place nearest the meridian where the necessary telegraphic facilities could be obtained, and the Western Union Telegraph Company, with its usual courtesy, placed its line at our disposal for the work.

Prof. Woodward commenced operations at Sierra Blanca May 16, and by May 29 had completed the necessary observations. He was able to exchange time signals with Prof. Pritchett at St. Louis on four nights, giving to the observation pier at Sierra Blanca the resulting longitude of 105° 21' 24.96" west. During this time Prof. Woodward also made latitude observations by the method of zenith distances on 28 different pairs of stars for two nights, giving as the latitude of the pier 43° 35′ 57.98" north. The astronomic observations being concluded, Prof. Woodward returned to Washington, leaving Mr. Davis to locate the position of the one hundred and fifth meridian.

To accomplish this, Mr. Davis first measured a short base line, connected it with the pier of the astronomic station, and then expanded a series of triangles to the eastward, computing his positions as he proceeded until the meridian was reached. The point thus determined was indicated by permanent marks and an azimuth line laid out, which was extended by the local authorities south to the Rio Grande River.

Upon the completion of this work Mr. Davis was directed to return to Washington, D. C.

Topography.-The topography of the area surveyed by the New Mexico division is of both the mountain and plains type, nearly the whole of the Santa Fé and the west half the Las Vegas sheets being of the former, the east half of the Las Vegas and all of the Watrous sheets of the latter.

The mountain area, which is really the southern continuation of the Culebra Range of Colorado, though known locally as the Santa Fé and Vegas Mountains, is drained by the tributaries of the Rio Grande, Pecos, and Canadian Rivers. It presents the general features of two short ranges united at the northern edge of the Santa Fé and Las Vegas sheets, terminating separately near their southern boundary and inclosing a segment of an oval basin drained by the Pecos River.

The western rim of the Pecos basin, or the Santa Fé Mountains, rises to an altitude of over 12,000 feet. Its slopes are comparatively smooth and gentle, descending from the summits into the valley of the Rio Grande River on the west, and on the east into the valley of the Pecos. The southern extremity of this rim is more broken, often presenting mesa-like forms. The draining streams of the western slope are all small, flow in valleys rather than canyons, and furnish but little water except during the melting of the snows. The streams of the eastern slope flowing into the Pecos River are of the same character but have a more permanent flow of water.

The eastern rim of the basin, or the Vegas Mountains, though rising only to an altitude of a little over 10,000 feet, is more broken and presents somewhat the character of a double range, with draining streams having their sources in the higher interior range and cutting through the front or outer one in narrow canyons. This is an especial feature of the northern portion, where the branches of the Mora River have cut the front range into a maze of mesa-like forms. The plains area presents the general features of a level region bearing mesas and crater-like masses of volcanic rock on its surface. The draining streams in the eastern part of the Watrous sheet have cut narrow canyons with comparatively straight vertical walls.

Forests.-A large part of the mountain area on the Santa Fé sheet is covered by a forest growth suitable for milling or lumber purposes, pine and fir being the principal varieties.

This growth is not dense, but thinly scattered over nearly the whole region between the altitude of 8,000 feet and the timber line, which is here about 11,000 feet.

Below 8,000 feet the hills and mesas are usually scantily covered by a growth of cedar and piñon pine, fit only for firewood and other domestic uses.

The plains area is destitute of trees, except a few cottonwoods along the immediate valleys of the streams.

Reservoir sites.-Eight possible reservoir sites of considerable area were reported by the topographers within the limits surveyed. Two of these are in the plains region on the Watrous sheet, four in the foothills on the Las Vegas sheet, and two in the mountains on the Santa Fé sheet. Four are in the drainage basin of the Mora River and furnish sufficient capacity to store all the waters of that stream during the entire year. The others are on smaller streams and have but limited catchment areas.

All over the plains region are numerous depressions that are usually converted into ponds during the winter months, but are dry in summer. It often happens that a number of these could be easily connected by canals, and so furnish storage for considerable bodies of water. The contour interval of the maps is too great to show more than the principal of these depressions.

Irrigable lands.-Within the mountain region considerable areas of irrigable lands are reported along the valley of the Mora and Pecos Rivers. At least 75 per cent of the area of the plains region is reported good fertile land, lying at such altitudes and slopes as permit irrigation, the only question being the limited water supply.

SUMMARY.

The following table shows the locality of each atlas sheet, the scale upon which the final drawing was made, the intended scale for publication, and the contour intervals:

RESERVOIR SITES.

The following table shows the number of possible reservoir sites reported by each division, the atlas sheet area, and drainage basin within which each is situated:

The character of work deemed essential for the requirements of the maps of the Irrigation Survey was set forth in my last annual report. It embraced the correct determination of the drainage areas of all streams and the correct representation of all hypsometric features by means of contour lines representing equal vertical intervals and showing absolute and relative altitude as well as the slope of all surfaces. Certain features of culture were also added, such as the correct location of towns, wagon-roads, railroads, and the larger waterditches.

No attempt has been made to make these maps cadastral in character, though where the scale is sufficiently large they are admirably adapted to serve as a base for such maps. Neither have the maps been made with the expectation that they would give the engineer all the details necessary for the estimates for the construction of his works. To make maps with that degree of accuracy would be to increase their cost to a prohibitory amount. Nor does the constructing engineer require such maps except for a narrow area along the immediate line of his work.

But the irrigation engineer does require, for any systematic or comprehensive investigation of the problems of irrigation, maps which in their construction shall show the areas of the drainage basins of the steams, the slopes of the declivities of their watersheds, the location of all possible reservoir sites, the practicable routes for canal lines, locations for their headworks, and the situation of irrigable lands. It is hardly conceivable that an investigation worthy of the claim to be scientific and comprehensive, or worthy of the patronage of a great Government, could be made without these data. Experience has shown that the maps constructed by the topographic branch of the Irrigation Survey do furnish just such information, and the fact that they are the very first material to be called for by the irrigation engineer is ample proof that their value is appreciated.

The field work of all the divisions was conducted essentially by the same methods, though the manner of its execution varied with the local conditions. It consisted in the determination of linear distances, or horizontal control; of altitudes, or vertical control; and of the conventional representation of topographic forms after the methods explained in my last annual report.

In the California-Nevada and Colorado divisions the horizontal control was still derived from the primary stations of the transconti nental triangulation of the Coast and Geodetic Survey; in the Idaho division, from the base line measured the present season near Boisé City; in the Montana division, from the stations derived from the base measured by the U. S. Geological Survey near Bozeman; and in the New Mexico division, from stations derived from a base line measured by the U. S. Geological Survey near Fort Wingate.

Plane-table traverses, using the compass for directions and some form of odometer for distances, were extensively used in addition to regular triangulation and plane-table work from stations. All public roads, streams, cliff-edges, valleys, and frequent lines across the country were thus traversed and the whole area surveyed covered by a network of accurately determined and adjusted distances.

VERTICAL CONTROL.

The altitudes of points in the area surveyed were determined by horizontal or angular leveling or by the use of aneroid or mercurial barometers. In all cases a number of bench-marks were accurately located on each atlas-sheet area by leveling, and to these were referred all other level or barometric observations, thus reducing the vertical element of the work to a few accurately determined points.

REPRESENTATION.

The representation of topographic forms was obtained by planetable sketching from all the stations occupied either in traverse or