278 DISCOVERY REPORTS 



2146 krill taken by oblique pairs below 100 m. In one of these the N70B catch was slightly larger 

 than that of the N looB, in another it was distinctly larger. In the third instance the N looB produced 

 by far the larger catch. 



It is somewhat surprising that for the deep night pairs the N70B produces the slightly larger 

 average catch-figure. This result, however, is heavily biased by the three relatively large paired 

 gatherings to which I have already called attention, and it simply means that in deep water the 

 N 70 B sometimes strikes and samples the deep and evidently very rarely occurring rafted swarms with 

 better effect than the NiooB, just as it occasionally does above 100 m. 



CORRECTIONS APPLIED TO THE CATCH-FIGURES 



We have seen that by night or by day the surface population of larvae and early adolescents can be 

 sampled with reasonable consistency and in significant amount in one or other of two ways, (i) by 

 the i-m. diameter stramin net towed horizontally on the surface, and (2) by the same net hauled 

 obliquely open to the surface from a depth of approximately 100 m. Good samples as a rule can be 

 obtained by either method. Both methods can be used too to sample the older krill, the surface net 

 however sampling them effectively only at night when avoidance by these active animals is at a 

 minimum and they are heavily massed it seems above the lo-m. level. Owing to the vertical migrations 

 undertaken by the older swarms by day, and the comparative ease with which such as remain on the 

 surface avoid the conventional stern net, we cannot, or virtually cannot, sample the over 20 mm. 

 population in the ordinary way with the surface net by day. We can, however, sample it on the surface 

 with enormous success in daylight using a lateral net, or a net fishing far back in the wake. We can 

 sample it too with fair success with the oblique net both by night and by day. 



At many stations following the reintroduction of the surface (0-5 m.) net in 1932 (p. 54) both this and 

 the oblique ( 1 00-0 m.) net were fished simultaneously. Even after this practice became general, however, 

 it was often impossible to keep it up, heavy seas frequently preventing the shooting of the surface net and 

 leaving the oblique net to be fished alone. Because of this, and because the practice did not in fact become 

 a well established routine until early in 1935, there are many more oblique samples than surface ones. 



Comparison of the catches of the under 16 mm., 16-20 mm. and over 20 mm. classes of the surface 

 population obtained in surface and oblique nets fished simultaneously shows that if the daylight 

 gatherings of the staple population be disregarded, far more often than not the surface nets both by 

 day and by night produce significantly larger samples than the oblique. Details of these simultaneous 

 samplings arranged in the descending order of abundance of the surface catch when it was the larger 

 (or at any rate not the less), and in the descending order of abundance of the oblique catch when it was 

 the larger, are set out in Tables 57-9. 



Taking the three size groups in turn it will be seen that of 138 simultaneous samplings involving 

 the under 16 mm. class the surface catch was the larger in 109 instances, the smaller in 27 and the 

 same as the oblique catch in two. Similarly, of 1 1 1 simultaneous samplings involving the 16-20 mm. 

 class the surface catch was the larger in 96 instances, the smaller in nine and the same as the oblique 

 catch in six; and of 208 simultaneous samplings involving the over 20 mm. class the surface catch 

 was the larger in 164 instances, the smaller in 40 and the same as the oblique catch in four. It will 

 be seen, too, that on the few occasions when both catches were the same the numbers involved were 

 negligible, and that in the great majority of instances where the oblique catch was the larger the catch- 

 figures on the whole were also negligible or at any rate not very large. The oblique catch in fact was 

 greater on only six occasions involving the larvae (Table 57) and on only two occasions involving the 

 adults (Table 59) where the krill were really abundant and hauls exceeding 1000 were obtained. 



