Senator JOHNSON of California. Is the proposed all-American canal indicated upon that map, Mr. Davis?

Mr. DAVIS. Yes, sir.

Senator JOHNSON of California. Would you please designate it upon the map?

Mr. DAVIS. The proposed all-American canal would utilize the Laguna Dam, about 15 miles above Yuma, which is now in use for the Yuma project in Arizona and California; would enlarge that canal to a point about 13 miles below its head, and then build a new canal on that same grade down to where they would very nearly, but not quite, reach the Mexican line. And then follow this red dotted line from the Laguna Dam to the Mexican line, and then westward along the Mexican line.

Senator ASHURST. You say along the Mexican line, but in California?

Mr. DAVIS. Yes; along the northern side of the Mexican line.
Senator ASHURST. How far from the line?

Mr. DAVIS. Oh, as short a distance as possible for a few miles there. Perhaps about a thousand feet from the line. And then follow the best means it can through the sand hills here; at this point [indicating on the map] it takes it a little to the northward, going back to the Mexican line on the west side of the sand hills, and again following just north of the Mexican line, and reaching the Imperial Valley in that way. There is more than 100 feet cut for a long distance through these sand hills, and when it gets out to the surface, then it can send off branches to the northward and water lands that are now too high to be reached by the present works. The present works bring water along this line marked in blue, diverting it a little north of the Mexican line at what is called the Rockwood Gate, a little north of the station called Algodones on the railroad. It roughly parallels the railroad.

Senator JOHNSON of California. All of this being in Mexico?

Mr. DAVIS. All of this being in Mexico after Algodones, and finally comes back to the American line east of Mexicali and west of the sand hills. So that all of this canal has to be maintained in the Republic of Mexico. And there are very great difficulties in doing that because of the laws.

Senator JOHNSON of California. So that the Imperial Valley is wholly dependent upon the flow of the water in Mexico at the present time?

Mr. DAVIS. That is the fact.

Senator JOHNSON of California. And about what length would the all-American canal that you have just designated be?

Mr. DAVIS. The main canal would be between 50 and 60 miles in length to where it comes to the westward edge of the hills, and then it is out on the plain.

The two ridges that the river has formed in recent history that we know of I will refer to. The ridge along this old channel, which is built up to such a height that it was unstable, I have explained. That is a wide, flat-top ridge, but a good deal higher than the country on the west side. East of that is high country and a range of hills, where it would be impractical to put the river, because it is too high. It built another ridge along the Bee River Channel, as I explained, to Volcano Lake, and filled that up so high that it was unstable, a

very imminent menace to the Imperial Valley by reason of the fact that the floods would reach the top of the levee that was built along the north of the Bee River that was built to protect the Imperial Valley, built that up to about 14 or 15 feet on an average, and finally the river was turned down here into the Pescadero, which runs down between these two high ridges. Now, there is that narrow strip that is still low and where the river can flow safely as long as it remains. Senator GOODING. From what country, Mr. Davis, does most of that silt come? It does not come from the higher altitudes, does it?

Mr. DAVIS. No; the Grand River, which carries more than half of the water, is almost clear at its mouth. And the Green River, while it is muddy, is not anywhere near as muddy as the Colorado River below. Most of it comes from the side streams. The San Juan is a very muddy stream, coming from New Mexico and Utah. And most of the small streams have a high declivity and are eroding their sides very rapidly and bringing sand and clay, and that material is rolled down through the canyon of the Colorado and finally ground up to very fine sediment, and the river carries that on down, but when it reaches the lower regions south of the Boulder Canyon the declivity is too small to allow it to carry that entire load, and the river begins to deposit it, and also the water begins to be dissipated by evaporation. The discharge at Boulder Canyon is greater than at Laguna Dam, although the distance between the two points is nearly 300 miles, and the losses are much greater than the increase in that region, on the average. Of course, occasionally some of the small streams get in flood and increase the flow, but that is rare. And the river is carrying that large load of sediment along and is depositing some of it in this region where the declivity is small, and it is only a question of time, and not a very long time, when it wll build up this interior part of its delta so high that it will again become unstable and again be a menace, as it has been for a long time, and then it will break through all over again and flow out on lower ground. We have had that happen heretofore, and as it was running originally here it would run over to Volcano Lake. And that made it safe for a period of 10 years or so, and it ran that way for 14 years.

