Oceanic Circulation and Temperature 

 ward to the surface, thus continually replacing the 

 surface water that is forced away from the coast 

 by the wind (figure 17). In addition, owing to the 

 fall in pressure as the coast is approached, and 

 because of the deflecting force due to the earth's 

 rotation, there must be a flow parallel to the coast 

 at all depths in the general direction of the wind. 



Now if, instead of being vertical, the coast is 

 inclined, the velocity of upwelling, and conse- 

 quently the cooling effect on the in-shore water, 

 will be less marked. Also, as the average depth 

 of the water decreases below that of twice the 

 depth of the "wind current," the effect of the wind 

 will be less modified by the deflecting force, and 

 the motion will take place more in accordance 

 with the commonly accepted laws. 



Winds of the North Pacific. — Numerous ob- 

 servations have shown that the winds over a wide 

 belt of the Pacific along the west coast of North 

 America are usually directed to the southeast. More- 

 over, these winds which form a part of a vast 

 anti-cyclonic or clockwise circulation of air over 

 the North Pacific are subject to an annual variation 

 corresponding to the annual fluctuation of the dif- 

 ference between the air pressure over the conti- 

 nents and over the oceans. This difference is 

 caused by the familiar fact that bodies of land 

 acquire different temperatures from bodies of water 

 under similar conditions. Thus in summer, when 

 the land is relatively warmer than the ocean sur- 

 face, there is a relatively high air pressure over 

 the latter which results in a correspondingly greater 

 wind velocity along the coast. On the other hand, 

 in winter the pressure difference and consequently 

 the wind velocity decreases owing to the relatively 

 low temperature of the land. Hence an upwelling of 

 bottom water would be expected along the Pacific 

 Coast, and the rate of this upward flow would vary 

 from a maximum in August to a minimum in Feb- 

 ruary. This accounts qualitively for the peculiar 

 temperature distribution previously described. Also 

 the observed surface drift along the coast agrees 

 well with what would be expected from the theory. 



As an additional test of the theory I obtained 

 a mathematical relation between the wind velocity, 

 the surface temperature of the in-shore water, and 

 the normal value for the latitude. A very satis- 

 factory agreement was found between the tem- 

 peratures computed from the above relation and 

 the observed monthly in-shore temperatures, which 

 at times are as much as 14* below the normal. 



138 



