south of Klamath Falls and below the lake level were visited. In the evening various projects were discussed with prominent citizens of Klamath Falls, and it was agreed that the project which appeared most feasible was one in which Clear Lake, about 7 miles south of the border line (in California), should be used as a reservoir and the waters carried down Lost Liver to Langells Valley and below.

On the 14th, in company with Mr. Hamaker, Clear Lake and Langells Valley were visited, returning to Klamath Falls on the 16th by way of Bonanza, Poe Valley, and Olene. Practically all of the land. bordering on Clear Lake is in private ownership. Willow Creek seems to be the only available source of water supply for this lake as a reservoir, and the adequacy of the supply is questionable. It is understood that no water whatever flows through Langells Valley during the dry season, and it is not believed that a very large amount comes down Willow Creek. The waters of Willow Creek also are controlled by private interests, a dam diverting the flood waters directly into Lost River and thus partially reclaiming swamp lands lying northeast of Clear Lake.

At Bonanza Lost River appears again in some very large springs, the discharge increasing until at Olene it is estimated to be not less than 75 second-feet. Some of the springs are hot. From the amount of water appearing at these springs it is deemed possible that the supply in Willow Creek has been underestimated.

In Link River, between Upper Klamath Lake and the town of Klamath Falls, there is a fall of about 65 feet in a little over a mile. The river at the outlet of the lake is from 300 to 500 feet wide, and its bed is of hard basalt. The river is similar to the Deschutes in many respects. Its discharge is probably between 2,500 and 3,500 second-feet throughout the year. At this time at the bridge in Klamath Falls, it was carrying, as determined by floats and from the best information as to depth, 2,500 second-feet. This is said to be about the time of lowest water. High water is not over 2 feet above this.

Upper Klamath Lake covers about 70,000 acres. From these data it is estimated that a controlling dam 30 feet high would hold back the entire flow for a year. A dam 20 feet high would retain it for all except 135 days of the irrigating season, when 8,000 second-feet would be available. Assuming that 100,000 acres could be watered by a gravity canal, 50,000 by a 100-foot lift, and 50,000 by a 200-foot lift, and that one-third of this flow would be required for irrigation, there would remain available 40,000 horsepower for 135 days.

Inasmuch as there would be no long lines for the transmission of power, it is thought that at least 25,000 effective horsepower could be estimated upon. About 8,000 effective horsepower would be required to raise the water for the 50,000 acres on the 200-foot level, making nearly the 25,000 horsepower available.

At Keno, Klamath River begins to drop very rapidly, probably not less than 40 feet to the mile, and it is possible to lower the bed of the river by blasting at that point, so that not only all the swamp land lying north of Lower Klamath Lake can be reclaimed, but a large portion and perhaps all of the lake itself.

A slough now connects Lost River naturally with Klamath River, and at very small expense, comparatively, Lost River could be diverted into Klamath River, and thus more or less of Tule (Rhett) Lake could be reclaimed.

This project would affect the swamp lands north of Upper Klamath Lake, but their value is small compared to that of the area which would be reclaimed.

The altitude of the lands referred to is about 4,200 feet, the same as that of Harney Valley. The climate is similar, though probably slightly warmer. Two full crops of alfalfa are grown, besides using the land for pasture for a short period each year. Irrigated and improved land is selling for $40 and $50 per acre.

INVESTIGATIONS ALONG SNAKE AND COLUMBIA RIVERS. a

UMATILLA AREA.

An investigation was made in January, 1903, to ascertain the feasibility of irrigating the large body of land lying on both sides of Columbia River for 20 miles above the town of Umatilla.

Levels were run back up the Columbia Canyon to Wallula, then up Wallawalla River along the line that would be followed by a canal and back to a point east of Wallula. Here the elevations were checked from a United States Geological Survey bench mark in front of the station, which is at an elevation of 322 feet. The line that had been run was 241 feet higher, or at an elevation of 563 feet. The elevation of low water at Riparia, Wash., 50 miles up the canyon of Snake River and 75 miles from Wallula, is 510 feet. By making allowances for grade from commanding point 24 miles below Wallula to Riparia, the ditch line opposite Riparia would be at an elevation of 663 feet, or 153 feet above the river.

In view of these facts, the scheme of irrigating the Umatilla lands from Snake River was deemed impracticable.

PALOUSE-PASCO AREA.

The practicability of irrigating the lands near Pasco, Wash., north and east from the junction of Snake and Columbia rivers, was also investigated. There is a fine body of land lying between the wheat belt on the highest benches and these rivers, but as all but a narrow strip of perhaps 2,000 acres would require a ditch fully as high as for the Umatilla lands, it is equally impracticable.

a Extract from report by James G. Camp.

The possibility of obtaining water from Columbia River at Priest Rapids, 40 miles above Pasco, was next investigated. In the 10 miles of rapids there is a fall of 75 feet, and in the 30 miles from the foot of the rapids to Pasco the fall is about 32 feet, making a total of 107 feet. The lowest practicable grade for the canal, 1 foot to the mile, would give for the ditch line at Pasco an elevation of 67 feet above the river. From this height only a narrow strip along the river under the bench could be irrigated, and even that is rendered impracticable owing to what are known as the White Bluffs. These are almost perpendicular

FIG. 45.-Investigations along Snake and Columbia rivers in Washington and Oregon.

bluffs of clay and sand over 100 feet high, and they are constantly caving into the river, making a ditch or flume along their face out of the question.

These conditions lead to the belief that the irrigation of these lands by a gravity ditch from Columbia or Snake River is not feasible. What might be done with pumps using steam, gasoline, or electricity generated at a big power plant located at Priest Rapids, and transmitting power by wire to different pumping stations, was outside the scope of these investigations.