The saturation point is often incorrectly stated by omitting the contributing areas. Thus, it is said that California, in her irrigation lands, can support 500 per mile, but this is only one part of the real area needed, for the vast mountain forests from which water is obtained must forever remain unsettled. Thus, while California has now but ten per square mile, and the whole coast only four, the chances for much more increase are not so great as is often stated. Likewise the crowding of Egyptians in the little Nile strip, depends on the enormous watershed to the South. Prof. E. W. Hilgard, University of California, makes this very mistake*  when he states that it required but ten to twenty acres to a family in irrigated California colonies as compared with the forty to 160 acres in humid regions. It was likewise erroneous to state that certain Arizona and New Mexico Indians could exist in dense masses, for very recently white men have damned up the streams and taken the water formerly used by the Indians, who are now said to be decreasing in numbers because they cannot irrigate their crops.

The saturation point for population closely corresponds to the mean annual rainfall. That is, the more rain there is, the more grass and grain, and therefore more flocks and herds for man to eat. Thus, very little wheat is grown in the United States in Western lands having less than twenty inches of rain. In England and Scotland wheat grows only where the rainfall is more than thirty inches. In South Australia, Sir Charles Todd has shown that there is a very close relationship between the mean annual rainfall and the mean number of bushels of wheat per acre. It is matter of common knowledge that the annual variations in our crops are proportional to the annual fluctuations in the amount of rain.

* Draper: Intellectual Development of Europe, *  North American Review, Sept., 1902.

Likewise the number of sheep depend upon the grass and that upon the rain. In Australia, Mr. J. T. Wills has found that where they have less than ten inches of rain the land is worthless, unless irrigated. If there is ten inches, they can raise eight or nine sheep per square mile; if there is twenty inches of rain, they can raise 640. In Buenos Ayres with thirty-four inches, they raise 2,560. When they overstock in wet years there will, of course, be insufficient food in dry years, and it is possible to calculate ahead how many sheep will die in Australia for every inch deficiency of rain.

Hence, a map of mean annual rainfall, such as those made by Dr. A. J. Herbertson, Lecturer on Regional Geography, of Oxford University, is a very fair map of the density of population. There are minor differences, of course, due to the fact that some places of heavy rainfall are too mountainous, or are too hot and light for the higher races to survive and build up a civilization with its higher saturation point. Herbertson shows that the amount of moisture in the air diminishes with the temperature and therefore with the elevation. Hence, a map of the United States, shaded to represent elevation, generally approximates a map shaded to represent density of population. It seems as though population, like a real fluid, settles in the low lands. For instance, the 100th meridian of longitude separates arid from wet regions; west of it are dry elevated plains and a density of population less than two per mile; east of it is a narrow strip, lower and with more rain, but with two to six people per mile; then another lower zone with six to eighteen per mile; then on the ninety-eighth meridian begins the real population of eighteen to forty-eight per mile, increasing as we go east, until it reaches ninety and over in the area around the Great Lakes as far east as Massachusetts and as far south as Kentucky.*

L. W. Dallas, the English statistician, proved*  that the population of India depends directly upon the rainfall, being checked in its increase or actually decreasing in years of drouth. It merely shows how intensely sensitive population must be to the food supply, for when there is less to eat some must die, and if nothing to eat, all must die. What Dallas proves for India is a universal law, common to all countries, but only more evident in India, where there are so many people. We have essentially the same cycles of maxima and minima of rainfall as in India. A writer in Popular Science Monthly some years ago proved that all our financial panics and periods of industrial distress have followed the minima in the cycles of rain, and are wholly disconnected from the particular political policy the nation may have adopted. Panics were simply due to the fact that the country did not produce as much wealth in the dry years, and there was distress in place of the famines which trouble other countries.

Soil Exhaustion

There is a counteracting factor to food production to which the late Prof. N. S. Shaler called attention in a very able geological paper in The International Quarterly, May, 1905. He showed that the roots of wild vegetation hold the soil and prevent it being wasted by rains into the sea, but when man clears a field, it is subjected to so much washing that it loses in one heavy rain as much as it would ordinarily lose in several centuries. Consequently the food production is lessened and the density of that population must diminish until the land becomes feral again and recuperates. He says: "There is no basis for an accurate reckoning, but it seems likely from several local estimates that the average loss of tillage value of the region about the Mediterranean exceeds one-third of what it was originally. In sundry parts of the United States, especially in the hilly country of Virginia and Kentucky, the depth and fertility of the soil has in about one hundred and fifty years been shorn away in like great measure. Except in a few regions, as in England and Belgium, where the declivities are prevailingly gentle, it may be said that the tilled land of the world exhibits a steadfast reduction in those features which give it value to man. Even when the substance of the soil remains in unimpaired thickness, as in the so-called prairie lands of the Mississippi Valley, the progressive decrease on the average returns to cropping shows that the impoverishment is steadfastly going on." Man, then, by the very processes he starts, lessens his own food supply and lessens his own density of population, just as every other animal does when it exists in such great numbers as to destroy its food supply. Arid lands contain more plant food than humid ones, as the latter are being constantly washed out by rains, and this fully explains the enormous crops obtained by irrigating the deserts, but there is evidence that even they are washed out in time, unless the soil is constantly renewed, as in Egypt.

* Census Atlas.

*  Quarterly Journal of the London Meteorological Society, 1905.