cerning the number of sheep which can be pastured per square mile with different rainfalls.
Such investigations have not been made for the United States, but the data indicate clearly the enormous variations in the food-supply, both vegetable and animal, which attend variations in rainfall; and they suggest how these variations must affect the producer, the transporter, the merchant and the consumer. Hence it is easy to imagine the great influence which variations in rainfall may have on commerce and through this on politics.
The accompanying table gives the variations in the amount of rainfall in the Ohio Valley and in the Mississippi Valley which lie about the center of the food-producing area in the United States and include a large part of this area. The data are derived from tables prepared by Professor A. J. Henry, of the United States Weather Bureau, and published by the Bureau as Bulletin 'D,' entitled 'Eainfall of the United States,' Washington, 1897. The average rainfall for each district was made up from a number of stations in the district, the same stations being used so far as the records would permit throughout the period 1830 to 1896. The sharp, irregular fluctuations which characterize the rainfall were toned down by Professor Henry by taking the means of several successive years. This process is called 'smoothing' and it renders more evident the long-period oscillations. The average rainfall for each district was obtained, and the departures of the rainfall for each year from tins mean are given in the accompanying table. The plus sign indicates that the rainfall for the given year was above the mean, and the minus sign that it was below. The figures give the amounts in inches and tenths of inches. The figures for the Mississippi Valley from 1848 to 1857 are derived from the observations at one station only.
The table also gives the departures from the mean value of the level of the water in Lake Michigan. These data have been carefully collected by the engineers on the lakes and were kindly furnished by General John M. Wilson, Chief of Engineers, U. S. Army. The figures show, in feet, the departures of the annual means from the general average. The lake may be regarded as an enormous rain-gauge. When the rainfall is in excess, the water level rises above the mean; but
- These facts are largely derived from Hann's 'Climatology,' a standard work on climate. (See Ward's English Translation.)
- Professor Henry also gives the rainfall for New England. Although the oscillations run roughly parallel to those in the interior valleys the data are not reproduced here, (1) because Mr. E. B. Weston has shown that the early measurements are probably deficient on account of the methods of measuring the snowfall; (2) because New England has largely ceased to be an agricultural region.