zoo DISCOVERY REPORTS 



at the surface at Arica, where it was still recognizable as a distinct layer; elsewhere 

 upwelling occurred solely within subtropical water. Thus the Antarctic intermediate 

 water was never touched. 



An accurate idea of the depths affected by upwelling is not easily obtained, because 

 the return current, hugging the coast in the upwelling depths, interferes with the normal 

 trend of isotherms and isohalines. The coastal character of the return current is sug- 

 gested not only by our sections, but by the fact that no evidence of it can be found in 

 the section representing the ' Carnegie's' oceanic stations 60-70. Its shape in section is 

 roughly that of an elongated isosceles triangle lying on its side. The base is flattened 

 against the coast, the apex projects into the open ocean between the sub-Antarctic layer 

 above and the Antarctic intermediate layer beneath. Thus its upper margin slopes up- 

 wards towards the shore. This upward slope might be accounted for partly by up- 

 welling, but the possibility of a centrifugal effect as the current pressing to the left, 

 presses against the coast on its southward flow should not perhaps be excluded. At 

 Pichidanque Bay, where upwelling was minimal, the upper margin of the return current 

 sloped inwards and upwards from a depth of about 160 m. out to sea to less than 50 m. 

 inshore. Where upwelling was more vigorous the upper margin had a steeper slope. 

 Thus the trend of isohalines in the Figures illustrated on pp. 138 169 while reflecting 

 the influence of upwelling, is partly determined by the structure of the layers. 



In view of the interest taken in the depth affected by upwelling, it will be appropriate 

 to estimate the depths likely to be involved. On the evidence of isotherms, Figs. 

 19-41 suggest that upwelling is usually restricted to the upper 200 m., but ofT Callao 

 in August, and off Caldera, depths of 280 and 320 m. appear to be disturbed. The 

 evidence of salinity (Figs. 18-50 on pp. 138-43 and 164-9), which is perhaps more 

 reliable, suggests that depths exceeding 200 m. are very rarely disturbed. In Table XVII 

 a minimum and maximum estimate of the depth affected by upwelling at each locality 

 have been listed : and the estimates based on salinity have been plotted as curves on the 

 salinity section in Fig. 42. The minimum depth from which water wells up may be 

 taken as the depth offshore at which salinity values are found to correspond with the 

 surface inshore salinity. But as alteration of its salinity is inevitable in upwelling water 

 through its admixture with the surrounding water layers, the depth from which it 

 arises must be greater than that represented by the minimum chosen. At localities 

 where mixture may have been extensive, a better guide to the upwelling depth might be 

 furnished by inspection of the isohalines in Figs. 18-50. The maximum estimates in 

 Table XVII have been obtained by this method and are plotted in Fig. 42 as a broken 

 line. 



In five of the localities — Cape Carranza, Antofagasta, Arica, San Juan and the 

 Guanape Islands — the minimum and maximum depths so estimated correspond with 

 fair precision within 20 m. or so of one another. At these localities the upwelling process 

 may be supposed to be at its height and to be attended by comparatively little mixing. 

 On the other hand, at Pichidanque Bay, Caldera, Callao, the Lobos Islands and Punta 

 Aguja, where discrepancy is shown between these estimates of minimum and maximum 



