162 DISCOVERY REPORTS 



strong temperature discontinuity layer, through which the salinity remains almost constant. This 



discontinuity is therefore a layer of great stability. 



The distribution and depth of the convection layer is set out in Table 5, zero indicating where none 

 existed or where it was too thin to be recorded by our observations (< 10m). 1 



Comparing this table with the distribution of surface temperature and salinity (Figs. 7 and 8) it is 

 clear that the convection layer is well developed in the oceanic surface-water, but is notably absent 

 or ill defined in the cooler coastal waters, in which the temperature decreases more steadily from the 

 surface to the sea-bed. 



The Orange river line (28 30' S.). The vertical sections of temperature and salinity (Figs. 12 and 13) 

 show that both the isotherms and isohalines trend upward to the coast. Although this pattern may 

 suggest an active vertical uplift of water against the coast, such a movement does not appear likely in 

 view of the well-developed discontinuity and the impoverished nature of the surface-waters (fig. 50). 

 The great reduction in salinity of the surface-waters at the inshore stations has already been referred 

 to, and the resulting decrease in density of the surface layers has led to very great stability at these 

 stations. 



It is suggested that this section shows an advanced stage of the upwelling process : a relic of previous 

 upwelling rather than a stage of active uplift. 



The Sylvia Hill line (25° S.) (Figs. 14 and 15). The whole water column at the inshore stations, 

 WS 988 and 989, was very stable. Clearly there is no indication of inshore upwelling on this line, but 

 then it will be remembered (p. 160) that the inshore stations were apparently influenced by a convergent 

 tongue of oceanic water from the north, and that the cooler water had been displaced offshore. In this 

 cooler water the lower stability and patterns of the isohalines in particular are strongly indicative of 

 vertical motion at stations WS 986 and 987. 



The Walvis Bay line (23 ° S.) (Figs. 16 and 17). The salinity section shows a prominent tongue of 

 highly saline water near the surface. Within this the temperatures are high, and it is characterized 

 by a pronounced thermal discontinuity. This is the northern edge of the intrusion of oceanic water 

 which curves to the south and influences the inshore station on the preceding line. On this Walvis 

 section the discontinuity remains well marked at the inshore stations, and the relatively high tempera- 

 ture and salinity of the latter show that no active upwelling was in evidence. The nature of the inshore 

 water masses here is somewhat problematical, for they appear to exhibit characteristics of a mixture 

 of the oceanic and coastal waters. The most plausible interpretation is that previously upwelled water 

 has subsequently become mixed with oceanic water and the resultant mixture occupies the area in the 

 proximity of the coast at Walvis Bay. 



The Mowe Point line (19 44' S.). Extending throughout the area to the north of Walvis Bay the 

 oceanic surface-waters form a thin layer about 50 m. deep, overlying a very strong temperature 

 discontinuity. A section through this layer is shown in Fig. 18, which shows the most northern line 

 of stations (the Mowe Point line). It will be seen that on the inshore boundary of this layer the dis- 

 continuity weakens and slightly cooler water lies against the coast. Although no salinities were taken 

 in this region it is highly probable that the coastal water is again an admixture of the oceanic and 

 coastal water types. There is a slight indication that the upward trend of the isotherms towards the 

 coast may represent a very early stage of upwelling. 



1 Xo salinity observations have been available for the 'bathythermograph stations', and so certain assumptions have been 

 necessary. At all full stations the character of the upper layer density distribution, and consequently the convection layer, 

 was primarily dependent upon variations of temperature, and the slight variations of salinity have affected the density curves 

 only slightly. It has therefore been considered justifiable to define the convection layer from the temperature curves at the 

 bathythermograph stations, as the layer within which the temperature remained almost constant. 



