A station at the bottom of a valley near a more elevated plateau receives almost as much rain as that plateau-Settons, at 1,955 feet (596 mètres), is as wet a place as neighbouring stations at 2,959, 2,625, and 2,133 feet (902, 800, and 650 mètres); the rainfall exceeding in 1872 78.74 inches (2,000 millimètres). At stations which are not in the vicinity of these heights the rainfall is considerably less; for instance, at Saulieu and Château-Chinon, 1,827 feet (557 mètres) and 1,768 feet (539 mètres) respectively, though nearly the altitude of Settons, it amounted only to 37.79 inches (960 millimètres) and 50 inches (1,271 millimètres).

The number of rainy days near the sea is much greater than at any other part of the basin; in 1872 there were at Yvetot 223, at Caudebec 200, at Fatouville 207, while the mean number for the entire basin does not exceed 164. The mean rainfall over the entire basin in 1872 was 34.65 inches (880 millimètres); the average for eight years being 27.87 inches (708 millimètres). The mean results from observations over the great river-basins during 1872 were—

The Paper also describes the tranquil and torrential streams of the Seine basin, and the peculiarities of their different floods.

L. G.

The Hydrology of the Basin of the Seine. By M. A. DELAIRE.

(Annales du Conservatoire des Arts et Métiers, No. 138, 1874, pp. 335-392.) The entire basin of the Seine is subject to similar meteorological influences, and, though the amount of rain varies with the locality, yet the season, whether wet or dry, is of the same character along the whole extent of the basin, as proved both by rainfall observations, and by the height or lowness of the streams. The winter rainfall determines the moisture of the soil throughout the year, as the summer rains produce little effect on the rivers or springs. The sources and the nature of the rivers vary according to the geological character of the district; thus where the soil is impermeable-as, for instance, the granite of Morvan, the lias of Auxois, and the clay of Champagne-the streams are numerous but small, and, as the water runs quickly off the surface, the streams and rivers swell rapidly, in time of rain, forming torrents, the beds of which are generally dry in summer. In permeable soils, such as the öolitic limestone of Bourgogne, the white chalk of Normandy, and the sands of Fontainebleau, the streams rise in

the marshy meadows of deep valleys, the springs are considerable, and are frequently found at the junction of permeable with impermeable strata; the rivers are few, not generally liable to dry up, and their course is gentle. The impermeable soils occupy 7,722 square miles (20,000 k. q.) out of the total area of 30,501 square miles (79,000 k. q.) of the Seine basin.

The torrential rivers have frequent, high, and rapidly rising floods; the rivers flowing through permeable strata rise slowly, and only swell moderately; but, owing to successive rainfalls being merged together, they continue in a state of flood for a long period, and consequently are more injurious to the adjacent lands. Some rivers fed by torrential and other tributaries occupy an intermediate position, and their highest floods occur when a torrential flood follows several other successive risings. The torrential floods flow through Paris at the end of three or four days, and the other flood waters follow three or four days later. An ordinary rise of flood at Paris is about 11 feet (3.5 mètres), and a rise of 23 feet (7 mètres) has only occurred eight times since 1649; extraordinary floods being the result of a thaw followed by excessive rain.

The bridges over a torrential river have to be larger and more numerous than those over rivers flowing through permeable strata, but their span need not be increased proportionately at the lower points of the river, as the flood from one torrent subsides before another from higher up arrives; whereas floods in rivers flowing through permeable or mixed strata are increased by each affluent. In most parts of the basin the rivers have a navigable depth of water of 4 feet 11 inches (1.5 mètre), but the locks vary from 108 to 590 feet (33 to 180 mètres) in length, and from 16 to 39 feet (5 to 12 mètres) in width. The torrential rivers are of little use as water power, their flow being irregular, but for agricultural purposes, and for supplying canals, the flood waters are led into large ponds to provide against the summer drought. For mill-dams across rivers flowing over impermeable strata large weirs and no sluices are required; over slightly permeable strata both weirs and sluices, and over permeable strata sluices only, are

necessary.

Water for domestic use is obtained purer from springs than from rivers, even though containing in general more salts of lime. The purest sources are in sandy soils, and the hardest water comes from the limestones, lias, and gypsiferous strata. The water supply of Paris is obtained from the Dhuis and the Vanne.

L. V. H.

Flow of the West Branch of the Croton River.

By J. JAMES R. CROES.

(Transactions of the American Society of Civil Engineers, July 1874, pp. 76-86, 2 pl.)

The first storage reservoir for impounding a portion of the surplus flow of the river, begun in 1866, on the western branch of the Croton, in the town of Kent, 60 miles north of New York, receives the drainage of 20.37 square miles; the surface of the watershed being broken, the hillsides steep and rocky, the area covered with timber and grass, and the rock underlying the area a compact gneiss. Two thousand five hundred and ten observations of the flow of the stream were made near the dam, from April 1867 to November 1872, during certain periods as often as three times a day. All the water flowing from the drainage area to the reservoir was caused to pass over a weir with a horizontal crest, and provided with vertical side boards 1 inch thick. In the first series of gaugings the weir was 24 feet long, in subsequent ones 21 15 feet, afterwards reduced to 18.02 feet. The channel just above the weir was 3 feet wider than the weir was long, with parallel sides, planked for 15 feet, above which it widened. The height of the crest of the weir above the bottom of the channel on the upper side was about 18 inches, and there was a clear fall of about 30 inches on the lower side. With this weir and heads of water of 0.15 foot to 4 feet perfect contraction was obtained. The head was measured by a float gauge inclosed in a box, and attached to a graduated rod. Ordinarily there was no difficulty in reading this gauge to 0.005 foot. When the float was unsteady, the mean of the oscillations for several minutes was taken. The gauge was placed far enough above the weir to be unaffected by the slope of the surface of the water in passing over the weir. In gaugings later than 1868, the channel was obstructed by a dam with two openings of 4-feet diameter.

