Banks of Rivers, are those natural boundaries within which . 7 stream is confined, accord-to the magnitude and velocity of its current. But as the course of rivers is frequently rapid and irre-r, taking different directions, and often turning at acute angles, extensive inundations, especially in high spring tides, necessarily happen from the overflowing of their banks. Hence it is of the utmost importance to every inhabitant in the. vicinity of rivers, to possess some knowledge of the proper method of forming embankments, for the prevention of floods.

Although we cannot, consistently with our limits, attempt a full mathematical analysis, yet we shall lay down a few general hints, maxims, by which the reader may be guided in the practical view of this subject. I. The principal point to be ascertained, is the elevation, or the heights necessary to be given to such hanks. This must be regulated by the additional quantity water which, according to former experience, the river brings down during its freshes; and likewise by the distance, at which the artificial bank is to be constructed, from the natural boundary- of the stream. On this important point, mistaken economy frequently defeats its own purpose. If, therefore, theembankment is to be raised at some distance from the natural banks of the river, both a comparatively smaller height and base will be re* quired ; the saving will be in the duplicate proportion of the former, and the works will he likewise the more durable, nearly in the same ratio; because, by enlarging the additional bed given to the swollen river, its velocity and power of ruining the works are, likewise, •accordingly diminished. Unless therefore, the freshes of the stream be loaded with fine sand, which might decompose the turf, the embankment should always be'undertaken at a considerable distance from the edge of a river. By placing the artificial bank at half the breadth of the stream, from its natural banks, its channel will thus be nearly doubled, and the detached space, in general, afford excellent pasturage.

2. The next circumstance to be attended to, is, that the river will rise higher, when embanked, than it did at the time when it was suffered to overflow ; and hence the difficulty of ascertaining to what height it may rise, from the greatest swell which has been observed in its former floods. For this son, the utmost rise in some gorge, when the river could not extend farther, should be accurately marked, as far as can be remembered by the oldest inhabitants. Now the increased section in this place should be measured ; and, as the water rises in a much greater proportion than the section, the latter must be increased nearly in the same proportion as the gorge already mentioned. Those who neglct this method of regulating the proper height of the embankment, by the greatest swell that has in former floods been observed in the plain, are in danger of constructing their banks too low, arid consequently rendering them totally useless.

3. The whole embankment should, as much as possible, be conducted in an uniform line, and by the concurrence of the proprietors of both banks; • because the general effect to be aimed at, consists in rendering the course of the stream straighter than it was before. All bends should be made less abrupt, by keeping the embankment farther from the river in all convex lines of the natural bank, and approaching to it nearer, where the latter is concave.— Thus the action of the waters on the embankment will be considerably diminished, and the duration of the work insured. On the same principles, we ought to proceed in fencing rivulets, or brooks, which empty themselves into a larger river ; and whatever bends are given at its mouth to the two lines of embankment, they should always be made less acute than those of the natural brook; at the same time an opportunity should be taken, of reducing the angle of this transverse brook, or, in other words, of condu6ting it with a more gentle flexion into the main river.

4. Particular care should be taken, to cover the outside of the dyke with compact pieces of turf, or green sods, closely united. For if it admits the water, there is great danger of drenching the interior and more porous part of the wall, while the statical pressure of this fluid body tends to burst the bank on the land side and thus the labour of months or years may be suddenly destroyed. Hence too great attention ca. not be bestowed on making and keeping it perfectly tight; so that the whole be one continued fine turf, and even bare-spot must without delay be care-ftrfly covered with firm and fresh sods : nor should the rat and mice-holes be neglected.

lastly, it deserves to be remarked, that a dry earthen bank, not firmly conjoined by grass-roots, will scarcely maintain itself the pressure of the water, slope of forty-five degrees; while a canal conveying a moderate stream cannot be supported, even with such a declivity. Those banks, however, the base of which is as four to three of their height, will in a moist soil: and this is not only the slope usually given them, but also observed in the spontaneous operations of Nature, in the channels which she forms in conducting rills and rivulets through the higher and steeper grounds-. This natural form possesses both mechanical and mathematical properties which justly claim the admiration of those who adopt her beneficent hints and maxims.