SUBTROPICAL WATER 61 



the Agulhas current brings almost tropical water, whilst on the west coast of South Africa 

 upwelling of cold water increases the phosphate content in the surface. In late August 

 1930 a section (Sts. 414-421) was made from the subtropical convergence in 40 28' S, 

 16 52' E towards Cape Town. The surface content decreased from a value of 58 mg. at 

 the southern end of the zone (where a mixture with sub-Antarctic surface water had 

 increased the content) to 35 mg. at the last deep station before the coast was reached. 

 At the actual coastal station which was situated on the continental shelf, upwelling had 

 increased the surface content to 51 mg. 



West of Saldanha Bay five observations were made within 100 miles of the coast in 

 the third week of August 1930. These stations (408-412), were made in the Benguela 

 current and had a surface phosphate content ranging from 20 to 27 mg. In early 

 September 1930 a section was made south-east from Port Elizabeth (Sts. 424-435) ; this 

 section cuts across the Agulhas and return Agulhas currents. The coastal station had a 

 high surface phosphate of 49 mg. which was due to upwelling caused by the Agulhas 

 current, whose surface water had a content of 26 mg. This value increased towards the 

 south-east to about 35 mg. and then at the most southerly station where the salinity was 

 slightly lower than is generally associated with subtropical water, it rose to 56 mg., 

 possibly owing to admixture with sub-Antarctic surface water. East of Durban five 

 stations (436-440), in late September 1930 showed that the surface phosphate content 

 in the Agulhas current at this position and time of the year was 16 mg. 



The observations around South Africa which have been referred to above were all 

 made in August and September which are late winter months. In late March 1935, 

 however, two subtropical stations (1554 and 1555), which were situated south of Cape 

 Town, showed that the surface content of phosphate in subtropical water may be reduced 

 to zero at this time of the year. The phytoplankton catches at these two stations were 

 negligible and far too small to have been the immediate cause of the withdrawal of all 

 the phosphate from the photosynthetic layer. There remains the possibility that a secondary 

 outburst of phytoplankton activity in February or early March was responsible for the 

 removal. It would require a season's intensive work to test such a possibility, which 

 however, is supported by inshore investigations near Cape Town (private communica- 

 tion). If, as Dr Hart believes, the main phytoplankton outburst in subtropical waters 

 occurs in September, when as our own results have shown there is an appreciable 

 quantity of phosphate in the surface layer, it is reasonable to suppose that this available 

 phosphate is quickly removed. Sts. 1554 and 1555 were taken in late March, and it is 

 again reasonable to argue that regeneration of phosphate will have occurred in situ be- 

 tween September and March, and yet there was no phosphate in the surface layer at these 

 stations. If, as we have postulated, a secondary outburst of phytoplankton occurred in 

 February it would account for the absence of phosphate in late March. The September 

 observations were of course made in 1930 and the March data in 1935, and the 

 possibility of annual variation must not be forgotten. For instance, in the western 

 part of the South Indian Ocean some observations made in the subtropical zone 

 (Sts. 1566-1569), in the second week of April 1935 between Marion Island and 



