservatively estimated at $200 to $250 per acre. With a water supply for one crop each year they are worth $75 to $100 per acre. 6. The water supply available without storage is ample for safely maturing one crop annually in at least 7 years our of 10. 7. By boring wells and installing pumps it may be possible to develop 50 cusecs for auxiliary supply in dry seasons, at a capital cost of $130,000, and a total annual cost for pumping 10,000 acre-feet of $22,000, assuming that the United States Government will furnish electric power at 0.5 cent per kilowatt hour. 8. It is feasible to build a safe diverting dam of the type used in India and Egypt for $62,540 (net). 9. The canal projected and surveyed will carry 180 cubic feet per second on the grade and dimensions adopted. 10. The canal as projected can be built at a cost not to exceed $118,820. 11. A storage reservoir on the upper Gila or the San Francisco River, in New Mexico, may possibly be a cheaper means of securing an auxiliary supply than wells and pumps, and should be investigated. 12. A thorough test of the well-water supply should be made under expert direction and advice. 13. By the expenditure of $181,360 a system of irrigation can be provided for 25,000 acres and not less than 12,000 acres safely irrigated in all ordinary seasons. If the burden of cost is borne by 12,000 acres alone, the average of about $15 per acre would be a moderate outlay. By the expenditure of about $12.50 per acre, distributed over 25,000 acres, it may be possible to increase the irrigable area to cover 25,000 acres safely and to the fullest extent desired. INVESTIGATIONS BY THE PRESENT BOARD. 58. The duty assigned to the board is to determine the feasibility and to report its opinion on the advisability of the San Carlos project for the irrigation of Indian, public, and private lands. 59. The feasibility (physical possibility) of the project depends upon that of constructing a dam of proper height and assured safety in the San Carlos Canyon; the feasibility of all the other elements of the project is manifest. 60. The borings of 1899 and 1903 had shown that bedrock at the 1899 site is at considerable depth, and this site was known to be open to the further objection that it is intersected by a fault. The board at its first inspection decided to begin boring about 1,000 feet upstream from the 1899 site, at a place where abutments of favorable appearance are found, and where the canyon is sufficiently wider at the stream level than at the 1899 site to suggest that bedrock might be encountered at less depth. If this site proved to be unfavorable, and in any event if time permitted, the board decided to supplement by additional borings the work done in 1899 and 1903 at the lower site. 61. The desirability or advisability of the project ("reasonability" in the act) depends upon the relation between cost and the value of obtainable benefits. It involves among other things a determination of the acreage that can be irrigated and a comparison of the producing value of the irrigated land, with the cost of the project per acre. Involved in the determination of the acreage is the determination of the water supply at the reservoir site, the quantity which can be stored and drawn from storage, the reduction of storage capacity by silt accumulations, the losses from the reservoir due to evaporation and seepage, and the various losses which would occur between the reservoir and the lands, and finally the determination of the proper allowance or duty of water per acre. 62. As to water supply, the board realized that in the time at its disposal no great addition could be made to existing data, but hoped from a study of all available data, collateral as well as direct, to obtain more probable values for the mean, minimum, and maximum flows than those which had been deduced from the more limited data of earlier years. 63. It decided also to make a closer determination of the quantity of silt carried by the river, including that rolled along the bottom, than had heretofore been done. It was realized that if the silt approached in quantity that heretofore assumed, the life of the reservoir, measured in years, would be very short, and that the irrigated land would soon relapse into its present desert condition, or that desilting operations must very soon be undertaken. It has long been known that one of the most serious questions affecting the desirability of the project was whether silt accumulations in the reservoir could be disposed of within practicable limits of cost. The board, therefore, has made a study of methods of removing silt from the reservoir and an estimate of the cost of the most promising method. DAM SITES, IMPOUNDING DAM. (Appendixes B, C, and D.) 64. Maps Nos. 4 and 5 and Plates Nos. 2 to 11 show the canyon in which the dam sites now to be