The value of water for irrigation purposes is steadily increasing throughout the arid region. In some of the well-watered States bordering on the continental divide, canal water rights some fifteen years ago were considered-high at $10 an acre. The same rights now sell readily for double and treble that amount. The estimated cost of water rights under the Government irrigation projects has been increased from time to time and now averages nearly $40 an acre. In localities possessing a climate adapted to high-priced products, like that of southern California, the value of water is much higher. The last water rights sold under the Gage Canal of Riverside, Cal., cost the purchasers $250 an acre.

By reason of the scanty rainfall a western farmer must not only secure a water right for all the arable portions of his farm, but he must likewise contribute an annual water rental ranging from 50 cents to $5 an acre to defray the expenses of operating and maintaining a canal system. These charges when added to the cost of preparing his fields for irrigation and of applying the requisite amount of water make the total average cost per annum about $10 an acre.

The results of investigations by the Office of Experiment Stations on the use of water in the West during the past decade point to the conclusion that by far the greater part of the water diverted from natural sources of supply for irrigation purposes is wasted. This line of investigation in the warmer, fruit-producing localities of the West, where water is scarce and valuable, led to a consideration of how much water passes into the air in the form of vapor from newly irrigated orchards and fields. This work was begun at Riverside, Cal., several years ago and the results are summarized in Bulletin No. 177 of the Office of Experiment Stations. These results showed the great influence which dry, granular, soil mulches exerted in checking evaporation from the damp soil beneath. So large a percentage of irrigation water was conserved by this means that it was decided to broaden the field of operations and determine what effect such mulches would exert in other localities of the West. Accordingly, experiments were established in the spring of 1908 at Davis, Cal., Wenatchee, Wash., Reno, Nev., Mesilla Park, N. Mex., and Bozeman, Mont., the experiments being carried on in cooperation with the State experiment stations.

The equipment consists of eight or more water-jacketed tanks, the outer vessel being 27 inches in diameter and 43 inches deep and the inner 23 inches in diameter and 46 inches deep. The inner vessel has sufficient capacity to hold nearly three-quarters of a ton of soil, and when placed inside the larger vessel the 2-inch space between the two was filled with water so as to maintain an even temperature approximately equal to that of the adjacent soil. Figure 22 shows the tanks and their position relative to the ground surface. After being installed the outer tank was not disturbed, but the inner tank, containing the soil, moisture, and mulch, was hoisted and weighed at frequent intervals. In filling the inner tank with soil an effort was made to secure conditions similar to those in the natural soil. Each foot in depth of soil as it was excavated was kept separate, was placed in the tanks in 2-inch layers, and slightly compacted by tamping.

Fig. 22.—Water-jacketed tanks used in evaporation experiments.

The soil used was carefully examined and the amount of moisture it contained determined. The temperature of the air was likewise observed and recorded, as well as that of the water and soil surfaces, during the time of each test. The results are briefly summarized by means of the following diagrams and tables.

The site selected for the experiments was part of an open field on the university farm, bordering on the town of Davis, located in the Sacramento Valley, 76.miles north of San Francisco and 13 miles west of Sacramento. The soil is a rich brown loam to a depth of 2½ to 8 feet, and gradually merges into a light sandy subsoil.

From June 10 to July 1, a period of twenty-one days, the length of the first test, the daily maximum temperature of the air in the shade averaged 87.4°F. and the minimum temperature 49.3° F. During the day the temperature of water in an open tank varied from 60° to 88° F and averaged 76°, while the average temperature of the soil taken 6 inches below the surface was several degrees higher than that of the water.

In filling the tanks with soil it was arranged that Nos. 1 and 2 should have no mulch, 3 and 4 a 3-inch mulch, 5 and 6 a 6-inch mulch, and 7 and 8 a 9-inch mulch. Before beginning the test each tank received a medium irrigation of 6 inches in depth, equivalent to 94 pounds. The tanks were weighed on Wednesday and Saturday of each week. The evaporation losses occurring between these dates are shown in the following table and graphically in figure 23.  The dailyrate of evaporation computed on the basis of percentages of the total amount of water applied in one irrigation is shown in figure 24. By this it will be seen that the tanks having no mulch lost on an average 11 per cent in three days after the water was applied to the surface, while the loss at the end of twenty-one days was 21.92 per cent.  The loss in the tanks protected by dry-soil mulches was as shown.

Fig. 23.―Evaporation at Davis, Cal., 1908.

A second trial was made at the same place in the same manner and with the same equipment between September 1 and October 3, 1908, a period of thirty-two days. The weather conditions as recorded did not differ materially from those of June, when the first trial was made. The results of the various weighings are shown in the following table and also graphically in figure 25.

The daily rate of evaporation (fig. 25), as in the June test, shows an excessive evaporation from the unmulched soil during the first few days after water was applied. This loss in three days amounts to more than 17 per cent of the water used, and in thirty-two days to nearly 35 per cent. The loss in the tanks having a 3-inch mulch was 14.71 per cent, in those having a 6-inch mulch 5.94 per cent, and in those having a 9-inch mulch only 0.78 per cent.

