acceleration just sufficient to overcome their resistances, requires that the acceleration should be not only the small component of gravity acting on the barometric gradient, but the much smaller resultant of this component acting with the deflective force arising from the motion of the wind itself. The course adopted by the established interchanging circulation between the equator and the poles consists for the most part of a great circumpolar whirl from west to east; and the deflective forces here in play reduce the polar high pressures to low pressures. A reactionary relation therefore exists between the winds and the pressures, by which the distribution of pressures according to temperature alone is greatly modified. Instead of finding high pressure at the cold poles, a low pressure is produced there by the great circumpolar whirl of the general winds, and the air thus held away from the poles accumulates around the tropical belts of high pressure, of which Ferrel had first learned from Maury's book. The absence of northeast return currents (in this hemisphere), except in the trade-wind belt, is as important a feature of Ferrel's theory as the reversal of polar high pressure into low pressure. Maury's erroneous explanation of the winds gained great acceptance from the attractive style in which his book was written; but it is time that his explanation should be abandoned even in elementary teaching, and replaced by more serious views, less easily acquired but of more permanent value.
Ferrel's theory of the winds not only explains the general distribution of atmospheric pressure over the world, as no other theory can do; it introduces broad correlations among many phenomena in meteorology, greatly to the advance of the science. The legitimate analogies that may be drawn between the great circumpolar whirl of the terrestrial winds, the smaller whirls of tropical cyclones, and the concentrated whirling of tornadoes show the unity of action of the convectional processes in the moist atmosphere of a rotating planet. In earlier years, meteorology consisted chiefly of rules for observation and statistical study. The broad generalizations taught by Ferrel raise the science from this simple inductive condition and complete the philosophical rounding of its parts.
Ferrel was not an observer, but he carefully based his studies on well-ascertained facts. He was not an experimenter, but he followed the results obtained by the best physicists. He was a reasoner, able to employ the stronger methods of mathematical analysis. He was sincere and judicial, single-minded and simple-hearted. No one criticised his results more carefully or deliberately than he did himself. He was indifferent to popularity, and took little trouble to enforce his views on the world. He lived a quiet life, more with books than with men, although the few to
