296 DISCOVERY REPORTS 



again we may have failed completely to strike the main mass of the eggs. The March capture, for 

 instance, in the East Wind drift of a very large number of Second Nauplii very close to the bottom on 

 the Antarctic continental slope (p. 93, Table 14 and p. 303, Fig. 71) suggests that there at any 

 rate, in these high latitudes, spawning might still be in full swing. Evidently, however, judging from 

 the unbroken negative coverage extending in April from 90° W eastwards to 90° E, it ceases every- 

 where on a major scale before the end of March. It may continue on a much reduced scale in the 

 East Wind zone into April, for although no free eggs are recorded then, as late as the 20th of the 

 month (p. 177), at a station near Enderby Land, a recently spent female was captured in which, 

 still clinging to the otherwise empty ovary, were a few unsegmented globular eggs of the same 

 size, colour and translucency as those recorded from time to time from the upper layers of the 

 plankton. 



There is still one free record, although a minor one, from the coastal waters of north Graham Land 

 suggesting that spawning there may be continuous from November to March. For the rest it will be 

 seen the occurrences of eggs, such as they are, are confined as before exclusively to the Weddell drift 

 and to the East Wind drift, in the latter locality mainly near the land. The South Georgia area, in 

 spite of the thoroughness of the sampling there, is again for all practical purposes barren, and so too, 

 it would seem, is the Pacific sector up to very high latitudes, though here again the East Wind drift is 

 scarcely sampled. 



The principal inferences to be drawn from the distribution of the eggs, a comprehensive view of 

 which is given in Fig. 67 (see also p. 301, Fig. 70 and p. 303, Fig. 71), may be summarised as 

 follows : 



(i) Although the main mass of the eggs has hardly if ever been properly sampled, spawning it 

 appears may be a phenomenon associated in its major aspects with the shallow shelf or slope water 

 of the north and east Graham Land peninsula^ and with the corresponding shelf or slope waters of 

 continental Antarctica in the higher latitudes of the East Wind drift. Yet paradoxically both spent and 

 gravid females have in fact been recorded (p. 194, Figs. 24 and 25) far from the slope and coastal 

 waters of the continental land. 



(2) There is an early spawning, associated with the shelf or slope waters of the north Graham Land 

 region, in November-December. 



(3) There seems to be a discrete or restricted spawning there of which we have continuous 

 evidence from November to March. 



(4) There is a later spawning associated with the shelf or slope waters of the East Wind drift 

 which, beginning on a reduced scale in January (or rarely perhaps in very late December), is in full 

 swing in February and March and continues possibly even into April. 



(5) Throughout the spawning season the mass shedding of eggs in the upper layers of the oceanic 

 water far away from land seems to be an uncommon event. Yet, again paradoxically, the vertical 

 distribution of the spent and gravid females (p. 188, Table 39) suggests that spawning is in fact a 

 surface phenomenon. 



(6) Spawning at deeper levels, if it takes place at all, must take place far beyond the reach of our 

 deepest, 1 500-1 000 m., vertical nets. 



(7) Spawning in general, whether it be a deep or shallow phenomenon, is restricted to the East 

 Wind- Weddell surface stream although virtually none it seems takes place on the South Georgia 

 whaling grounds. 



The overriding conclusion suggested in (i), that spawning is a phenomenon associated with shelf 

 or slope regions, is strongly supported by the distribution of the eggs in, and in the neighbourhood of, 

 1 And doubtless, too, with other shelf or slope regions in this locality, but farther south, as yet unexplored. 



