2o8 DISCOVERY REPORTS 



the circular drift of the ship from St. WS 630 to 635 on her return journey and the 

 presence of a southerly counter-current off Bahia Herradura have been regarded as 

 evidence of a coastal eddy. Less force is probably required to draw surface water for 

 considerable distances horizontally than to lift deep water even for a short way vertically ; 

 and it is possible that for a local reason surface water is available here to flow in together 

 with upwelled water as compensation for the divergence. The wind was rather stronger 

 off^ Antofagasta than elsewhere in the vicinity. A coastal inflow of warm water from the 

 north might, then, have been a factor contributing to the rise of surface temperature 

 inshore, but a factor of only limited importance. Assuming a constant velocity of 38 

 miles a day, a breadth of 2 miles, and a surface temperature of 15° C. (Table I and Appen- 

 dix IV), its heat capacity was obviously far too small materially to affect surface tem- 

 peratures over a wide area. The temperature rise from 13-93° C. on June 8 to 15-31° C. 

 on June 10 on the position of St. WS 625 at 7 miles from shore, must evidently have 

 been brought about by some other mechanism. 



The theory advanced earlier, that the shift in surface isotherms may have indicated 

 a subsidence of the cool water, deserves then to be considered. The change of surface 

 temperature simultaneously with change of wind at St. WS 630 argues a dependence of 

 water movement on wind: and the vigorous upwelling on the outward journey argues 

 strong divergence of surface water from the coast. Dependence of the latter on wind 

 may be inferred as a probability and indeed is to be expected from Ekman's theory. 

 From this the southerly wind is to be looked upon as a force capable of raising water 

 from the lower layers. And as these are heavier than the surface layers offshore, they 

 may not unreasonably be supposed to sink when the wind fails or reverses its direction. 

 That this was happening on the return journey may be gathered not only from the 

 shoreward shift of surface isotherms above noted, but also from the shoreward drift of 

 the ship during the course of St. WS 630 (Fig. 9). 



In our analysis of the conditions at Antofagasta, we are hampered by having no 

 simultaneous observations outside the immediate neighbourhood. While it cannot be 

 known whether an eddy in the north co-existed with upwelling in the south on June 8, 

 we do know that on June 10, simultaneously with evidence of eddy in the north, up- 

 welling in the south had ceased: the surface temperature at St. WS 625 in the upwelling 

 area of June 8 was 13-93° C-) ^^^ after the change of wind it had risen to 15-31° C. on 

 June 10. Thus while the ship was drifting to the northward and finding warm water off 

 Bahia Herradura, upwelling in the former upwelling area had definitely slackened and 

 the surface temperature had risen. This and the fact that changes in surface tempera- 

 ture and changes in the wind force were simultaneous, lends definite support to the 

 view that subsidence played some part in this disappearance of the low temperatures. 



Guanape Islands 

 The shift of surface isotherms with change of wind off the Guafiape Islands forms an 

 interesting contrast to the conditions off Antofagasta, since movement was here anti- 

 cyclonic and not cyclonic. Here during the journey towards the shore (towards 



