,6 DISCOVERY REPORTS 



The monthly surface temperature range of the grand total of euphausians captured in the surface 

 (loo-o m.) layer is shown in Table 4.1 The figures given here, and in the regional presentation of the 

 data (Tables 5-9) that follows, show the average monthly catch per every 1° C and are based through- 

 out on the material obtained in the surface (0-5 m.) and oblique (loo-o m.) stramin nets with which 

 night and day we sampled the surface population (see again, however, p. 65). In both general and 

 regional presentations of the data the population as before has been divided into three broadly dif- 

 ferentiated developmental phases: (i) krill under 16 mm. long (vast majority larvae), (2) krill 

 16-20 mm. long (very young adolescents), and (3) krill over 20 mm. long (older adolescents and adults). 

 Anomalous occurrences of E. superba have been disregarded. There are two such occurrences in our 

 records, both of them in February, at Station WS 139 (surface temperature 6-15° C) and at Station 829 

 (surface temperature 6-29° C), one specimen of the over 20 mm. class having been taken in each 

 instance. Both stations lie in Subantarctic water immediately to the westward of the S-shaped bend 

 in the convergence (p. 58, Fig. 4) which occurs in 50° W about half-way between the Falkland 

 Islands and South Georgia. In the neighbourhood of this bend hydrological conditions are 

 complex, somewhat unstable and, as John (1936) has remarked, characterised by particularly 

 strong water movements both across and beneath the convergence which resuh in the carrying 

 of certain species (for example Euphausia frigida), normally inhabitants of the Antarctic zone, 

 north of their usual habitat. A strong tongue of cold water flows towards this region from the 

 Weddell Sea and no doubt it is this movement that is responsible for the occasional occurrence of 

 E. superba in this unusual position, in conditions which as Table 4 shows Ue well outside the 

 normal limits of its absolute temperature range. In the concluding remarks of his paper on the 

 distribution of Rhizosolenia curvata Zacharias, Hart (1937) also calls attention to the water move- 

 ments in this area, suggesting that it is there apparently that the position of the Antarctic convergence 

 is most variable and there too apparently that mbcing across the convergence most frequently takes 



place. 



Disregarding these anomahes it will be seen that the maximum or absolute range of temperature 

 in which the surface population of the krill may occur lies between —2-00° and +3-99° C, the mini- 

 mum surface temperature at which it has been recorded being -i-Sg°^ and the maximum +3-90° C, 

 the total range thus extending over nearly 6°. Tolerance of the maximum range it will be seen is 

 confined as might be expected to the warmer months of the year, namely, from November to May, 

 and even then it is the older stages including the adults (represented by the over 20 mm. group) 

 and the early surface larvae (represented by the under 16 mm. group) that alone are involved, the 

 former from November to March, the latter from March to May. It will be seen too, again as might 

 be expected, that it is only in February, the warmest summer month, that the older stages occur in 

 any real measure of abundance from extreme to extreme of the maximum range. At the end of autumn, 

 with the onset of winter conditions, the temperature range noticeably shortens, the vast bulk of the 

 total surface population, as the average catch-figures show, existing from June to November in sub- 

 zero conditions. After October it lengthens again and continues to do so until it reaches its February 

 maximum. 



1 In this and Tables 5-1 1 I use the temperature measured at the extreme surface, because at many stations where towed 

 nets were used this is the only point at which the sea temperature was measured. It can differ from the temperature lower 

 down within the loo-o m. horizon sampled by the oblique nets, but seldom by very much except when the extreme surface 

 is occasionally warmed up in summer. In any case there is reason to believe that the vast majority of krill taken in the oblique 

 nets were quite close to the surface. Thus there should be no appreciable error in Tables 4-9. Tables 10 and 11 it is true 

 show the catches from 250 m. to the surface, but even here any variations from the surface temperature should not be enough 

 to affect the conclusions given at the end of this section. 



2 At Station 1781 at the ice-edge in Weddell East in June 1936. 



