282 DISCOVERY REPORTS 



water movements in the Benguela current, and is a diagrammatic and somewhat idealized version of 

 them, but we feel that it is a fair representation of the circulation which is probably common to the 

 major upwelling regions (but the mirror image of those in the northern hemisphere). 



A striking feature of the Peru current is the very large apparent size of the anti-cyclonic eddies, 

 which far exceed the dimensions of those found in the Benguela and California currents. There is 

 some indication that these large eddies off Peru may conceivably have within them smaller eddies, but 

 in themselves they appear to be peculiar to the Peruvian coast, and one can only assume that it must 

 be constancy in the driving forces which enables them to persist and achieve such large dimensions. 



Fig. 96. A perspective diagram showing an idealized picture of the principal horizontal and vertical water-movements in the 

 process of upwelling. The isosteres are represented by the thin lines on the 'cut' faces of the water masses. The shaded 

 sinuous line on the sea-surface represents the continuation of the boundary convection between the upwelled and oceanic 

 surface-waters. The name of the deep compensation current has been abbreviated. 



The Peru current itself, of course, is very much more extensive than any of the other upwelling 

 regions, for it runs along the South American coast for nearly 2000 miles. Furthermore it is unique 

 in having within its range the subtropical convergence, and consequently involves subantarctic 

 waters in addition to subtropical waters. By comparison the subtropical convergence in the South 

 Atlantic lies well to the south of South Africa (Deacon, 1933). In spite of its great size, however, the 

 Peru current has a slightly smaller negative temperature anomaly ( — 8° C.) than the Benguela 

 ( — 9°C), and both exceed by far that off California ( — 3° C.) and Morocco ( — 3° C), (Dietrich, 1950). 



Upwelling affects similar depths in all of these regions. Gunther (1936) quotes figures of 40-360 m. 

 with a mean at 133 m. in the Peru current. Sverdrup (1941) states that in the California current 

 upwelling was evidently confined to the upper 200 m., and in the Benguela current we have placed it 

 at depths between 200 and 300 m. This removal of water from the 200-300-m. level appears to be 

 accompanied in most cases by a subsurface compensation current. We have described this compensation 



