ANTARCTIC SURFACE WATER 25 



In the curve for the northern region the average phosphate content falls from August 

 to the end of January and then rises to 1 June ; the fall from August is very steep until 

 1 November, after which a nearly flat curve results until January when the gradient 

 becomes steeper only to rise subsequently very steeply in March. 



The steep fall in the southern curve from August to the end of October is entirely 

 at variance with the observed phytoplankton concentration in this area; the fall of 

 phosphate content is taking place over a period when phytoplankton activity is practically 

 nil and must be due to the effect of the arrival in the southern area in earlier months of 

 water of relatively low phosphate content in the upper layers of the warm deep water. 



The fall of 20 mg. during October in the northern curve is a little early for the 

 time that Dr Hart considers to be the average for the main phytoplankton outburst. 

 However, it has been previously stated that the dates of the main outburst of the 

 phytoplankton and its duration are average dates subject to ±2 weeks. In a private 

 communication from Dr Hart the big rise of phytoplankton concentration is shown as 

 apparently beginning in October, but the exact date is uncertain. It will be noticed 

 in the northern curve that the low level of phosphate content persists from 1 November 

 to 1 March, a period of 4 months, with a distinct minimum at the end of January. 

 We know from phytoplankton data that in the northern region effective phytoplankton 

 production lasts from November to December, that the phytoplankton concentration 

 is falling off rapidly from mid-December onwards, and that by 10 April it is back 

 at the figure prior to the main outburst of November. Most of the fall in phyto- 

 plankton concentration occurs from the last week of December to the third week of 

 January. Thus the period of 4 months when the phosphate content remains low over- 

 laps the time of effective phytoplankton production. This overlap may perhaps be 

 explained by the arrival in the northern region of surface water from the southern 

 and intermediate regions in which phytoplankton has been continuously removing 

 nutrient salt from November to April, and thus the period of low phosphate content 

 in the northern region is lengthened beyond that which is due to phytoplankton 

 production in situ. 



The rise during November, which interrupts the fall in the southern curve from 

 August to mid-January, is interesting because phytoplankton production is starting 

 in November in this area and phosphate should be withdrawn, but apparently it is not 

 until December and January that the effect of production is correlated with a downward 

 gradient of the phosphate curve. The reason for the rise during November must be 

 sought in the northern curve. In the northern curve the minimum occurs at the end 

 of January, after which regeneration causes the phosphate content to rise. After the 

 Antarctic surface water leaves the Antarctic zone it travels with a northerly component 

 as part of the Antarctic intermediate current (see Fig. 6). Vertical sections in the South 

 Atlantic Ocean for example, show that at about 38°-43° S the phosphate content is a 

 maximum in the mixed water between the Antarctic intermediate current and the warm 

 deep water. This mixed water forms the warmer and less saline part of the warm deep 

 water and travels southward as the upper layers of this current. At the Antarctic con- 



