334 DISCOVERY REPORTS 



to the population found in the "oldest" type of Weddell Sea water (o-i-o° C.) at the 

 beginning of the month (Sts. 795, 798, 806 and 808), and represents that the part of 

 the population of the Bellingshausen Sea current in the Drake Passage which had failed 

 to reach maturity by the end of January, through having drifted into Weddell Sea water 

 with a temperature lower than i-o° C. At St. 825 (Fig. i b), however, we find the stock 

 consisting almost exclusively of stage iii. This must result from a fairly recent spawning 

 and part of the parent generation remains, particularly in the lower haul. It will be 

 noticed that the stock here in the Scotia Sea at the end of January is of the same age as 

 that found during the first week of the month at Sts. 796, 802 and 803 (Fig. 21). Now 

 the population sampled at St. 825 must result from a spawning which took place to the 

 south-west in the Scotia Sea in water of Bellingshausen Sea origin, where the tempera- 

 ture, according to the isotherm map (Fig. 5), may have been less than 0° C. This 

 spawning, then, must have taken place, if the rate of development is the same in all 

 waters, at least three weeks later than that which took place east of South Georgia at 

 the end of December and which gave rise to the stock at Sts. 796, 802 and 803 (Fig. 21). 

 Alternatively the spawning took place at the same time, but the growth rate in the water 

 of lower temperature has been slowed down. 



Falkland Islands to South Georgia, mid-February, 1932 (Fig. 24, Table VI d) 



On the line from the Falklands to South Georgia in the middle of February we find 

 small catches at Sts. 828 and 829 (Fig. i b), in sub-Antarctic water, consisting of stage iii 

 and no trace of a parent generation. 



The disappearance of Rhincalanus gigas from sub-Antarctic water at the end of the 

 summer has already been commented upon (pp. 299, 311,312). In the light of what we 

 have seen of the course of events during the summer two explanations may perhaps be 

 advanced to account for the diff'erence between the summer and winter conditions on the 

 sub-Antarctic side of the convergence. There is at present, however, no evidence to prove 

 which, if either, of the suggested explanations is correct. It might be supposed that the 

 Rhincalanus population begins the descent to its winter level much earlier north of the 

 convergence than south of it. This was found to be true to a certain extent by Mackintosh 

 (1935), who has shown that it was to the north of the convergence in March 1934 that 

 R. gigas first began to seek its winter level. But this hardly seems sufficient to account 

 for the great reduction in the catches within the surface 250 m. in sub-Antarctic water 

 so early as February in the seasons 193 1-2 and 1932-3. If, however, this theory is 

 correct the population sampled at Sts. 828 and 829 in February 1932 was part of a 

 stock of which the main body lay at a level below 250 m. and was therefore out of range 

 of the nets. An alternative, and possibly more probable, explanation is that spawning 

 is very reduced in sub-Antarctic water during the summer, so that when the over- 

 wintered generation dies out in late January or February there is only a very greatly 

 diminished summer spawned advancing generation to replace it. The absence of a parent 

 generation from the stock at Sts. 828 and 829, in which stages ii and iii predominate might 

 suggest that this population was not spawned in the water in which it was taken. It is 



