RHIZOSOLENIA CURVATA 



443 



more closely worked areas in the Scotia Sea and in longitude 8o° W. It is impossible 

 to include all these on a chart of this scale without overlapping. 



The direct application of the evidence accumulated by measurements and estimates 

 of the numbers in the net hauls to the problem of the distribution of the organism is 

 shown in Table X. Here the observations have been grouped according to their distance 

 from the average position of the Antarctic convergence. The percentage frequencies of 

 the different diameter classes of R. cnrvata have been calculated by the methods already 

 described, while the average number of individuals per net haul is shown in the right- 

 hand column. 



Table X. Size variation and relative abundance of Rhizosolenia curvata at 

 different distances from the Antarctic convergence 



It may be added that records of the species were obtained at four stations north of the 

 distance scale shown in the table. At these the mean diameter of the individuals was 

 only 54^ and their average number only 450 per net haul. Evidently the normal limit 

 of the species on the sub-Antarctic side of the convergence is about 300 miles, and from 

 the temperature records discussed in earlier sections it seems probable that this will be 

 found to coincide roughly with the 8-5° isotherm, except in the eastern South Pacific. 



It is obvious that the size frequency distribution shown in Table X should be a broad 

 reflection of the size temperature correlation already demonstrated, but it is bound to be 

 somewhat masked by seasonal effects. However, it is sufficiently clear that the highest 

 proportion of small individuals was recorded at those stations farthest on the Antarctic 

 side of the convergence, while the largest proportion of large individuals was found in 

 the sub-Antarctic Zone. The small fluctuations in the actual position of the convergence 

 probably account for the proportion of large individuals found immediately to the south 

 of it being larger than the theory ideally demands. It will be readily appreciated that in 

 drawing up a general comparison such as this, it has only been possible to consider the 

 "probable average" position of the convergence. 



Wimpenny (1936) and Hendey (1936), as a result of their work on phytoplankton from 

 very different aspects, have been led to conclude that the size/temperature relation of 

 marine plankton diatoms is the reverse of that which has long been known to exist in 



