Another partial explanation is found in the meteorological condition. The moisture of this region is brought from the Pacific Ocean and the Gulf of California by the prevailing southwest wind. As this wind ascends the elevations toward the Colorada Plateau its temperature is lowered, which reduces its capacity for holding moisture and increases its relative humidity. When this quantity reaches 100 per cent in any part precipitation occurs. This influence continues until the wind passes the summit, where the process is reversed.

As might be expected, therefore, the hydrographic resources of the country immediately southwest of the Colorado Plateau are disproportionately great when compared with those to the northward. For instance, the precipitation at Fort Apache, as shown by the mean of twenty years' observations, is 19.75 inches; the elevation being 5,050 feet, while the precipitation at Holbrook, at an elevation of 5,047 feet on the northern slope is 8.47 inches, as indicated by the mean of ten years' observations. This is an important fact, especially when taken in connection with the fact that the great areas of valley land with a semitropic climate lie in the southwestern portion of the Territory, and are easily covered by the streams which are formed by the conditions above described, and which constitute the main features of the great Gila River system.

The drainage area of Gila River, including a number of small lost basins which are topographically tributary, but which seldom or never furnish any run-off to the main stream, is about 72,000 square miles, of which nearly 57,000 lie in the Territory of Arizona, about 14,000 in New Mexico, and something over 1,000 in Mexico. The areas in the United States are distributed, with respect to elevation, approximately as follows:

Gila Basin is conveniently divided into four parts. Of these, the most northerly is Salt River Basin, which includes all the territory tributary to Salt River. This again is sharply divided into its valley and mountain portions. Salt River Valley may be taken as including all the territory adjacent to Salt River from its mouth up to the junction of the Rio Verde. Above that point the greater portion of the basin is mountainous, with small valleys on the Rio Verde and in Tonto Basin, and at a few points on Salt River and other tributaries. The Lower Gila district may be taken as the portion lying below the mouth of Salt River to the Buttes above Florence, and including the Pima (Gila River) Reservation and the great Casa Grande Valley. The Upper Gila district includes the valley in the region of Camp Thomas and Solomonville and the tributary mountainous districts. In addition to these main divisions of the trunk streams may be taken the subordinate divisions of tributaries, such as San Pedro, Santa Cruz, Hassayampa, and Agua Fria creeks.

The tables of temperature for the Gila and Salt River valleys show that the climate is very warm. But care should be taken not to exaggerate this feature, for the actual conditions in their relation to human life and comfort are by no means as unfavorable as they might appear to persons comparing these tables of temperature with those of some eastern localities. For about eight months in the year the temperature of this valley is delightful. Cool nights, bracing mornings, and bright, pleasant days are the rule, except in the months of June, July, August, and September. In these months the heat becomes intense, and though there is of course some variation, the temperature remains continually high throughout the greater part of this period.

The physiological effect of this heat is markedly modified by the aridity of the climate. The human economy provides that when the temperature of the healthy body rises above the normal the perspiration glands begin to act and furnish the skin with moisture, the evaporation of which lowers the temperature of the body. An essential condition of this natural safeguard against excessive heat depends upon the ready evaporation of the moisture furnished by nature. In a very

humid climate this evaporation can not occur; and in those portions of the country where the humidity is comparatively high such evaporation must be proportionately tardy and sluggish, so that any considerable temperature above normal blood heat produces great suffering and exhaustion, and even prostration. In an arid region on the contrary, the low percentage of humidity causes prompt and quick evaporation of the moisture and the consequent success of nature in its attempt to prevent uncomfortable and injurious bodily temperature.

In southern Arizona these favorable conditions for resistance to heat are at their maximum. Though the temperature is high, the relative humidity is very low, and every particle of moisture which reaches the surface of the skin is promptly evaporated-so promptly that its presence is not perceived and while the body is thus kept at its normal temperature the unpleasant effects of excessive moisture are not experienced; and the sultry, sticky days so common in the east are unknown in Arizona. Farm labor, the construction of canals, the rounding up and branding of cattle, and other active hard labor are performed at any time in the summer-with less comfort of course-but with no worse effects, than at at any other time of the year, and without actual suffering, the only requisite being plenty of drinking water.