In storms the flow increased as soon as the rain began to fall, the maximum being reached six to eight hours after the rain had ceased. On the 1st of August, 1867, the flow was 28,000 cubic feet per hour. Between 7 P.M. and noon next day 1.96 inch of rain fell, and by 6 P.M. the volume had increased to 336,000 feet per hour. On the 15th, rain having fallen in the interval, the flow at 7 P.M. was 242,000 cubic feet per hour. At 10 A.M. of the 16th 3.38 inches of rain had fallen, and by 6 P.M. the flow had reached 2,432,000 cubic feet, an average hourly increase of 95,000 cubic feet. This was the greatest discharge observed. On the 17th of February, 1870, the ground being covered with snow, thaw set in, with 2.41 inches of rain in twenty-eight hours. The water came in such quantities that the openings in the dam were insufficient to carry it off; it rose 20 feet, nearly 50,000,000 cubic feet having accumulated behind the dam. For twenty-four hours the flow was 2,200,000 cubic feet per hour, making 52,800,000 cubic feet per day;

the maximum discharge must have been about 3,500,000 cubic feet per hour; while during the month of September 1870 the total discharge of the stream for thirty days was only 4,508,000 cubic feet. Accompanying diagrams show the monthly rainfall from June 1866 to January 1874, and the flow of the stream during the forty-nine months in which gaugings were made; a set of tables gives the depth of rainfall on the entire watershed, the ratio of flow to rainfall, the yearly proportion of flow to rainfall, and comparisons of different gaugings.

J. D. L.

Relation between Water Levels of Main Rivers in Holland.

By J. P. DELPRAT.

(Tijdschrift van het Koninklijk Instituut Van Ingenieurs, No. 1, 1874, pp. 1-12.)

The purpose of this investigation is to find a simple formula from which the height of the water level at any point of a river may be deduced, when the height at any other station along the same river is known by observation during two consecutive days.

Supposing the increase and decrease in height of water level along the river to be changing in the same ratio, the formula becomes:

H = a+bh ch1, where

a, b, and c denote constant co-efficients, which depend upon the form of the river bed, and the distance from each place to the station, taken as the basis of the calculations and observations;

h and h1 represent the height of the water line during two consecutive days at the main station;

H the same height on the latter of the two days at any other point along the river.

The constant co-efficients are to be deduced from a series of observations. In Holland these observations are made daily with regard to the height of the water along the main rivers, at distances varying from 4 to 30 kilomètres. The correctness of the formula is checked by comparing the results of the calculations with the observations at an earlier or a later date than those chosen for the basis. The Author has calculated the value of a, b, and c for several places along the two main branches of the Rhine, called the Waal and the Neder-Rhijn, or Lek, where observations showed a considerable rise and fall of water during a short time. When applying the formula to cases of an earlier or a later date, and comparing the results with those already obtained, the differences were found to be very small, seldom amounting to a décimètre. As soon as the rivers get into an abnormal state-for instance, in consequence of the breaking or overflowing of a damthe formulæ are no longer applicable; the differences between

observation and calculation amounting in some cases to 50 or 70 centimètres.

Here the formula indicates the influence of accidents upon the height of the water. If, for instance, the heights at the main. station and at any other be observed, the height for the latter is deduced from the formula, the result giving the height of water under normal circumstances; nearly the whole variation therefore is to be considered the consequence of the abnormal conditions. Several examples and tables prove the correctness of the formula, and its practical value. J. M. T.

Observations on Subterranean Water in Dresden.

By HERR MANCK.

(Protokolle des Sächsischen Ingenieur-Vereins, Sept. 7, 1874, pp. 4–9.) It having been asserted by the Munich physician Von Pettenkofer that the lowness of subterranean water was distinctly connected with outbreaks of epidemics, observations of subterranean watercourses were commenced in Dresden in 1867, and have been continued down to the present day, at ninety-two wells selected in various parts of the old town, the new town, and the suburbs. Eighteen principal wells were examined every Monday morning at six o'clock, and the other seventy-four on the first of every month; the operations being under the immediate direction of the Author. As a measuring instrument he employed an impervious tape with a slate stave attached to the end. The relation between the surface of each well and the level of the Elbe being exactly known, a simple calculation was sufficient to determine the height of the subterranean water-level above that level. The heights thus ascertained varied considerably; on the left bank of the river between 13.1 feet and 72.2 feet (4 mètres and 22 mètres); on the right bank between 16.4 feet and 59 feet (5 mètres and 18 mètres), the difference being partly caused by the variation in the surface, partly by the impervious layers of stone on which the water moved. These layers consisted principally of ragstone at a depth of from 39 4 feet to 54 feet (12 to 16.5 mètres) below the surface; the upper strata being on the left bank coarse gravel and pebbles, and on the right bank fine gravel and sand. It was found that in the strata of coarse gravel and pebbles the variations in the water level were much more pronounced and sudden. By means of horizontal curves drawn on the plan of the city to represent the simultaneous observations, and by lines connecting different wells, the Author was enabled to calculate with tolerable accuracy the rapidity with which the water travelled from one pump to another, and having taken a line running from three such wells to the Elbe, he found that in thirty-five days the water percolated from the Königsbrücker well [1874-75. N.S.]

2 B