between herself and a sister State. And this is especially true when the waters are, except for domestic purposes, practically useful only for purposes of irrigation. The Arkansas River, from its source to the eastern end of the Royal Gorge, is a mountain torrent, coming down between rocky banks and over a rocky bed. Along this distance it is of comparatively little use for irrigation purposes. After it debouches from the Royal Gorge it enters a valley, in which it wanders from one side to the other through eastern Colorado, southwestern Kansas and into Oklahoma, with but a slight descent, and presenting but little opportunities for the development of water power through falls or by dams. Its length in Kansas is about three hundred and fifty miles, and the descent is only 2,320 feet, or less than seven feet to a mile. There are substantially no falls, no narrow passageways in which dams can be readily constructed for the development of water power; and while there are some in eastern Colorado, yet they are of little elevation and mainly to assist in the storing of water for purposes of irrigation. So that, if the extreme rule of the common law were enforced, Oklahoma having the same right to insist that there should be no diversion of the stream in Kansas for the purposes of irrigation that Kansas has in respect to Colorado, the result would be that the waters, except for the meagre amount required for domestic purposes, would flow through eastern Colorado and Kansas and be of comparatively little advantage to either State, and both would lose the great benefit which comes from the use of the water for irrigation. The drainage area of the Arkansas River in Colorado is 26,000 square miles; in Kansas, 20,000 square miles; and all this area, unless the stream can be used for purposes of irrigation, would be left to the slow development which comes from the cultivation of the soil.

The testimony in this case is voluminous, amounting to 8,559 typewritten pages, with 122 exhibits, and it would be impossible to make a full statement of facts without an extravagant extension of this opinion, which is already too long,

and yet some facts must be stated to indicate the basis for the conclusion to which we have come. It must also be noted that, as might be expected in such a volume of testimony, coming as it does from three hundred and forty-seven witnesses, there is no little contradiction and a good deal of confusion, and this contradiction is to be found not merely in the testimony of witnesses, but also in the exhibits, among which are reports from the officials of the Government and the two States. We have endeavored to deduce from this volume those matters which seem most clearly proved, and must, as to other matters, be content to generalize and state that which seems to be the tendency of the evidence.

Colorado is divided into five irrigating divisions, each of which is in charge of a division engineer. That which includes the drainage area of the Arkansas is District No. 2, divided. into eleven districts. Under the laws of Colorado, irrigating ditches have been established in this district and the amount of water which each may take from the river decreed. In addition some reservoirs have been built for storing the surplus waters which come down in times of flood, and this adds largely to the amount available for irrigation. The storage capacity of six of these reservoirs is shown to be 8,527,673,652 cubic feet. The significance and value of these reservoirs can be appreciated when we remember that the Arkansas, like many other streams, has its origin in the mountain districts of Colorado, and that by the melting of the snows almost every year there is a flood. The amount of water authorized to be taken by the ditches from the river is, as alleged in the bill, 4,200 cubic feet, and from its affluents and tributaries 4,300 feet. (Whenever this term is used in reference to the flow of water it means the number of cubic feet that pass in a second.) The average flow of the river, as it comes out of the Royal Gorge at Cañon City, is as shown by official measurements for a series of years, 750 cubic feet. So that it appears that the irrigating ditches are authorized to take from the Arkansas River much more water than passes in the channel into the valley. It is not clear

what surplus water, if any, comes out of the tributaries. There are some twenty-five of them, the average flow from four of which into the Arkansas is 313 cubic feet. Aside from this surplus water some may be returned through overflow of the ditches or from seepage. What either of these amounts may be is not disclosed. Indeed, the extent to which seepage operates in adding to the flow of a stream, or in distributing water through lands adjacent to those upon which water is poured, is something proof of which must necessarily be almost impossible. We may note the fact that a tract, bordering upon land which has been flooded, shows by its increasing vegetation that it has received in some way the benefit of water, and yet the amount of the water passing by seepage may never be definitely known. The underground movement of water will always be a problem of uncertainty. We know that when water is turned upon dry and barren soil the barrenness disappears, vegetation is developed, and that which was a desert becomes a garden. It is the magic of transformation; the wilderness budding and blossoming as the rose. The writer of this opinion recalls a conversation with Bayard Taylor, the celebrated traveler, in which the latter stated that nothing had contributed so much to secure the steady control of the French in Algiers as the fact that after taking possession of that territory they sank artesian wells on the borders of the desert, and thus reclaimed portions of it, for the Arabs believed that people who could reclaim the desert were possessed of a power that could not be withstood.

Further, adjacent barren ground is slowly but surely affected, and itself begins to increase its vegetation. We may not be entirely sure as to the methods by which this change is accomplished, although the result is undoubted. It may be that water percolating under the surface has reached this adjacent ground. Perhaps the vegetation, which we know attracts moisture from the air, may increase the rainfall, and thus affect the adjacent barren regions.

It appears that prior to 1885 there was comparatively little

water taken from the Arkansas for irrigation purposes-certainly not enough to make any perceptible impression on the flow of the river-but about that time certain corporations commenced the work of irrigation on a large scale, with ditches, some of which might well be called canals. Thus, in 1884 work was commenced on ditches capable of carrying off 450 cubic feet; in 1887 others capable of carrying off 1,481 cubic feet, and in 1890 still others carrying 1,705 cubic feet. Most of these were completed within two years after the commencement of the several works. By the year 1902, according to the report of the Census Bureau of the United States, there were 300,115 acres, in 4,557 farms, actually irrigated.

The counties in Colorado, from Cañon City eastward, through which the Arkansas runs are Fremont, Pueblo, Otero, Bent and Prowers. The following tables prepared by the defendants from various census reports show the population, number of acres cultivated and total value of farm products in these several counties for the years 1880, 1890 and 1900:

These tables disclose a very marked development in the population, area of land cultivated and amount of agricultural products. Whatever has been effective in bringing about this development is certainly entitled to recognition, and should not be wantonly or unnecessarily destroyed or interfered with. That this development is largely owing to irrigation is something of which from a consideration of the testimony there can be no reasonable doubt. It has been a prime factor in securing this result, and before, at the instance of a sister State, this effective cause of Colorado's development is destroyed or materially interfered with, it should be clear that such sister State has not merely some technical right, but also a right with a corresponding benefit.

It may be asked why cultivation in Colorado without irrigation may not have the same effect that has attended the cultivation in Kansas west of where it was productive when the territory was first settled. It may possibly have such effect to some degree, but it must be remembered that the land in Colorado is many hundred feet in elevation above that in Kansas; that large portions of it are absolutely destitute of sod, and that cultivation would have comparatively little effect upon the retention of water. Add further the fact that the rainfall in Colorado is less than that in Kansas, and it would seem almost certain that reliance upon mere cultivation of the soil would not have anything like the effect in Colo