THE OLDER STAGES 211 



Intensive exploration of the shelf waters at all levels, although primarily directed towards determining 

 whether the spawning that takes place there is a near-surface, bottom or near-bottom phenomenon, 

 might at the same time help to clear up the mystery of our failure so far to produce evidence of large- 

 scale hatching in these shallow conditions, a failure which suggests the possibility that the majority 

 of the shelf or slope laid eggs might get carried away. 



(5) Direct measurement, using the neutral-buoyancy floats that Swallow (1955) has recently 

 employed with such success below the Gulf Stream (Swallow and Worthington, 1957; 1961), of the 

 rate at which the bottom water is travelling below the Weddell stream. 



No doubt there are other matters that might seem to call for attention, such as, for instance, our 

 failure to find anything but the most meagre evidence of both spawning and hatching on the South 

 Georgia whaling grounds. Much of what still remains in major doubt, however, will, I feel sure, 

 be largely cleared up as a result of the measures suggested. 



Finally, in view of the immeasurable abundance of the krill in the Antarctic seas and the fantastic 

 numbers of eggs that must be liberated annually to maintain such a population, I feel sure there must 

 be some simple explanation of the so far almost complete failure of our nets to sample the eggs in 

 the massive concentrations in which they must naturally occur in the sea. Our failure, except once, 

 to find the hatching eggs is of course easy to explain, for they, except it seems for very rare instances, 

 lie far beyond the range of our deepest vertical nets. But if they are laid near the surface, where the 

 mass of the gravid and spent females seems to be, then surely somewhere in the enormous area we 

 have explored, we should have struck them there, or struck them, as they went down, at some deeper 

 level. It is not that they are too small to be readily taken in the vertical net, for twice it has sampled 

 them with at least some measure of success. Nor it seems can it be that they are so widely scattered 

 that only negligible numbers can be captured, for again everything points to the contrary, that they 

 occur naturally in concentrations, as for instance at Stations 540 and 2594. The gravid swarm in 

 fact (p. 250, Fig. 53) must liberate an enormous mass of highly concentrated eggs, the ripe females 

 (Bargmann, 1945) being capable of producing over 11,000 ova each, while the concentrations of 

 Nauplii and Metanauplii so regularly encountered deep down in oceanic water suggest strongly 

 that neither can spring from a scattered stock. I feel sure, therefore, that the mass of the eggs must 

 be laid somewhere where we have not explored at all, or not explored very thoroughly, and the 

 inaccessible south-western side of the Weddell Sea and the slope waters of the continental land seem 

 the most likely places. For it must be significant, in so far as the slope waters are concerned, 

 that on the rare occasions when we approached the continental land, as for instance at Stations 

 1671, 1713, 2600 and 855 (p. 93, Table 14), Metanauplii and First Calyptopes, obviously sprung 

 from recent deep hatchings, were encountered rising towards the surface, and that on the still 

 rarer occasions when we worked right on the slope itself (Stations 1662 and 2603) we struck the 

 Nauplii and Metanauplii very close to the bottom. 



To maintain the krill population at its existing fantastic level, having regard to the enormous 

 wastage it must suffer during the larval phase, the number of eggs liberated annually must in fact be 

 astronomically large.^ Thorson (1946), for instance, suggests that the 478 or more million eggs released 

 in under 5 months by the Californian sea hare (MacGinitie, 1934, 1935) are produced with the expecta- 

 tion that only one of the resultant larvae will survive to reach maturity. Somewhere, therefore, in these 

 southern waters, at certain times and in certain as yet it seems unexplored situations, there must be 

 vast masses of eggs, presenting an easy target for the right apparatus if used at the right depths, in the 

 right places and at the right time of year. Whether in fact our apparatus is the right one or not, 



^ Parrish, Saville, Craig, Baxter and Priestley (1959) estimate that a single patch of herring eggs measuring 320 by 320 m., 

 distributed in an almost continuous carpet over the sea bed, at its maximum might contain 258,000 million eggs. 



23-2 



