ANTARCTIC INTERMEDIATE WATER 65 



consumption of a considerable amount of phytoplankton : also diatoms tend to die and 

 sink out of the Antarctic surface layer. Thus at the time of the main phytoplankton 

 outburst a large amount of phosphate is removed from the photosynthetic zone of the 

 surface layer, and is not all transported northwards to sink across the convergence into sub- 

 Antarctic water. Similarly in the southern end of the sub-Antarctic zone phosphate is 

 removed from the surface layer, and some of this is lost to the Antarctic intermediate 

 current by migration of plankton and sinking of dead phytoplankton organisms. Thus 

 if two seasons are considered, (1) the winter months when phytoplankton activity is at 

 a minimum and the phosphate content of the surface layer is greatest and (2) the time 

 when the main phytoplankton outburst is at its height and large amounts of nutrient salts 

 are being consumed, it is possible to trace the underlying causes of the seasonal variation 

 of phosphate content in the Antarctic intermediate current. In the Antarctic and sub- 

 Antarctic zones the phosphate content of the surface layer is at a maximum in winter, 

 and in consequence, the Antarctic intermediate current which has its origin in the area 

 of mixing 100-200 miles north of the Antarctic convergence and is mainly composed 

 of Antarctic and sub-Antarctic surface waters, will have its greatest nutrient salt content 

 in winter. Similarly the nutrient salt content will be least when the waters which 

 make up this current have their minimum nutrient salt contents. This will be 

 immediately after the time of the main phytoplankton outbursts of October-November 

 in the sub-Antarctic zone, and November-December in the Antarctic zone. If the path 

 of the intermediate current is followed northwards by plotting the phosphate content at 

 a definite position within its structure, i.e. at the depth of minimum salinity, it will be 

 possible to see what kind of a curve is obtained for the south to north distribution of 

 phosphate within this current. In Fig. 18 the phosphate values in Table V have been 

 plotted against latitude. The resulting curve consists of a series of maxima and minima. 

 The positions of these maxima and minima correspond very well with those found in the 

 curve (which is also shown in Fig. 1 8) of the minimum salinity of the current when plotted 

 against latitude. Similarly, a very close correspondence is found if the oxygen and silicate 

 contents at the depth of minimum salinity of the layer are plotted against latitude (see 

 Figs. 19-21). Deacon {Discovery Reports, VII, p. 224 et seq.) has used the oxygen and 

 salinity values at the depth of minimal salinity of the intermediate current as a means 

 of obtaining an estimate of the northerly component of this current. I have heard 

 criticisms of this method on the ground that the various maxima and minima in the 

 curve are due to some sort of wave action and not to seasonal variation. It is remarkable, 

 however, that the curves show very many correlations; for instance, if we consider the 

 effect of winter formation of the intermediate current which involves high phosphate, 

 silicate and salinity, and low value of oxygen content, we find the curves show that 

 phosphate, silicate and salinity are maximal and oxygen content is minimal at the same 

 positions of latitude. Similarly, summer formation of the intermediate water would 

 involve low contents of phosphate, silicate and salinity, with a high oxygen content : 

 the positions of minimal contents of phosphate, silicate and salinity, and maximal oxygen 

 content agree extremely well. It is, therefore, reasonable to suppose that the crests 



