S o DISCOVERY REPORTS 



two stations has already been considerably warmed; it has also without doubt been 

 diluted. Typical temperature, salinity and density values of the bottom water at 1500 m. 

 in the deepest part of the strait were found at St. 543 to be - 1-72° C, 34-65 °/ 00 and 

 27-91. The temperature of this water is higher than the freezing-point of water of 

 salinity 34-50 °/ 00 . It is not difficult to imagine that in winter the surface salinity 

 increases to a value greater than that of the bottom water and the temperature decreases 

 to the freezing-point of such water, with the consequence that in areas where lanes of 

 open water are found, sudden and intense freezing takes place with the formation of 

 instabilities. In November and December we have noted that the water on the conti- 

 nental shelf is still very saline and cold from surface to bottom. In the winter months, 

 water still heavier and of uniform temperature and salinity will be formed on the shelf. 

 Within this layer intermittent sinking of water of higher salinity and lower temperature, 

 formed by freezing processes, will take place and this heavier water will follow the slope 

 of the bottom contour. In this way the bottom water in the strait is renewed and 

 maintains its high oxygen content. This renewal of the bottom water will cause the 

 water above to flow out of the strait, and it is significant that the oxygen content is at a 

 minimum at a depth of 400-600 m. at St. 543 in the deepest part of the strait. This 

 depth almost corresponds with the depth of the gap in the ridge south of Clarence 

 Island through which the displaced water must flow. 



Thus in the Bransfield Strait the formation of the bottom water is considered to take 

 place in winter as a result of freezing processes mainly on the extensive continental shelf 

 of west Graham Land and Trinity Peninsula. This mode of formation probably takes 

 place also in other parts of Antarctica, and in particular some of the Weddell Sea 

 Antarctic bottom water is undoubtedly formed on the continental shelf in the southern 

 part of this sea as suggested by Brennecke (1921, p. 38). From the observations hitherto 

 made it cannot be determined whether a renewal of the bottom water in the deep basins 

 of the Bransfield Strait may also be affected by horizontal flow across the ridges, 

 particularly across that to the east of the strait. 



FEBRUARY 1929 

 The movement of the surface water in February 1929, as shown by the topography of 

 the surface of the sea for the two principal lines of stations (Fig. 59), shows in general the 

 two main features noticed for November 1929 and December 1930; there are, however, 

 some modifications. The characteristic bend of the dynamic isobaths on the line from 

 Livingston Island occurs almost as far across the strait as Astrolabe Island, so that the 

 north-westerly flow of the surface water back to the north-easterly flowing current 

 along the South Shetland Islands is found over a greater distance than in November 

 1929. The north-easterly flow on the South Shetland side of the strait is concentrated 

 chiefly between the second and third stations from King George Island. A difference 

 of height of 9-6 dynamic cm. occurs between these two stations, which are only 5-8 miles 

 apart, and this corresponds to a surface current of 69 cm./sec. The gradient may be too 

 great owing to considerable stray on the water-bottle wire at St. WS 383 causing the 



