RHINCALANUS GIGAS 353 



usually In the upper nets in all but the " oldest " type of water (o-i-o° C). No observa- 

 tions were made in the South Atlantic water east of the South Sandwich Islands during 

 the second season comparable with those taken in January 1932. 



In 1 93 1 -2 the population in waters of Bellingshausen Sea origin around South Georgia 

 and in the Scotia Sea came to maturity in the middle of December, and males appeared 

 in the lower hauls in increased numbers at the same time of year. Nauplii were taken at 

 one station in the South Atlantic, in Antarctic surface water of Bellingshausen Sea origin, 

 on the eighteenth of December. At the end of the first week in January, in the same 

 water, an advancing summer generation was found in stage iii. Spawning evidently began 

 in the middle of December and possibly continued throughout that month in Antarctic 

 surface water of the South Atlantic east of South Georgia. Spawning and hatching 

 of the eggs apparently took place within a very short time. In 1932-3, however, the 

 population in the Scotia Sea and around South Georgia was found to be mature at the 

 end of November, and males began to increase in the catches at the same time. The 

 maximum number of males appeared at the beginning of December during this season, 

 as against the middle of that month in 193 1-2. It is, therefore, justifiable to suppose 

 that spawning began correspondingly earlier — a fortnight or three weeks — in 1932-3 

 than in 193 1-2 in the waters around South Georgia. The condition of the summer 

 generation in the Drake Passage in early February (stages iv, v and vi) confirms this 

 supposition. 



Comparison of the isotherm charts for the seasons 193 1-2 and 1932-3 (Figs. 5, 6) 

 shows that the former was a colder and "later" season than the latter. In the spring 

 of 1932-3 the isotherms for i-o° C. and over occupied a more southerly mean position 

 than in the spring of 193 1-2. This is less true of the isotherms for o" C. and tempera- 

 tures below 0°. Warm water, therefore, extended farther south at the beginning of 

 1932-3 than at the beginning of 193 1-2 and the transition from warm to cold water 

 from lower to higher latitudes was correspondingly more abrupt. The pack ice also 

 was farther north in 1931-2 than in 1932-3. In the Drake Passage at the end of 

 November 193 1 the pack ice was met with in about 64° S. (St. 735) and was followed 

 through the Drake Passage to about 60" S. (St. 741) where it trended away southwards. 

 It thus lay far to the north of the South Shetlands. It was met with again near the 

 South Orkneys at the beginning of December and was again followed north-eastward 

 to about 57° S. (St. 767) in the longitude of South Georgia. At the end of October 1932, 

 however, the pack ice was met with in 67' S. in the western Drake Passage (St. 995) 

 and the Bransfield Straits were entirely free during early November. At the end of 

 that month the ice was followed from about 62° S. in the longitude of the South 

 Orkneys (St. 1035) north-eastwards to about 58° 30' S. (St. 1045, between Bristol and 

 Montague Islands) in the South Sandwich group. Thus appearances strongly suggest 

 that the difference in the spawning time of R. gigas in the two seasons may be con- 

 nected with these differences of temperature and ice conditions. 



In 193 1-2 no spawning at all took place in Weddell Sea water, at least before the end 

 of January ; but the line taken from South Georgia through the waters of the Weddell 



