GENERAL CIRCULATION OF THE PACIFIC OCEAN 547 



former theories which have been mostly propounded with respect to the 

 mass transport. 



Since the problem is three-dimensional, the numerical computation 

 is rather tedious. For this reason the author has not yet been able to 

 finish the computation below the level % = SD^. Still we should be 

 able to expect some important conclusions from w^hat has been com- 

 pleted thus far. 



First of all it is very interesting that the wind-driven currents exist 

 in a layer much deeper than that expected from, the classical theory of 

 Ekman (1905). According to Ekman's theoi-)', a wind-driven current is 

 confined to the surface layer about D^ thick, and we can expect prac- 

 tically no drift current at a deeper level except very close to the equator. 

 From our computation, it can be shown that the current velocity does 

 not drop as low as half the surface value even at a level oD^. If we 

 take D^ = 75 meters, this depth is 225 meters. 



This conclusion will be helpful for us to understand the fact that 

 wind-driven currents can penetrate into a layer several hundred meters 

 deep, a layer several times as deep as that expected from Ekman's theory 

 as the limit of the wind-driven currents. This implies that the motion 

 of water in most parts of the oceanic troposphere could be produced 

 by the stresses of the permanent wind system prevailing over the oceans. 

 In other words, the winds are responsible not only for the currents in 

 the skin layer of the ocean, but also for the most part of the circulation 

 in tlie oceanic troposphere. 



We have long considered that the winds are responsible only for 

 the current motion in the surface layer of about 100 meters thick. This 

 depth is nothing but the "depth of the frictional influence" defined 

 by Ekman. To explain the circulation in deeper parts of the tropo- 

 sphere, we had to assume a very strong convection current and slope 

 current. Still we had a distinct difference in the circulation patterns 

 between the troposphere and stratosphere. These circumstances have 

 made several problems very much complicated. Defant (1928) defined 

 the troposphere as the part of the ocean in which we can expect strong 

 currents due to violent turbulence and convection. Still we can have 

 violent convection in the seas of higher latitudes beyond the polar 

 fronts which are no longer defined as troposphere. The conclusion 

 that the drift currents penetrate into much deeper layers than D^ is 

 much in favor of the definition of troposphere that this is the upper 

 layer of the ocean in which strong currents are present. 



The explanation of the result that we can have a strong motion 

 even in a layer several hundred meters deep might look possible by 

 assuming slope currents which would be produced as the effect of purely 



