during January, 1895, was about 40 per cent less than that of January, 1911-the lowest during the last three years but that may be due largely to the fact that during the earlier period water was being stored in the Sandy Lake and Pine River reservoirs with practically no flow from them whereas during the later period water was flowing from both reservoirs. As the rainfall in 1894 was more than twice that in 1910, the natural flow during January, 1895, would probably have been greater than during January, 1911. On the upper Minnesota and Crow rivers the ground water was so much depleted by the drought of 1910 that the extreme cold weather during the early part of 1912 caused the flow to fall even below that of the preceding winter. On Zumbro River, the normal flow of which is derived largely from ground water, the full effect of the drought of 1910 was felt in July, 1911, when the flow was lower than during the preceding or the following winter. It therefore appears that, although the minimum flow did not occur during 1910, it was directly traceable to the drought of that year.

It is manifestly unreasonable to base estimates of available. water power wholly on minimum flow which in all probability occurs only once in many years. It is more reasonable to consider chiefly the mean flow for the lowest month of the average low year. The years of low flow from 1895 to date, as shown by the records of the Mississippi and the long-time records of the Minnesota and Red rivers, are 1909, 1910, and 1911. Accordingly, to determine the low flow for the rivers lacking long-time records, the mean of the lowest monthly flow in each low year has been selected. If the flow for the lowest year was very much less than that for the other low years, the record of the lowest year has been disregarded in estimating the available water power.

At some sites it may be possible to install auxiliary stream power to be used only during a part of the year. By fixing rather arbitrarily a period of six months during a low year and a shorter period during a high year for the use of auxiliary power, the mean flow for the lowest of the six high-water months of the low years has been determined.

Thus, for each river that has been surveyed, the possible sites have been selected, and for each site estimates have been made covering

(a) The continuous horsepower at 80 per cent efficiency possible of development during the lowest month on record.

(b) The continuous horsepower (80 per cent) possible during the lowest month of the average low year.

(e) The continuous horsepower possible for the six high-water months of an average low year.

These estimates are based on the short formula HP (80 per flow x head.

cent)=

11

For a few streams where power can be developed at reservoir sites of considerable capacity the power available from regulated flow is estimated.

For other streams, which have not been surveyed but for which elevations at various points are approximately known, a skeleton profile has been drawn to show the fall at certain localities. As topographic maps are not available it is not possible to indicate dam sites, but of these streams the total horsepower in each section is estimated.

No attempt is made in this report to determine the cost of power-either steam or water. Those interested in the subject are referred to "The Cost of Power" by Seth A. Moulton in the Second Annual Report of the Maine State Water Storage Commission.

DRAINAGE AREAS.

Drainage area boundaries were determined from the few special topographic maps available, the approximate contour maps in the reports of Geological and Natural History survey of Minnesota, and the topographic maps in the State drainage engineer's report on the Topographic Survey of Minnesota. These boundaries were then transferred to a hydrographic base map of the State prepared by the State drainage engineer on the scale of 1 inch to 5 miles. The areas within these boundaries were determined by polar planimeters. The Canadian areas for the boundary waters were obtained from the official maps published by the Canadian Government. The small areas in the bordering States were obtained from PostRoute maps of those States.

The total areas of the major basins in the State were adjusted slightly to equal the total official area of the State, and the same. percentage of adjustment .applied to the minor basins comprising the major. The drainage areas in each basin are arranged in the following order:

The main river in the group heads the list and the drainage areas of different points on the river are arranged in descending order, beginning nearest the source. Next the different streams discharging directly into the main river are arranged in descending order beginning with the stream nearest the source. Where more than one area is measured on a stream; these areas are arranged in similar manner. The same order is used in applying to areas of streams flowing into the principal tributaries. (These smaller streams following immediately the streams into which they flow.)

SANITARY CONDITION OF RIVER WATERS.

In order to determine the sanitary condition of the river waters with reference to pollution by sewage and the extent to which they are used for municipal supplies, letters of inquiry were sent to the officials of all the towns which are situated on streams and which contain 500 or more inhabitants. This limit of population makes it practically certain that the inquiries reached all towns having municipal sewage systems and waterworks plants. Settlements having neither waterworks nor sewerage systems have been included with rural population. Information obtained in this manner was supplemented by data obtained from the State Board of Health.

The statistics collected show distances between points of pollution, the average fall of the river, and any ponding that may occur. No examination of the waters, either chemical or bacteriological, has been made to determine the actual spread of pollution, as that subject is too broad to be covered in a general report on the water resources of the State. This report indicates only in a general way the sources of pollution. A study of the quality of surface waters was made by the United States Geological Survey in cooperation with the State Board of Health some years ago.1

Quality of Surface Waters in Minnesota: Water-Supply Paper U. S. Geol. Survey No. 193, by F. F. Westbrook and R. B. Dole. This report contains many analyses of the various river waters.

