THE LARVAL STAGES 123 



of hatching before they appeared in the Antarctic surface layer. If, however, the developmental ascent 

 does take very much longer than has been suggested,^ let us say it took not 8 + 5 days but 24 + 1 5 days, 

 and if the higher deep current speed indicated by the South Georgia experiments be accepted, then 

 a major southerly displacement of the larvae, involving distances of 100 miles or more, can well be 

 imagined. A displacement on this scale would mean that the deep larvae in the East Wind zone, never, 

 it will be shown (p. 200, Fig. 28), very far away from the land, would constantly be carried deeper (in 

 the latitudinal sense) into this coastwise current, while those hatched deep down in the far southern 

 reaches of the Weddell drift, such as, for example, the larvae at Station 2594 (p. 90, Table 13), 

 would, when they appeared at the surface, no longer it seems find themselves in the east-flowing 

 current under which they were born but in the west-flowing East Wind stream some considerable 

 distance to the south. In other words in high latitudes all round Antarctica, wherever in the East 

 Wind zone there should be an adult breeding stock (p. 194, Figs. 24 and 25) its progeny, whether 

 hatched near, or some distance from the land, would tend automatically to become compressed within 

 a relatively narrow coastal belt in which, first as larvae, and later as adults, they would move westward 

 in the surface stream until eventually some of them would find their way into the Weddell Sea where 

 they would become augmented by fresh influxes of larvae deep-hatched in the far southern reaches 

 of the Weddell drift. 



On the Atlantic side of Antarctica between 30° W and 30° E the warm deep water appears to be 

 traveUing mainly towards the east (Deacon, 1937, p. 86, Fig. 22, and pp. 90 and 92). There is, however, 

 between 10° and 30° E a strong indication of south-westerly movement reaching far south into the East 

 Wind zone where it penetrates well west of 0°. Model (1958) shows the same movement even more con- 

 spicuously. As Deacon (1937, p. 81) points out, since there are general movements towards the north 

 in both surface and bottom currents all round the continent, these currents ' can only exist if there is 

 a compensating movement towards the south in the intermediate deep water '.^ Between 30° W and 

 30° E the deep water is generally warmer below the westerly current near the continent than it is 

 below the cold more northerly surface current flowing eastwards from the Weddell Sea. This gives a 

 clear indication that the main southward movement in the warm layer occurs in a sweep round the end 

 of the Weddell drift, but the fact that every deep observation in the Weddell Sea shows a warm layer 

 sandwiched between two very cold ones indicates everywhere a very active supply of warm deep water. 

 There can in fact be little doubt that here as in the Ross Sea (p. 124) its influence as a major instrument 

 of southerly dispersal does not cease until it comes up against the continental slope, as for instance in 

 0° (Discovery Reports, Station List, 1937-9), where we do in fact find it penetrating practically up to the 

 coast. But perhaps the overriding consideration is this. Unless there be somewhere, between 30° W 

 and 30° E, active penetration of the East Wind zone by warm deep water from the north the continued 

 existence of the Atlantic krill population it will be seen (p. 432) becomes hydrologically impossible. 



The developmental ascent in relation to the local distribution 



OF krill in the Ross Sea 

 At the head of the Ross Sea there is a vast expanse of shallow water from which the warm deep 

 current (Marshall, 1930; Deacon, 1939; Discovery Reports, Station Lists, 1927-9 and 1935-7) ^^ 

 virtually excluded.^ The distribution of adult and adolescent whale food in and on the outskirts of 



1 There is some evidence (see p. 312 in the main distributional part) that it might in fact occupy the better part of a month, 

 if not longer. ^ See also Deacon (1957a). 



^ Wright and Priestley (1922), referring to the annual calving of bergs from the Ross Barrier face, state that the prime 

 cause is melting by the warm water lying beneath the barrier, which is most effective close to the edge. 'This action', they 

 continue, 'suggests the presence of a warm current from the north in the lower layers of the Ross Sea'. There is nothing, 

 however, in our hydrological data to suggest (see below) that such a current exists. 



