I40 DISCOVERY REPORTS 



May 28, wind was either absent or blew gently from the south-west, and no immediate 

 connection is to be supposed between this wind and water at this depth. On the other 

 hand, the association between cool water at the surface and the easterly wind at Sts. 

 WS 604 and 605 is suggestive of a direct relation. 



After completion of Sts. WS 606 and 607 in the open ocean, a strong southerly wind 

 arose which drifted the ship northwards as shown in Fig. 6. The possibility that the 

 upwelling noted in Fig. 21 may be an indirect effect of this wind is discussed on p. 210. 



30-25° S: CALDERA 



The inshore water at Caldera showed signs of active upwelling, and immediately to 

 the southward cold water was met at the surface as far out as 33 miles. Elsewhere, how- 

 ever, warm water was close to the coast and inshore winds were northerly. Irregularities 

 of surface temperature as found in this region (Fig. 17, Tables XX and Appendix IV) 

 are frequently met with where warm and cool water meet ; but while convergence of the 

 warm water with the cool inshore, and a ragged boundary between the two, may be 

 consistent with a recent change of inshore wind to the north, the question whether 

 the cool water was upwelling at the same time has not been decided. 



The changeable condition met with on the surface, also noted in reference to wind 

 and drift in Fig. 8, extends to the subsurface layers and is illustrated in the temperature 

 section off Caldera in which almost phenomenal irregularity is shown (Fig. 23). An eddy- 

 like movement beneath the surface at 10-15 miles offshore, Sts. WS 616 and 618 may 

 indicate a sinking of newly mixed water. Witte's (19 10) account of the theory under- 

 lying this phenomenon may be quoted :^ 



When on the boundary of an ocean current, warm water of high salinity is brought into contact 

 with colder water of less saline character, but having approximately the same specific gravity, then 

 the resulting mixture will, as may easily be proved by the Knudsen tables, be of greater density than 

 either of its component parts. It will consequently sink down, giving rise to the peculiar phe- 

 nomenon known as "cabbeling". 



Comparison of the temperature, salinity and density of the upper 20 m. at Sts. 

 WS 614-620 shows that these conditions come near to fulfilment (Table II). Thus 

 if surface water from Sts. WS 617 and WS 615 were mixed in the ratio of 5 : 8, the 

 mixture would have a salinity similar to water at St. WS 618 (34-36 "j^J but with a 

 density of 25-94. This value would not be in equilibrium with surrounding water until 

 it had sunk to a depth of about 40 m. 



Inshore upwelling at Caldera (Fig, 22) cannot be traced to greater depths than 

 100 m., but offshore, upwelling may be traced to a depth of 250-300 m. The inshore 

 water at these depths is seen to have a higher temperature than water in the open 

 ocean ; later it will be shown to be a highly saline return current having a southerly 

 flow — it is a coastal current: the offshore water is lighter, is sub-Antarctic in character, 

 and probably has northerly flow (Fig. 22). 



No adequate idea of water movement can probably be obtained from a single line of 



''■ Translation from Sandstrom (1919). 