A fair comparison of the sensible temperatures of two places may be obtained by a comparison of the readings of wet bulb thermometers. The difference between the readings of wet and dry bulb thermometers here often exceeds 30°. The summer is, therefore, far from being as uncomfortable as might be supposed, and the delightful autumn, winter, and spring fully compensate for the discomforts of the summer months; and the climate, taken as a whole, with its extreme aridity, its mildness, and its large proportion of sunshine, is exceptionally healthy and especially beneficial to those suffering from bronchial or pulmonary troubles.

RAINFALL AND FLOODS.

Water being one of the most valuable of Arizona's resources, its proper conservation and economical use are of prime importance. There are periods of the winter season when but comparatively little irrigation is necessary, during which the greater part of the entire flow of the streams could, with reservoirs, be held back and stored for use in the spring and summer, thus greatly enhancing its value for the crops during their season of most rapid growth. Reservoirs, therefore, would be very useful even were there no floods to be considered. A storage system would put an end to the lavish use of water in winter, a practice to which irrigators now resort because the water is at hand and going to waste.

But it is in the extent to which the floods can be utilized in the reclamation of the vacant lands that public interest chiefly centers. Much of the rainfall is so torrential in character that a high percentage of the run-off is carried by the streams in flood. Sudden floods are caused also by warm winds and rains attacking the snow in the mountains in the latter end of winter. From long observations of these floods the older residents of the Territory are convinced that the land already under irrigation is but a tithe of that which could be reclaimed were the water stored which runs to waste to the sea. No one doubts that the area reclaimable under the water supply available for storage is many times greater than the area already irrigated. These general statements must suffice, because of the absence of exact data as to the quantity of rainfall and the average yearly volume of water carried in flood. In no other section of the country, probably, are the streams so capricious and fluctuating in volume. Storms of a cloudburst nature in the mountains often add in a few hours many thousands of cubic feet to the flow of the rivers.

The difficulty of making accurate measurements under such conditions is obvious. Records of the rainfall are not as complete as is desirable, but the measurements taken through a long series of years

and at points fairly representative of the whole Territory yield data from which very close calculations can be made as to the annual precipitation.

Observations were begun at Forts Whipple, McDowell, and Bowie as early as 1866-67 and carried forward to a recent period. Other stations were established later, so that for the last twenty years the records of a score of stations are obtainable. It is thus demonstrated that while the mean precipitation in the lower valleys is very slightranging from 3.16 inches at Yuma and 5.34 inches at Casa Grande to 7.60 inches at Phoenix-in the drainage acres proper-i. e., in the region of great run-off-it is much heavier. For illustration, Fort Apache shows a mean precipitation of 19.75 inches; Fort Whipple, 16.06 inches; Fort Verde, 13.24 inches; Fort Grant, 15.45 inches; San Carlos, 13.03 inches; Pinal Ranch, 20.46 inches. But the sections of the drainage area from which the least official information has been gathered are the sections over which the greatest precipitation occurs, as they are the sections of highest altitude, least accessibility, and of sparse settlement, and the observatory stations have been usually located at military posts or in the towns of the lower altitude.

When consideration is given to the wide extent of the drainage areas tributary to the irrigable valleys; to the precipitous, rocky, and barren condition of the mountains, so conducive to a maximum of run-off with a minimum loss from seepage, and to the character of the rainfall, the conclusion is inevitable that in Arizona the water supply can be increased enormously if the larger portion of the floods is stored. And if the Geological Survey's eminently conservative estimate of 2,000,000 acres be accepted as the maximum quantity of land in this Territory for which sufficient water can be developed, I regard it as well within the probabilities to estimate that that amount of land will, when properly irrigated, directly support 1,000,000 people.

