About 42 inches of rain may be taken as a fair general average of the rain-fall in the United States. If this supplies as much ammonia to the soil as 3 cwt. of Peruvian guano to the acre, which is considered a liberal manuring, and which is valuable principally for its ammonia, we at once see the importance of retaining the rain-water long enough upon our fields, at least, to rob it of its treasures. But rain-water has a farther value than has yet been suggested: "Rain-water always contains in solution, air, carbonic acid, and ammonia. The two first ingredients are among the most powerful disintegrators of a soil. The oxygen of the air, and the carbonic acid being both in a highly condensed form, by being dissolved, possess very powerful affinities for the ingredients of the soil. The oxygen attacks and oxydizes the iron; the carbonic acid seizing the lime and potash and other alkaline ingredients of the soil, produces a further disintegration, and renders available the locked-up ingredients of this magazine of nutriment. Before these can be used by plants, they must be rendered soluble; and this is only affected by the free and renewed access of rain and air. The ready passage of both of these, therefore, enables the soil to yield up its concealed nutriment." We see, then, that the rains of heaven bring us not only water, but food for our plants, and that, while we would remove by proper drainage the surplus moisture, we should take care to first conduct it through the soil far enough to fulfill its mission of fertility. We cannot suppose that all rain-water brings to our fields precisely the same proportion of the elements of fertility, because the foreign properties with which it is charged, must continually vary with the condition of the atmosphere through which it falls, whether it be the thick and murky cloud which overhangs the coal-burning city, or the transparent ether of the mountain tops. We may see, too, by the tables, that the quantity of rain that falls, varies much, not only with the varying seasons of the year, and with the different seasons of different years, but with the distance from the equator, the diversity of mountain and river, and lake and wood, and especially with locality as to the ocean. Yet the average results of nature's operations through a series of years, are startlingly constant and uniform, and we may deduce from tables of rain-falls, as from bills of mortality and tables of longevity, conclusions almost as reliable as from mathematical premises. The quantity of rain is generally increased by the locality of mountain ranges. "Thus, at the Edinburgh Water Company's works, on the Pentland Hills, there fell in 1849, nearly twice as much rain as at Edinburgh, although the distance between the two places is only seven miles."