Now, the last room for taking care of the river is being used, and every year adds to the menace and shortens the life of that possible solution. Of course, just how long it will take to fill that nobody knows, because it would involve a more intricate knowledge of the topography down here in Mexico than we have. It is in Mexico, and we can not explore it very feasibly or in very great detail. But we do know that some time, maybe 7 years, maybe 15 years from now, that will be filled up.

And even if the river were now desilted and regulated, it hasn't any greater storage capacity than we ought to have for the sand that will be deposited there anyway after it is regulated. And it means that the flood control, to be successful and economical, has got to be done quickly and thoroughly.

Senator JOHNSON of California. Mr. Davis, how far is the most southerly point on the canal which now serves the American side from the international line?

Mr. DAVIS. I suppose it is about 15 miles. Perhaps Mr. Rose can answer that question.

Mr. Rose. It is approximately 15 miles. That is, it serves a part of the valley; and some of it is farther than that. There are two canals that serve the valley. The main canal comes to here, and then the other canal that serves the west side of the valley comes around in this way as shown on the map.

Mr. DAVIS. The distance that has to be traveled, however, to reach any point on that canal is much greater, because the only points of supply are to the eastward in the neighborhood of Algodones, or Yuma, and to the westward in the neighborhood of Calexico, and it has to be approached from one of those directions, and along those lines we have a railroad, the Mexican International Railroad.

Senator KENDRICK. Can you tell us how much of the territory in old Mexico there is under irrigation from water taken out of the Colorado River?

Mr. DAVIS. The estimate that I have heard last is about 190,000 acres. They are increasing that gradually and intend to increase it largely in the future if they can, I understand.

There has been something said, Mr. Chairman, about the interests of the upper basin, and that is a subject to which I have given a great deal of thought, and upon which I addressed the League of the Southwest in the city of Denver shortly after we began the investigation of Boulder Canyon dam site and reservoir. That investigation of the Colorad Basin had then been in progress over 20 years, but we were about approaching conclusions at that time, actively investigating still, and I pointed out the need of storage in the lower basin, because the solution of the problems of storage and flood control in the lower river require a very large amout of storage capacity. This storage capacity must intercept the water supply to be efficient. and the only place, herefore, that it could exist and be availble for controlling the floods was in the lower basin.

That had been condemned as a solution by some engineers who had given some study to the basin, and some engineers in the Reclamation Service had considered that unfeasible, not appreciating the feasibility of building such a large reservoir that it would take care of the silt problems for a long time. It was considered by them unfeasible on account of the quick filling of sediment in any reservoir down there, and they are correct so far as it applies to a small reservoir, because a small reservoir would cost so much in proportion to its capacity, the foundation conditions and others being almost as great a problem for a small dam as a high one, and it would fill so quickly and last such a short time that it was unfeasible.

But by this proposal we have now that sediment would be taken care of completely for 100 years or more, and the reservoir would not entirely fill for something like 300 years, depending on how large we build the reservoir and how accurate are our measurements of sediment. Our estimates of sediment are based upon measurements made during several years.

But the upper basin has a number of reservoirs which have been explored and surveys have been made of them. Rough estimates, of course, have been made, and they are listed and published with the capacity at various heights in the various reports. Those that

are feasible and large enough to be considered for this purpose, and so situated as to intercept water supply enough to make them serviceable aggregate too small a quantity to solve the problem. They aggregate about nine or ten million acre-feet, depending on how high they are built, but two of those are of such a character that they would not be fully serviceable because of insufficient water supply to fill them every year, and one of them because of too great a water supply, so that the capacity of the reservoir is not as large as it should be for that particular location. And all four of them have a capacity of only about 9,000,000 acre-feet, which would be insufficient to solve this problem, and, furthermore, they intercept only a fraction of the water supply. So the flood problem would be scarcely touched. It would have some influence upon the normal floods, but the river would still be subject to very excessive floods even if those reservoirs were all built.