discussed are found, the location of holes drilled in 1899, 1903, and 1913, and the information obtained from the borings as to the depth to the bedrock. Two holes were drilled on line E (map No. 4) in 1899 by the United States Geological Survey. In 1903 the United States Reclamation Service completed line E and drilled lines D, C, and B and the easterly part of line A. Under the direction of the present board, line A was completed and line X drilled; also lines G, H, I, J, K, and M. 65. Considering nearness of the canyon walls to each other, the stretch of canyon shown affords two sites worthy of study; they will be referred to herein as the upper and lower sites. Faults cross the canyon obliquely at M and C. The former lies well below the upper site; the latter passes through the lower site, and therefore under or close to any dam that might be built there. 66. The dam discussed and recommended in 1899 (Water Supply Paper No. 33, p. 69 et seq.) was to be placed entirely below the lower fault. The topography of the abutments there is favorable for a dam of the height then under consideration, 130 feet from stream to spillway. The disadvantages of the site are depth to bedrock, 74 feet, and the existence of warm springs on both sides of the canyon there and just below, which taken in connection with the fact that the rock at bottom and on both sides is limestone might well occasion some anxiety as to the soundness or water-tightness of foundation and abutments. Considering only spillway requirements and quantity of material in the dam, the most favorable of all the locations at either site for a dam 160 to 200 feet high would be near line B, but that would involve building the dam across the fault, which should be avoided if practicable. In the Second Annual Report of the Reclamation Service, page 89 et seq., a site just above the fault is referred to as follows: The bedrock problem is very important at this point, as there is a fault in the strata almost directly across the canyon at the dam site, and it became necessary to locate the dam either above or below the narrowest part of the gorge in order to avoid this fault. * * The general result (of the borings of 1903) shows that the depth to bedrock is somewhat less above the fault than below it, and this fact, together with the general desirability of leaving the fault entirely out of the reservoir, pointed to the location above the fault as the more feasible location for a high dam. 67. The borings of 1903 as supplemented by those of 1913 show that the maximum depth to bedrock decreases from about 50 feet at line B to 42 feet at line A and to 33 at line X. The borings of 1913 also show that at the upper site the maximum depth to bedrock is about 20 feet. In order to compare the lower site above the fault with the upper site, it will be necessary to anticipate the conclusion elsewhere reached in this report that the most suitable height of dam for this project, water storage and silt accumulation considered, is 180 feet, forming a reservoir of 709,626 acre-feet initial capacity. The height mentioned is from stream to spillway in the case of a dam with spillway or to crest in the case of an overfall dam. The board recommends for this project, and has considered in the following discussion and estimates, a masonry dam of gravity section, arched in plan. The curved plan is adopted for the usual reasons that it makes arch action available as reserve strength, though the section assumed is sufficient against all the forces acting, including upthrust, and that it lessens the tendency to cracking near the top due to shrinkage and temperature changes. It is also an additional safeguard against sliding, though that danger is absent in the case of this comparatively short dam standing between walls that converge downstream. 68. Conforming to current practice in the case of high dams the board adopts the spillway type, but has given the main part of the dam such form and dimensions that it will serve as an overfall dam if need be. At either site quarrying to obtain rock for the construction of the dam would go far toward providing the space in which to build a spillway 300 feet long. 69. It is judged that in 1905 a flood of approximately 150,000 second-feet passed San Carlos; but a flow of exceeding 50,000 secondfeet is of rare occurrence. Considering the moderating effect of a reservoir with so large an area, seldom, if ever, will more than 25,000 second-feet need to pass over the spillway. (See Appendix D for behavior of reservoir under assumed maximum probable flood.) With water running 10 feet deep over a spillway 300 feet long, the discharge would be about 33,000 cubic feet per second, and accordingly the maximum flow line is assumed at the 190-foot contour. In determining pressures to be used in designing the dam, however, the |