196 DISCOVERY REPORTS 



more far-reaching than this, since the low temperatures of recently upwelled water are 

 continually dissipated by admixture with its surroundings. 



Inshore current, wherever it occurs, must be a factor of importance in carrying low 

 temperatures northward: but the evidence examined on pp. 125 to 133 and on p. igo 

 shows that northerly current is not only irregular but that the inshore water is more 

 often than not setting towards the west ; and that drift in the open ocean is predominantly 

 west. Low temperatures near the coast would thus appear to be continually borne off 

 towards the west, and to be replaced by further upwelling. In this respect, the Peru 

 Current differs from a current like the Labrador Current which, converging with the 

 American coast, is probably able to carry water particles and a low temperature for 

 almost the entire length of its course. 



COASTAL UPWELLING 



ESTIMATION OF UPWELLING 



In order to compare upwelling in different localities, it is first necessary to find a 

 method of estimating its degree. The method most used has consisted in noting the 

 reduction of surface temperature as compared with an arbitrary standard. McEwen 

 ( 1 9 1 2), in his investigations of upwelling off California, selects as his standard the thermal 

 normal for the latitude, which he assumes (p. 261) "is the same as the actual tempera- 

 ture at a point in mid-ocean having the same latitude". Thus the difference between 

 the observed temperature and the normal temperature is assumed to be due entirely to 

 the mixture of cold water from the adjacent ocean bottom with the surface water. The 

 disadvantage of adopting this as the thermal normal for the latitude has in part been 

 shown by McEwen himself in a remark on p. 244 to the effect that "The question of the 

 distribution of temperatures in the sea is so intimately connected with that of the 

 character of its currents that it is practically impossible to separate them entirely". 



Schott (1931) uses as his standard the mean surface temperature at a distance of 100 

 miles offshore. This has the advantage that both the temperature here and close inshore 

 will be subject to the same local major variations : but it is open to the criticism that at 

 this distance from shore the temperatures are influenced by upw'elling off Peru where the 

 surface isotherms are far apart to a greater extent than off Chile where isotherms tend to 

 hug the coast. The effect is an apparent reduction in the amount of upwelling off Peru, 

 because the contrast between inshore and offshore temperatures is reduced. This is 

 seen by comparing the surface temperature inshore with that at 46 miles offshore both 

 at Caldera and San Juan. Off Caldera the difference was 3-36° C, off San Juan only 

 1-46° C; yet more upwelling seemed in progress off San Juan, for here a greater 

 volume of cool water was present inshore, and the difference between the temperatures 

 inshore and at 152 miles offshore was 5-0° C.^ 



^ It is interesting that this effect happens to be largely neutralized by another factor working in the 

 opposite direction — namely the greater difference, in the tropics, between surface and subsurface tempera- 

 tures than in higher latitudes. The net effect is to give Schott's curve a slope tolerably close to the standard 

 we have chosen. 