But the strongest reason there, in my opinion, is that it would be an economic crime to build and use the reservoirs in the upper basin for service in the lower basin, because it would deprive the upper basin of the local service to which those reservoirs are best adapted. If a reservoir is built in Colorado or Wyoming to control the floods, it must, of course, be used for that purpose. If it is built to extend irrigation in the lower basin, it would, of course, be used for that purpose, both of which would be entirely inconsistent with the uses that the upper basin has for those reservoirs, some of them in irrigation and all of them in power. And they are, none of them, any greater than ought to be built for purposes for which they can serve in the upper basin.

If there were no large reservoir sites in the lower basin, there would be some argument for using not all but a share of the reservoir capacity of the reservoirs in the upper basin in favor of the lower. But that is not the case. There are larger, cheaper, more efficient reservoirs in the lower basin. One at least, Boulder, can be built to a capacity four times as great or five times as great as the capacity of all those four reservoirs in the upper basin combined. The CHAIRMAN. Where are they located, Mr. Davis, and what are their names?

Mr. DAVIS. They are listed in the reports. One of them is at Flaming Gorge, on Green River, and a preliminary permit has already been issued by the Federal Power Commission for a power company to build that for the service of the city of Salt Lake and other points within reach.

The largest of those I mentioned is the Dewey Reservoir on Grand River. The backwater would be a short distance below the Grand Valley in Colorado.

And another is on the San Juan River. And another is the Juniper Reservoir on Yampa River.

On Green River the Flaming Gorge reservoir site has a capacity of 3,120,000 acre-feet, with an average runoff for the basin of 2,300,000 feet, and would cost, estimated, about $16,000,000.

On the Yampa River the Juniper reservoir site has a capacity of 1,550,000 acre-feet, with an average runoff of 1,200,000 acre-feet, with an estimated cost of $4,000,000.

The Colorado River, or Grand River, as it is sometimes called, at Dewey, has a reservoir site of the capacity of 2,270,000 acre-feet, with a mean annual runoff of 6,800,000 acre-feet, so that reservoir would be entirely inadequate to control the Grand River.

On the San Juan River the Bluff reservoir site, with a capacity of 1,300,000 acre-feet has an average runoff in acre-feet of 2,300,000. The total cost, so far as we know-though there have been no detailed estimates made on them-is $40,000,000.

They have a total capacity of 8,340,000 acre-feet, or nearly $5 an acre-foot, while the Boulder Canyon dam, as proposed, would cost less than $2 an acre-foot, having a cost about the same as these, and a capacity about three times as great. It is near enough to the lands that are menaced by the floods of the Colorado to protect them, while none of these other reservoirs are.

Furthermore, it is where it intercepts the whole water supply, and will have an adequate water supply, so that the river can be controlled and the water made useful.

The reservoir will pay for itself in the development of power. That is the practically universal conclusion now by those who have studied the question, and the only question is the place, and height to which the dam should be built. And the principal thing to be remembered in that connection is that any small reservoir would be an investment wasted. It would fill so quickly with sediment and cost so much per acre-foot that it would be a bad investment.

The fixed charge of building a railroad to the site, and diverting the river, and building camp, and installing machinery and everything, is nearly as much for a small reservoir as for a big one, and the larger the reservoir is built the cheaper per acre-foot is the storage, and the cheaper per horsepower is the power development, and, of course, to a greater extent, the more efficient it is in flood control. People talk a great deal about a flood control reservoir of about 8,000,000 acre-feet capacity to control; some put it at 6,000,000, some 8,000,000, and some at 10,000,000 acre-feet. Now it hardly needs to be said that the larger it is the better the control of flood, until it comes to a point where it is so big that no flood can happen. Those figures of eight or ten million are misleading. If you build a reservoir of that capacity, that is all the flood capacity you have there, and that is all there is, and in a few years will be very in-. efficient.

But if we build a larger reservoir, from twenty-five to thirty-five million acre-feet, or anything between those figures, and reserve six or eight milion acre-feet in the top of that reservoir for flood. control, we have just as much as we have in a smaller reservoir in that zone, and then use the stored water through the dry season for irrigation and power, we have the additional capacity available at the opening of the flood season of the Colorado, in addition to the six or eight million acre-feet in the top, and so how much more efficient that is, without any wasting of storage capacity.

The CHAIRMAN. Do you think that power developed there could be sold at a profit justifying the repayment to the Government of the amount of money advanced for the construction of the dam? Mr. DAVIS. Yes, sir; I am very sure.