RHIZOSOLENIA CURVATA 



427 



ticularly unstable in this position. It is very unfortunate that no observations north of the 

 convergence at that time are available. At the winter stations 1381 and 1382 the prob- 

 ability of southward mixing is strong. The normal direction of the surface drift is known 

 to be reversed at times during the coldest months of the year, and the surface tempera- 

 ture of 2-60° C. at St. 1 38 1 is so high for its position at the time it was worked as to be 

 virtually impossible unless this factor was operating. It will be seen that at St. 1391, 

 worked only 10 days later at a similar distance on the opposite (sub- Antarctic) side of 

 the convergence, the temperature was only o-io° C. higher. 



a Valdivia records 



# Scoha records 



• Positive records 

 G Negative records 



Fig. 2. Observations of Rhizosolenia curvata in the South Atlantic Ocean. 



The richest catch from our meagre South Atlantic collection was obtained at St. 11 97 

 on the sub-Antarctic side of the convergence. Another station of special interest is 670 

 which furnishes one of the very few records (only three in all) of the occurrence of 

 R. curvata at temperatures higher than 8° C. 



OBSERVATIONS IN THE SOUTHERN INDIAN OCEAN 

 AND SOUTH OF AUSTRALIA 

 Our records of R. curvata for this area are not very numerous, but the majority were 

 obtained on series of observations running more or less north and south through the 

 full width of the sub-Antarctic Zone. The data thus provided were, therefore, particularly 

 well suited to the purpose of this study, and will be seen to agree almost perfectly with 

 the general theory of the distribution of the species. The results are shown in Table IV, 

 and the positions of the stations have been plotted in Fig. 3. 



All the larger hauls of R. curvata in this sector occurred in the sub-Antarctic Zone, 