MISSISSIPPI RIVER BASIN.

MISSISSIPPI RIVER.

SOURCE, COURSE AND TRIBUTARIES.

Mississippi River drains the greater part of Minnesota and consequently is the most important stream in the State. The portions of Minnesota lying outside this basin are the northwestern section, which is in the Red and Rainy river (Hudson Bay) basins and the northeastern part which lies in the Lake Superior basin.

Mississippi River rises in a small lake called Hernando de Soto, situated in the northeastern part of Becker County. From this lake it flows north into Lake Itasca. Above Lake Itasca it is known as Nicollet Creek. From Lake Itasca to the mouth of Crow Wing River it flows almost in a circle, as at this point it is only 75 miles from its source, while the distance following the river is 350 miles. Leaving the lakes, its course is northward, but below the junction. with the Crow Wing it turns to the south and continues in this direction until it finally reaches the Gulf of Mexico.

The total length of the river from its source to the Iowa State line is about 660 miles.

The important tributaries of the Mississippi beginning at the source and following down the west bank, are Leech Lake, Willow, Pine, Crow Wing, Sauk, Crow, Minnesota, Cannon, Zumbro, and Root. On the east bank are Prairie, Elk, Rum, St. Croix, and Black.

TOPOGRAPHY AND GEOLOGY.

The entire drainage basin except the extreme southeastern part, which lies in the "Driftless Area," is covered with a drift sheet ranging in thickness from 100 to 300 feet. The bulk of the drift is composed of blue till, a compressed mixture of sand, clay, and gravel. In the eastern part of the basin, the blue till gives way to red till. In the southwestern part of the State the till is overlain by a layer of loam which is separated from the till by a distinct line of demarkation. Along the valleys of the Mississippi and mostof the larger streams flowing southward are deposits of stratified gravel and sand which are found also as isolated plains in Cass, Wadena, Meeker and Kandiyohi counties. As a rule these deposits lie on the till and in many places they are covered by a finer sand. Extensive lenticular beds of stratified gravel and sand constitute a large portion of the till in the rolling or broken tracts, including the Leaf Hills in the northwestern part of the basin and the Coteau des Prairies in the southwestern part. In the southeastern part of the State the basin is covered with a loess loam or stratified clay which in places is very sandy.

1Abstracted from N. H. Winchell, Final Report on the Geology of Minnesota, Vol. 1.

The surface of this drift sheet forms a somewhat undulating plain with comparatively slight irregularities which form long, low swells and hollows. Many of the depressions have no outlet and to them are due the multitude of swamps and lakes in the basin.

From Lake Hernando de Soto to the Falls of St. Anthony the river flows almost exclusively through a drift-covered region. Down to Pokegama Falls it occupies a valley which is in some places narrow, in others broad and savanna like, with many rapids. in the narrower, and with gentle or sluggish currents in the broader portions. In this part of its course it drains a number of lakes, among which Bemidji, Cass, Winnibigoshish, and Leech are the most important. The first rock in place is at Pokegama Falls, and thence to the south of Crow Wing River which enters from the west, the average width of the stream is 300 feet, the valley is less winding, and the current is good, with many rapids of small extent.

Below the mouth of the Crow Wing the river flows in a general southeasterly direction to the southern boundary of the State.

Within this stretch are several rapids-the chief being Little Falls and Sauk Rapids-and many timbered islands. The banks are abrupt, of clay or sandy loam, and lead to meadows that stand 60 feet above the river. At the Falls of St. Anthony the river pitches down a vertical fall and rapids amounting to 80 feet in half a mile, and in so doing leaves the prairie and clay banks for a channel that lies between rocky bluffs of limestone and sandstone, which continue for many miles down the river, gradually increasing to a height of 500 feet as the bed sinks below the general prairie level. The sides of the bluff are not vertical, bare surfaces of rock, but are composed of easily eroded stone and drift, which form well-wooded or grassy slopes. It is believed by geologists that the gorge from the mouth of Mississippi River to St. Anthony Falls was caused by the gradual wearing away of the falls which were originally at the mouth of the Minnesota.

Minnesota River enters the Mississippi about eight miles below St. Anthony Falls, and below its mouth the width of the main stream averages 1,000 feet. From this point to the State line it is a broad, placid stream. In many places, especially where tributaries enter, fertile flats lie between the river and the bluffs. Fiftyfive miles below the mouth of the Minnesota is Lake Pepin, an expansion of the river apparently caused by the immense quantities of sand brought down by the Chippewa.