accompanying or immediately following great battles is not unlike that which might have been expected in the course of natural events. Bearing in mind the fact, already stated, that throughout large areas rain occurs on an average once in three or four days, and also the subjective fact that rain associated with July 4 celebrations or with battles would doubtless not have been remembered had it not been for such associations, the hypothesis appears to have no foundation. In 1892 the U. S. Government disproved the idea by experiments in which violent explosions of dynamite were produced within clouds by means of kites and balloons, with no rain following as a direct or even as an indirect result. The practise, still followed in various European countries, of attempting to prevent hail by bombarding approaching clouds or of projecting vortex rings of smoke upward, also is without scientific basis. The relatively feeble convectional currents resulting from these artificial attempts to influence the weather are too meager to have any appreciable effect upon the massive convection accompanying storms and are wholly inadequate to influence precipitation.
It is often maintained that cold waves are produced by a descent of cold air from aloft. While it is true that the air aloft is colder than that at the ground, and that up to a height of about six miles there is a more or less uniform decrease of temperature with increase of height, cold waves owe their origin to a number of factors. Nearly all cold waves of the United States occur in the area forming the rear of a passing cyclone and the front of an approaching anticyclone. During the winter half-year this region is characterized by relatively strong northerly or northwesterly winds, clearing skies, decreasing humidity, and the conspicuous fall in temperature. There is a distinct gyratory movement in large disks of air, clockwise, outward from the center, and to a slight extent descending, in the anticyclone, while it is counter-clockwise, inward toward the center, and to some degree ascending in the cyclone. The sharp fall in temperature forming the cold wave is caused primarily by the horizontal transportation of huge masses of cold air from the cold continental interior, and is heightened by the increased radiation from the ground through clear, dry air thus brought in. Vertical currents are probably only of secondary importance in this connection.
In comparing the climates of different places too much stress is generally laid upon mean, and not enough upon extreme conditions of the weather. For example, the average annual temperature, often the only climatological fact quoted in the description of a place, may be very deceptive. Based upon the records of 33 years, the mean annual temperatures of Washington, D. C, and San Francisco, Cal., are practically the same, being 54.7° F. and 54.9° F., respectively. The climates of the two cities are greatly unlike, however. Washington has a semi-continental climate, with daily maximum temperatures in sum-
