9 8 DISCOVERY REPORTS 



neighbourhood of the ridge on the eastward movement in both layers (cf. Ekman, 1928). 

 The greater northward tendency appears to prevail too far to the east of the ridge in 

 section 9, but it must be remembered that the ridge runs almost south-east, and also that 

 there is another extensive area of shallow water a short distance farther east in the 

 neighbourhood of the Shackleton ice-shelf. 



The observations made south of 32 S in the sub- Antarctic and subtropical zones in 

 the eastern half of the Indian Ocean show that the deep water has for the most part a 

 maximum salinity of 3475-3476 °/ 00 , and is throughout less saline, colder, and poorer 

 in oxygen than the deep water of the same salinity farther west. These properties suggest 

 that it is formed partly by the mixing of the north Indian deep water with the colder 

 and less saline Antarctic bottom water, and the poorly saline though warmer water of the 

 intermediate current ; but there is also good reason to believe that it contains water which 

 flows eastwards from the Atlantic Ocean. The upper part of the layer has a particularly 

 low oxygen content which, although it may be due partly to the lack of a rapid circula- 

 tion in the eastern part of the ocean, also suggests that the water belongs to the north 

 Indian current. 



From a temperature section across the ocean in 34 S based chiefly on the observa- 

 tions of the ' Gazelle' (1874-6), Moller (1929, fig. 13, pp. 26, 37-8) concluded that the 

 deep layer was reinforced in the eastern part of the ocean by a second deep current, 

 similar to the north Indian current, formed along the north-eastern seaboard. In a more 

 recent paper, however (1933, p. 234), written after the salinity measurements of the 

 ' Dana ' and ' Snellius ' were in part available, she was able to show that there was no 

 such current, and that the evidence of the Gazelle section is evidently misleading. 

 Our observations suggest that the section is based on faulty data. The mean of the 

 observations at Gazelle Sts. 82, 83 shows that the water in 34 S, 102 E has a minimum 

 temperature of 1-9° C. at a depth of 1250 m. ; but such water could only be Antarctic 

 surface water near the convergence, or Antarctic bottom water. The second of these 

 possibilities is ruled out because the water is found at such a shallow depth, above 

 warmer water, and the first could only be realized if there was a much stronger north- 

 ward movement of Antarctic water in this region than there is even in the western half 

 of the Atlantic Ocean, an occurrence which is definitely contradicted by the observa- 

 tions in section 9, and by the surface temperature charts compiled from other sources 

 (Moller, 1929, figs. 17, 21). The Gazelle data also show that below the cold water, be- 

 tween 1500 and 2000 m., there is a stratum whose maximum temperature is as much as 

 3 C. higher, and the existence of such a large inversion seems to be almost impossible: 

 our observations farther south showed no sign of an inversion, and if the deep water in 

 34 S were so much warmer than the water at 1250 m. it would probably have to be 

 more saline than any deep water found in the Indian Ocean in order to make the water 

 column stable. 



The salinity, temperature, and oxygen distribution in the Southern Ocean south of 

 Australia (Figs. 19-22, sections 10-12, Plates XIX-XXVII) indicates that the deep water 

 flows principally towards the east and south. The southward movement is probably a 