In order that there may be an intelligent understanding of the question, in so far as it affects Arizona, it is necessary to review, in some detail, the situation of the irrigation industry as we find it to-day.

It is less than twenty years since the large investment of capital in irrigation enterprises in the West was begun, and hundreds of millions of wealth have been created by the new industry. Individuals and corporations have constructed in the agricultural sections ditches carrying all the normal flow of the streams, and a number of costly reservoirs have been constructed in different parts of the arid West, which impound a great additional quantity of flood waters. Influenced by the phenomenal development of the country under irrigation and by the profits derived from less costly irrigation enterprises, capital was led by too sanguine promoters to investment on a larger scale, which, for years, proved disastrous in many cases. The business of impounding water was new, and it was but natural that many mistakes were made. In some instances due investigation had not been made as to the average run-off of the streams. Incorrect deductions were made from exceptional floods, and after costly dams were constructed it was found that the drainage area and rainfall had not been accurately calculated, and in some years the reservoirs were practically dry. But more frequently, oversanguine calculations had been made as to the beginning of revenue.

Investors found that they had not taken sufficient account of the time which must elapse before these lands could be settled and cultivated sufficiently to be made productive of revenue.

Often there was litigation over the ownership of the water which prevented the revenues earned from reaching the pockets of security holders, and lands owned by the water companies were held at an exorbitant market price. Meanwhile, interest charges accumulated and stockholders and bondholders became discouraged. These things, added to extravagant and unnecessary cost of construction, overcapitalization, etc., served to give irrigation investments a bad reputation. Yet the fact remains that the storage enterprises which were carried to completion are at last, under intelligent and economie management, beginning to pay. With the experience gained from the past, and in possession of the extremely valuable data gathered in the past few years by the industrious engineers of the Geological Survey; with reliable information at hand as to the capacity of reservoir sites, the drainage area tributary to them, and the average precipitation which may be expected; with the cheapened cost of constructing dams and the knowledge which engineers have gained as to the proper mode of construction, the builders of storage dams, operating on conservative and business-like lines, hereafter should not fail to make any ordinary storage enterprise a profitable one for investors as well as for the country which they supply.

SOME RESERVOIR SITES AND STORAGE PROJECTS.

There are numerous reservoir sites in the Territory. On all the interior streams carrying volumes of consequence nature has provided storage sites, and reference will be made only to those of largest capacity.

Perhaps the most attention has been paid to the sites on the Gila by the engineers of the Government, because of the pressing necessity of providing the Gila River Indian Reservation with water for irrigation. The 4,000 Indians on that reservation were formerly selfsupporting, but the waters of the Gila having been diverted by the farmers in Pinal and Graham counties, the occupants of the reservation are no longer able to raise crops sufficient for their sustenance, and under existing conditions they are certain to become absolute wards of the Government at a heavy expense each year for their maintenance. The fact that they are peaceable and fairly industrious long since bespoke for them a careful consideration of the problem of supplying them with water through some system of storage.

Investigations made by the hydrographers of the Geological Survey have yielded valuable data concerning the most prominent sites on the Gila. At The Buttes, 14 miles above Florence, the first reconnaissance was made some four years ago, and since then measurements of the flow of the river have supplemented the information gained in the first investigation. Last winter Congress made an additional appropriation for the purpose of ascertaining the location of bed rock at the dam sites theretofore examined, and during the present year much additional information has been gathered.

It is now definitely known that at The Buttes a dam 150 feet high will store enough water to cover 174,000 acres to a depth of 1 foot, or as it is technically stated, will impound 174,000 acre-feet. The dam would be 300 feet long at the bed of the stream, and 800 feet at the top. The original round-figure estimate of cost was $2,000,000, but that estimate was based on the belief that bed rock is nearer the surface than this year's investigations have shown it to be. But within a recent period the cost of handling rock has been cheapened by