228 THE PHYSICAL GEOGRAPHY OF THE SEA. 



precipitation near the tropical calms is nevertheless sufficient to 

 mark the seasons ; for whenever these calm zones, as they go from 

 north to south with the sun, leave a given parallel, the rainy sea- 

 son of that parallel, if it be in winter, is said to commence. Hence 

 we may explain the rainy season in Chili at the south, and in Cal- 

 ifornia at the north. 



491. The Westerly Winds. — To complete the physical ex- 

 amination of the earth's atmosphere, which we have supposed an 

 astronomer in one of the planets to have undertaken according to 

 the facts developed by the Wind and Current C?ia7'ts, it remains 

 for him to turn his telescope upon the southwest passage winds of 

 the northern hemisphere, pursue them into the arctic regions, and 

 see theoretically how they get there, and, being there, what be- 

 comes of them. 



From the parallel of 40° up toward the north pole, the prevail- 

 ing winds, as already remarked, are the southwest passage winds 

 (Plate VIII. ), or, as they are more generally called by mariners, 

 the " westerly" winds ; these, in the Atlantic, prevail over the 

 "easterly" winds in the ratio of about two to one. 



Now if we suppose, and such is probably the case, these '' west- 

 erly" winds to convey in two days a greater volume of atmosphere 

 toward the arctic circle than those "easterly" winds can bring 

 back in one, we establish the necessity for an upper current by 

 which this difference may be returned to the tropical calms of our 

 hemisphere (§ 109). Therefore there must be some place in the 

 polar regions (§ 112) at which these southwest winds cease to go 

 north, and from which they commence their return to the south, 

 and this locality must be in a region peculiarly liable to calms. 

 It is another atmospherical node in which the motion of the air is 

 upward, w^ith a decrease of barometric pressure. It is marked P, 

 Plate I. 



If we now return to the calm belt of the northern tropic, and 

 trace theoretically a portion of air that, in its circuit, shall fairly 

 represent the average course of these southwest passage winds, we 

 shall see (§ 113) that it approaches the pole in a loxodromic curve ; 

 that as it approaches the pole, it acquires, from the spiral convolu- 

 tions of this curve which represents its path, a w^hirling motion, in 

 a direction contrary to that of the hands of a watch ; and that the 

 portion of atmosphere whose path we are following w^ould grad- 



