10,000 feet high skirting the Gulf and South Atlantic coasts. Of course this is impossible, but if nature had erected it there would be no question about floods in the Ohio and the Mississippi rivers and their tributaries, for there would be neither rivers nor tributaries; just as on the Pacific coast, the rain would fall on the ocean side of the mountains, and the world's greatest granary would be a barren waste.

Prof. Frank H. Bigelow, on page 17 of A Manual for Observers in Climatology and Evaporation, says:

All this distribution of the general circulating currents, and the consequent precipitation, would occur whether there were forests or not growing on the land masses. It may be proper to say that the forests follow the precipitation and do not precede it.

There can be no question but that the action of the sun on the waters of the Gulf of Mexico and the adjacent ocean heavily charge the air with water vapor, and that this vaporous atmosphere is carried inland by the circulation of the air in such storms as are described in a preceding paragraph and illustrated on chart A, and that the effect is shown on chart B in the form of heavy precipitation in the region of the Gulf, which gradually shades away toward the Rocky Mountains.

It is therefore apparent that the precipitation that causes floods in the eastern half of the United States is from the aqueous vapor that is raised up from the vast waters to the south and southeast of our continent, and that the supply is inexhaustible. Our rainfall, then, is the result of such fundamentally great causes as not to be appreciably affected by the planting or cutting away of forests, or by any of the operations of man in changing the character of the surface covering of the continent, although to statistically and positively settle the question beyond the possibility of argument it would be necessary to have scientific data of temperature, rainfall, and the height of rivers, beginning at the first settlement of the continent and continuing through to the present time. Such records, of course, are not in existence. But the fundamental fact that the precipitation of the United States is due to the great hemispherical circulation of the air, and to the relation of the great bodies of water to land, and the direction of the vaporous-bearing currents, and the trend of mountain systems is something that can be positively shown.

Mr. Willis further says, in the same issue of Conservation, page 265, that:

The mountains are wet because they are high, and they are heavily forested because they are wet. But there is also a reciprocal action of the forests on the wetness, for the radiation from the dark-green expanse is comparatively uniform and promotes frequent and steady rains. Were the mountains bare they would, like the bared sierras of Spain, receive occasional but violent downpours and send down excessive and disastrous floods, even more disastrous than now. * * For in so far as we

*

clothe the surface with green crops we lower the temperature of the rising air and favor precipitation on the verdure-covered plain.

It would be difficult to either confirm or disprove this statement of Mr. Willis. Certain it is that the rain is precipitated largely from air masses that exist at a considerable distance from the surface of the earth, and that the influence that Mr. Willis describes is in a thin stratum of air close to the earth. Rarely is this stratum saturated, even during the fall of rain. If, then, the processes that he describes do not bring the air to the saturation point, and if the

precipitation occurs in the regions above those affected by these local surface conditions, I am unable to see how the rain can be either increased or decreased in its amount. Certain it is that most of the leading meteorologists of the world are of the opinion that the rainfall on continents is caused by the fundamentally great operations of nature as described above.

EROSION.

Another effect of deforestation, that of erosion, is of importance, but of unequal importance in different sections. In level countries it makes but little difference in this particular whether the ground is waste, cultivated, o densely forested, while in hilly or mountainous sections the result is different. When the soil becomes well sodded with grass, erosion is little worse in fields than in the woods, but usually the fields are cultivated from time to time, and occasions come when the best of care and cultivation can not prevent the formation of bad gullies that injure both the gullied fields and those of the lower grounds that are overflowed.

Of course, though, a field with an occasional wash yields more food material than the same area covered by a forest of any kind, so that only in exceptional cases-those in which erosion would probably be unavoidable and ruinous-is this a sufficient argument against clearing away the woods and the planting of crops in their stead, for the time is come when we should not only increase the yield per acre by wise rotation of crops on cultivated ground, but clear up and seed to wheat, corn, grass, and fruits millions of acres that now lie idle under brush or forest. In other words, every acre that will grow food for the people, and thereby reduce its cost and furnish sustenance for our increasing population and the teeming millions that are on the way to these shores, should be so employed; the remainder should grow timber that should be protected in its growth. Man and beast love the cooling shade, and the eye is pleased by the beauty of the wooded landscape. Therefore begin with the children and teach them to plant trees along the highways and byways and on the barren spots that will not produce food. Thus may we approach this problem rationally, with the object of gaining the greatest good for the greatest number for the longest period of time.

RATIO OF THE FORESTED AREA, OR MOUNTAIN WATERSHEDS, TO THE TOTAL WATERSHED.

I am of the opinion that not enough consideration has been given to the relative magnitude of the areas involved in the creation of floods. A flood in any given stream is usually caused by the precipitation over its entire watershed or over those of the major tributaries and is affected but comparatively little in a region like that of the Ohio basin by the precipitation over the extreme upper reaches, usually the forested area, or any other area that could be reforested without seriously encroaching upon the rich alluvial plains.

A critical examination of Chart C, which shows the entire river system of the Ohio basin and gives the exact limits of its boundaries, and which also indicates the elevations, shows what a comparatively small area in relation to the total catchment basin lies at