CENTRES OF UPWELLING 215 



The variations from this order in September 1927 and November 1929 are slight, 

 amounting principally to a state of vigorous upwelling off Puerto Chicama in 1929. 



Our experience was much the same ; we found very much more active upwelling off 

 some parts of the coast than off others, and the regions noted by Schott can be dis- 

 tinguished both in the curves in Fig. 66 and in the chart of surface temperatures in 

 Figs. 16 and 17. Thus in region I from lat. 30 to 25° S, vigorous upwelling off Caldera 

 lowers the mean temperatures in lat. 27-28° S as far as 10 miles offshore. The second 

 focus of upwelling occurs off Antofagasta in region II in lat. 25-18° S. In region III 

 (lat. 18-8° S) the upwelling is at San Juan, but in region IV upwelling off the Lobos 

 Islands was not quite so vigorous. Points of exceptional warmth were found south of 

 Antofagasta in region I (Fig. 9) and in moderation off Callao in region III, but at no 

 other localities close inshore. This agreement in the localities of major upwelling makes 

 it very likely that these centres of upwelling are more or less permanent. As a natural 

 consequence follow differences in the breadth of the cool zone and in the gradient 

 between inshore and offshore temperatures (see Figs. 29 and 30). 



Particular interest attaches to these centres since they are found to correspond with 

 the anticyclonic swirls suggested on p. 192 and Fig. 63. The two swirls off Peru are 

 illustrated diagrammatically in Fig. 66 by arrows which curve away from the graphs in 

 the divergence regions of strong upwelling, but curve towards them in the warm con- 

 vergence regions. The very similar nature of the upwelling centres off Chile and their 

 comparative permanency makes the existence of similar swirls, though weaker and 

 probably less pronounced, to be expected off the Chilean coast. The probable positions 

 of the hypothetical swirls are also marked in the figure. 



IRREGULARITIES OF THE CURRENT 



No account of the current would be complete without some reference to its extra- 

 ordinary variability. In almost every particular, drift, temperature, breadth, colour, 

 etc., it is fraught with irregularity. In their persistent references to the uniformity of 

 the conditions on the west coast, earlier workers are apt to mislead. Thus Murphy 

 (1925) says: "Extraordinary uniformity is, after all, the outstanding oceanic feature of 

 the Peruvian littoral." And Schott (193 1) writes: "On the large scale the temperature 

 too is uniform, for in the water it varies little with the latitude and according to no 

 obvious law, and little with the season." He is accounting for the slight differences 

 between mean temperatures in low and high latitudes within the coastal region and 

 the equable climate that results: yet later (1932) he writes: " . . .therefore the quantity 

 of upwelling water also varies from place to place. In consequence of this situation the 

 surface temperature . . . are depressed in complicated and irregular variations along the 

 coast." These variations are a conspicuous feature of the current and may be due very 

 largely to the swirls above described ; we may illustrate their nature with a few selected 

 examples. 



Changes of temperature along the coast sometimes occur abruptly within short dis- 

 tances, and the following record as the ship left Coquimbo Harbour on a course of 353° 