FIG. 25.—Curves showing daily rate of evaporation, Davis, Cal.

The experiments in the State of Washington were made on an orchard in the fruit district of Chelan County, near the town of Wenatchee. The altitude of this locality is about 850 feet, and in both climatic and soil conditions it resembles much of the orchard lands of the north-central portion of the State. The soil is a sandy loam several feet in depth, and contains more or less grit. The annual precipitation is about 15 inches.

The equipment and the manner of conducting the experiments were similar to those previously described. The first trial extended from June 2 to 24, 1908, and during this period of twenty-one days temperatures of the air in the shade, taken at noon of each day, averaged 79.7° F

The graphic illustration given in figure 26 shows the great difference in the amount of water evaporated from freshly irrigated soils without mulch and those which contain an equal amount of moisture protected by dry-soil mulches of varying depths. Thus the average percentage of moisture lost by evaporation in the tanks having no mulch was 14⅓, while the loss from the tanks which were protected by 3-inch, 6-inch, and 9-inch mulches were 3.98, 2.10,and 1.06 per cent, respectively.

FIG. 26.—Evaporation at Wenatchee, Wash., 1008 (on the left), and at Riverside, Cal., (on the right).

RESULTS OBTAINED AT RENO, NEV.

The experiments were conducted on a plat of ground in a corner of the agricultural experiment station-farm near Reno. Other than a light woven-wire fence, there were no obstructions to wind, sun- shine, and rain. The soil used was a sandy loam containing small fragments of rock, and in drying after being irrigated the surface crusted over more or less.

On the morning of June 9, 1908, the tanks were filled with soil containing on an average 11 per cent of free moisture. On June 12 sufficient water was applied to equal a 6-inch irrigation, and after it had been absorbed the mulches were added and the weight of each tank recorded. Thereafter the entire set of tanks was weighed regularly twice a week for three weeks, and from the losses indicated by these weighings the diagram shown in figure 27 has been prepared.  Here as elsewhere the amount of evaporation decreases with the increase in the depth of soil mulch.

Fig. 27.—Evaporation at Reno, Nev.,, 1908.

The average percentage of moisture evaporation in the tanks having a 9-inch mulch was 1.96, that in the tank having a 6-inch mulch 4.74, and that in the tanks having a 3-inch mulch 8.26, while the tanks without any mulch lost on an average 20.39 per cent.

A slight amount of rain fell during the test, but it was allowed for in the determinations. The average temperature of the air in the shade, taken at 8 a. m. and 5 p. m. on each day, was 70.9° F.

The site selected was on a part of the State agricultural experiment station near Bozeman, in the Gallatin Valley. The soil is a silt loam on the surface, underlaid with a heavy clay loam. The altitude of the station farm is 4,750 feet. As one would naturally expect, the winters are usually dry and cold, the springs cool and rainy, and the summer days hot, with comparatively cool nights.

The mean monthly temperature for September for eight years has been 52.8° F., and the average precipitation for the same period 1.23 inches. The experiment was carried on with the usual form of apparatus, the tanks being filled September 1, and the experiment begun on September 2. The average of the temperatures at 9 a. m. and 4.30 p. m. during the experiment was 65° F.

The influence exerted by a layer of dry-soil mulch is again shown in figure 28. It will be observed that the amount of water evaporated is much greater in this experiment than it is in others where the weather was warmer and where one would expect a greater loss from this cause. This is readily accounted for by the excessive amount of moisture in the soil when the tanks were filled, the percentages ranging from 21 to 26 per cent of the dry weight of the soil.

FIG. 28 —Evaporation at Bozeman, Mont., 1908.

CONCLUSIONS.

The results of experiments herein briefly summarized are fairly uniform in character. This is a little surprising when one considers that the experiments were carried on in widely separated localities of the arid region by different agents and under different climatic and soil conditions. These facts seem to warrant the presentation of the following conclusions:
   It seems to be clearly demonstrated that a large part of the water which is spread over the surface of soils in summer irrigation passes from the soil into the atmosphere without serving any useful purpose.  In the tests recorded the time varied from fourteen to thirty-two days and averaged over twenty-two days, while the average percentage of irrigation water which was lost by evaporation from surface-irrigated soil was 27 per cent of the amount applied.

Figures 24 and 25 likewise show that in the case of surface irrigation and no mulch the greater part of the total loss from this cause occurs during the first few days after the water is applied. It is well known that the more water a surface soil contains the greater is the evaporation. On completely saturated soil at Riverside, Cal., the evaporation was found to be over 4 inches a week in midsummer.*

In devising ways and means of checking the excessive evaporation losses from irrigated orchards and fields, it has been found that the deep furrows and the dry, granular soil mulch are the cheapest and best preventives. The influence of the latter is clearly shown in the tables and diagrams of this article. Judging from these the deeper the mulch the less the evaporation, but there are practical considerations which limit the depth of soil mulches. A depth of less than 9 inches and more than 3 inches would meet the requirements of the arid region in general.

*Evaporation Losses in Irrigation. Eng. News, vol. 58, No. 12, pp. 304-306, Sept. 19, 1907.