A dam across the upper part of East Fork Cañon would create a lake behind it which might have an area of 12 to 15 square miles. Numerous reservoirs could be created at small expense in Grass Valley, upon the Fish Lake Plateau, and upon the Sevier Plateau, and in those valleys which are drained by Salina Creek and its tributaries. The Sevier River itself can be cheaply dammed at several gorges and made to overflow swampy flats above--notably at the head of Marysvale Cañon, and again just north of Van Buren’s ranch. Other things equal, it would be better, as well as cheaper, to build dams at higher levels, since the evaporation is much less there than in the valleys, and the natural facilities for creating lakes are also greater. In this way, I believe it to be practicable to reserve a store of water sufficient to irrigate every acre of ground in the Sevier Valley, which is by the nature of its soil and its situation suitable for irrigation. It may be noted, too, that the “tank system” thus suggested would not interfere with or take the place of the present system, but would be supplementary to it. The streams would in June and early July run through the lakes and over the dams, yielding about as much water as they now yield in those months, and the reservoirs would not have to be drawn upon before the middle of July. A very interesting subject connected with the peculiar conditions of agriculture in the west is the origin and distribution of alkaline salts in the soil. In moist regions such occurrences are rare. They are peculiar to arid regions, and, in truth, very few arid regions fail to exhibit them. The cause in a general way is well known. The small amount of rain which falls during the wet season penetrates deeply into the earth, where it gradually takes up such soluble salts as it encounters there. During the dry season which follows, there is always going on an evaporation from the surface, however dry it may appear to the senses. It is a mistake to suppose that because the saline soil is as dry as ashes no evaporation is in progress. In many cases this may be true; but often in the most arid regions there are many localities where the water collects far below the immediate surface. By capillary action, this water always tends to diffuse itself throughout the loose materials which make up the overlying soils. As fast as it is evaporated at the surface, more water from below rises by capillary action to take its place. When the air is exceedingly dry, as it invariably is in summer throughout the whole Rocky Mountain Region at moderate altitudes, the evaporative power becomes so great and extends to such a depth below the immediate surface, that we are unable to recognize the slightest traces of moisture indicating that evaporation is going on. The water which may have accumulated beneath has gradually risen by percolation through the interstices of the unconsolidated materials of the soil, bringing with it whatever soluble