little improved. Records as to the amount of water daily evaporated from a plate or dish on shore afford us no means of judging as to what is going on even in the same latitude at sea. Sea-water is salt, and does not throw off its vapour as freely as fresh water. Moreover, the wind that blows over the evaporating dish on shore is often dry and fresh. It comes from the mountains, or over the plains where it found little or no water to drink up; therefore it reaches the observer's dish as thirsty wind, and drinks up vapour from it greedily. Now had the same dish been placed on the sea, the air would come to it over the water, drinking as it comes, and arriving already quite or nearly saturated with moisture; consequently, the observations of the amount of evaporation on shore give no idea of it at sea.
281. Rivers are gauges for the amount of effective evaporation.—There is no physical question of the day which is more worthy of attention than the amount of effective evaporation that is daily going on in the sea. By effective I mean the amount of water that, in the shape of vapour, is daily transferred from the sea to the land. The volume discharged by the rivers into the sea expresses (§ 270) that quantity; and it may be ascertained with considerable accuracy by gauging the other great rivers as I procured the Mississippi to be ganged at Memphis in 1849.
282. Importance of rain and river gauges.—The monsoons supply rains to feed the rivers of India, as the north-east and south-east trade-winds of the Atlantic supply rains to feed the rivers of Central and South America. Now rain-gauges which will give us the mean annual rain-fall on these water-sheds, and river-gauges which would give us the mean annual discharge of the principal water-courses, would afford data for an excellent determination as to the amount of evaporation from some parts of the ocean at least, especially for the trade-wind belts of the Atlantic and the monsoon region of the Indian Ocean. All the rain which the monsoons of India deliver to the land the rivers of India return to the sea. And if, in measuring this for the whole of India, our gauges should lead us into a probable error, amounting in volume to half the discharge of the Mississippi River, it would not make a difference in the computed rate of the effective daily evaporation from the North Indian Ocean exceeding the one two-thousandth part of an inch (0.002 in.).
283. Hypsometry in the North Atlantic peculiar.—That part of the extra-tropical North Atlantic under consideration is peculiar
