192 DISCOVERY REPORTS 



The continuity of the wedge from north to south cannot therefore be substantiated, 

 and the alternative, that more than one wedge is in question, calls for examination. 

 A comparison of the temperature and salinity sections off Callao and San Juan (Figs. 32, 

 33, 44 and 45) shows a similarity in the structure of the wedge off each of these 

 localities which would be hard to explain if the water were supposed to flow from one 

 to the other. The two localities are separated by some 270 miles. While strong southerly 

 current was recorded in the wedge off San Juan, none was noted off Callao, and it is 

 unlikely that in travelling this distance the temperature, salinity and general structure 

 of the wedge could survive so little altered. 



From a theoretical standpoint, also, the wedge if it experienced a flow southwards 

 from Callao to San Juan should tend to deflect to the left and to converge with the coast, 

 and after leaving Callao would not presumably swing to the right into the open ocean. 

 Moreover, the presence of the warm water close inshore at Callao and Arica, and the 

 way in which surface isotherms slope inwards towards these localities from the open 

 sea, suggest that off Peru two wedges are involved (see Fig. 63). The serial continuity 

 shown by salinity (Table VI, p. 161) from one wedge to the other, while it may be regarded 

 as supporting the notion of a single counter-current, might equally be applied to the 

 hypothesis of two counter-currents, showing them to be homologous, to be drawn from 

 the same water mass but from different latitudes. 



The two wedges are seen to lie off strips of the coast — off northern Peru and north- 

 wards of Mollendo — where northerly current and westerly set are notorious. Northerly 

 drift and westerly set in these localities were not only noted by direct observation during 

 the present survey, but could be inferred from the prevalence there of cool upwelled 

 water shown in Fig. 63. The wedges flowing south-east and converging with the coast 

 immediately to the southward of these two centres imply the existence of two large 

 anticyclonic swirls. 



According to this view, the first wedge was converging with the coast northwards of 

 Callao to compensate, in part, for the strong current off northern Peru. Corresponding 

 to this, the second wedge, some 5-10° of latitude farther south, also originating from the 

 open ocean, was converging with the coast in the Bight of Arica to compensate in part 

 for the strong current off the San Juan-Mollendo region. 



Evidence of an indirect nature suggests that the swirls may be a recurrent if not a 

 permanent feature of the coastal current. Attention has already been drawn to the fact 

 that coastal current is traditional between Eten and Punta Aguja and in the Mollendo- 

 San Juan region (Somerville, 1923); likewise westerly set (South American Pilot, 1927, 

 Part III ; Ray, 1896) ; attention has also been drawn to the frequent absence of current at 

 Callao and Arica (Buchanan, 1886; Schott, 1891). Schott (193 1) has shown that the 

 former localities were foci of strong upwelling, whereas the latter were regions of high 

 temperature in 1927 and 1929 (see pp. 214-15). In view of the conclusions reached on pp. 

 229-33 that the unusual colours of the sea met by us off Pisco, Callao and the Guanape 

 Islands were a form of aguaje and may have been a result of the convergence of the 

 wedge with the coastal water, the appearance of aguaje during winter months described 



