290 DISCOVERY REPORTS 



which must exist has never been properly sampled, a fact which, there is good reason to believe (p. 1 82), 

 may be ascribed (i) to our persistent, and understandable, disinclination to hazard our apparatus too 

 close to the bottom in Antarctic coastal waters, or (2) to the possibility, somewhat remote (p. 189) 

 though it appears to be, that the eggs, being sometimes laid deep down in oceanic water, are always 

 in such circumstances beyond the reach even of our deepest nets. Even if spawning near the surface 

 were followed by rapid sinking of the eggs to such deep levels, it is hard to see how they could so 

 rarely be sampled if there were widespread oceanic spawning. 



Unsatisfactory, however, though they are, they do it seems provide some indication that spawning 

 on a major scale in or over deep oceanic water far away from the continental land is not a very common 

 event. For out of 858 vertical stations^ made over deep water during the spawning season, the 

 vast majority of them extending down to the 1000 m. level, but including 64 down to 1500 m. 

 or more, only two can be said (Fig. 65) to give any indication that such a spawning could have 

 taken place. 



November-December. The first small, though perhaps significant, capture of eggs, as Fig. 64 

 shows, is recorded from the Bransfield Strait in November. In the same month small to negligible 

 numbers, that is, less than 10 per station, were recorded at six other stations in this shelf region. 

 Apart from these occurrences there are no other November records throughout the circumpolar 

 sea. 



In December, at the extreme western end of the Weddell drift, where its course passes across 

 the eastern opening of the Bransfield Strait, we struck the second largest concentration of eggs, 

 3030 in all, that has so far been recorded. The vast majority were obtained in a 500-250 m. net 

 haul at a station where the depth was only 510 m., an occasion when the net had probably begun 

 fishing closer to the bottom, without actually fouling it, than on any other occasion either before 

 or after. 



Elsewhere, beyond the north Graham Land shelf region, on the South Georgia whaling grounds, 

 where the heavy negative coverage is particularly to be noted, and in fact throughout the entire 

 oceanic water of the circumpolar sea, there is still little indication in December that spawning has 

 taken place, and it is perhaps significant that such scanty indications as there are should be largely 

 confined to the Western Weddell drift not very far from the place where a major spawning seems to 

 have occurred. There are, however, two very small occurrences, both in the last 3 days of the month, 

 from the shelf water of the East Wind drift near the west Graham Land coast. 



Two major facts then seem to emerge from the November-December egg chart, and while both 

 have been fully discussed in an earlier section (pp. 176-212), they are important enough to merit 

 repetition. They are (i) the spawning of the krill, at its commencement at least, is to all appearances 

 a relatively high latitude, shallow water, coastal phenomenon, and (2) having regard to the extensive 

 and on the whole convincing negative coverage of other parts of the circumpolar sea, and especially of 

 the coastal and oceanic water round South Georgia, it might in fact at the same time be an event peculiar 

 to the north Graham Land shelf region (and no doubt to other adjacent shelf areas as yet unexplored) 

 and not as has hitherto been supposed a general and widespread occurrence all round Antarctica. The 

 first egg chart in short seems to justify the view that the existence of a discrete spawning ground in 

 the far western (or south-western) reaches of the Weddell stream must be more than mere hj^pothesis. 

 Moreover, the obvious objection to this, that spawning on an unknown but presumably comparable 

 scale might simultaneously be taking place under the vast ice-sheet which throughout November until 

 about mid-December still extends over the circumpolar sea, can readily be answered. For if such 

 a thing were happening, as Ruud (p. 44) once suggested it might, we should expect to find, when 



* Representing over 5000 operations of the vertical net. 



