154 HUMPHREYS: MSTeOROI^OGlCAL PARADOXJSS 



on the other side of the equator, for there, pushed north it blows 

 west, pushed south it blows east. 



The push that causes the winds to blow is due to the existence 

 of unequal amounts of air above a given level over adjacent re- 

 gions — more at the place from which the air is pushed than at the 

 place towards which it is pushed — and this in turn, usually, is 

 due to the temperature differences, level for level, between the 

 atmosphere at the two places. Obviously there tends to be, 

 and, initially, actually is, a horizontal flow of the air (that is, 

 a wind) at each level, in the direction of the most rapid horizontal 

 decrease of presstu-e at that level. Such winds, however, fre- 

 quently last so long (hours at least) that their directions are pro- 

 foundly altered by a certain obscure factor, namely, the rotation 

 of the earth — the secret of the above paradox — which is over- 

 looked by almost everyone, and overlooked simply because its 

 effect on the shooting of a marble, the pitching of a ball, and all 

 the thousand other similar phenomena with which we are in- 

 timately familiar, is always negligible. 



It is easy to demonstrate, as may be found in many books and 

 articles, that an object moving in any horizontal direction tends so 

 strongly to turn to the right north of the equator, and to the 

 left south of it, as to exert a force, against a restraint preventing 

 such deflection, given by the equation 



/' = 2 mojv sin^p, 

 in which ;n is the mass of the object, v its speed, ^ its latitude, 

 and CO the angular velocity of the earth's rotation. 



Consider, then, the effect of applying a horizontal push of 

 constant magnitude and constant geographic direction to a mass 

 of air, m, and assume this air to be free from friction, as it very 

 nearly is when appreciably above the surface. Let m, figure i, 

 be this mass of air, initially at rest with reference to the surface 

 of the earth; let it be in the northern hemisphere, and let p 

 be the push of constant magnitude and constant direction, north. 

 Immediately the mass moves it begins to deflect from the north 

 tdwards the east, and, owing to the curvature of its path, in- 

 troduces a small centrifugal force. A little later p may be re- 

 solved, as shown, into two components, one normal and the 



