THE OLDER STAGES i6i 



held at the surface before being paid away, considerably larger numbers might be captured, an 

 unknown proportion of which might well be retained as the net, still maintaining the measure of 

 fishing strain necessary to keep it clear of the warp, was lowered to fish in deeper water. The practice 

 in oblique fishing of shooting nets in pairs, a stramin net (N looB) with a silk net (N70B) immediately 

 above it, renders the first net to enter the water, the N looB, still more liable to error through surface 

 fishing. For in this operation the stramin net would be shot and kept fishing on the surface for 

 upwards of 30 sec. or a minute,^ and sometimes for even longer if the operation were bungled, while 

 the N70B was attached to its closing mechanism, shot and got clear away. It may be noted too 

 that the numbers of krill caught in a subsurface net, should it happen to be the lowermost of a flight 

 of three or more on the same warp, could be suspect on the ground that the catch could have been 

 made or contributed to, not only at the surface, but also, before the net reached its intended level, 

 possibly augmented on every occasion when the warp was halted for the attachment of the inter- 

 mediate nets. 



o 

 10 

 40 



70 



C 130 

 O 150 

 190- 

 220- 

 250 



r 



AVERAGE CATCH DAY 

 500 1000 I500 

 I I I 



150 

 14 

 71 



7a 



43 

 40 

 22 

 10 



2 



AVERAGE CATCH NIGHT 

 SOO 1000 I50O 



2000 



Fig. 17. Diurnal vertical distribution of the staple whale food, the number of net hauls made at each 

 10 or 30 m. depth interval being shown in the column of figures in the centre. 



Diurnal vertical distribution 



The diurnal vertical distribution of the staple whale food is shown in Fig. 17 in which the 

 histograms representing the average catches at all levels investigated by day and by night have been 

 constructed exclusively from data obtained at stations where open horizontal nets on the surface 

 and closing horizontal nets in the subsurface strata were fished simultaneously. Here again for the 

 subsurface nets, as in Tables 27-29, there has been a grouping of depths of tow into selected broad 

 horizons, in this instance horizons of 30 m. vertical range. The salient features of the diurnal 

 distribution thus disclosed are (i) the great abundance in which the krill are concentrated in the surface 

 (o-io m.) stratum at night, and (2) their apparent absence, or virtual absence, from this stratum 

 by day. At first sight, supposing there had been no simultaneous deep-level observations to accom- 

 pany the surface series, this would suggest that E. superba was a typical vertically migrating animal, 

 rising to the surface at night and seeking the deeper water by day. However, we see that the average 

 daytime subsurface concentration does not at any level, down to the deepest we have examined, even 

 remotely approach the average night concentration at the surface, and this suggests three possibilities : 

 (i) the older krill are migrating by day to deeper levels than are covered by our horizontal observa- 



1 In only 36 seconds, it will be recalled (p. 152), more than 140,000 large krill were taken by the boom (lateral) net used at 

 Station WS 540. 



17 



