236 DISCOVERY REPORTS 



on the surface (p. 149), although perhaps of considerable lateral extent, are disposed in shallow rafts 

 or plates no more perhaps than a metre or two thick in which the vertical nets, fishing only momen- 

 tarily, capture far smaller numbers than they would obviously do if the swarms were densely disposed 

 in depth. The orders of abundance of the Nauplii, Metanauplii and First Calyptopes taken at deeper 

 levels are also (Fig. 49) inclined to be low, suggesting that from the outset the larvae are assembled 

 in pockets of shallow draught. 



°/ 



'° DEPTH 



100 • 



7S- 



1000 - 750 



75- 



T 1 r 



10-100 100-1000 1000 -10000 lOOOO-IOOOOO 



Fig. 49. Orders of abundance of deeper living E. superba larvae and Ctenocalanus vanus expressed as percentages 

 of the total samples from the Warm Deep current (see legend to Fig. 45). 



Further investigation, involving a comprehensive and elaborate programme of staging and 

 measuring, may reveal that other plankton animals, like the krill, throughout much of their lives, are 

 assembled discretely on or near the surface in swarms, and I believe that a great deal of the patchiness 

 of the plankton as a whole^ will eventually be explained in terms of this phenomenon. In so far as 

 our Discovery collections are concerned, one is struck by the frequency with which the surface 

 (0-5 m.) nets produce enormously larger samples of certain species than are produced, for instance, 

 by oblique nets hauled open to the surface or by horizontal or oblique closing nets fished at subsurface 

 levels. One is struck, too, by how often these enormous surface samples are severally composed of 

 individuals that to the eye at least appear to be all of much the same size and in much the same 

 developmental phase.- Among the euphausians E. crystallorophias (p. 124) is a known surface swarmer, 

 and Thysanoessa macrura may well prove to be another, the recent discovery by the Japanese (Nemoto 

 and Nasu, 1958) that this species sometimes contributes substantially to the diet of baleen whales 



1 As Cassie (1959) has recently written, 'It is now well known and generally accepted that many, if not most, plankton 

 organisms show a marked departure from randomness in their horizontal distribution, even when samples are taken at 

 relatively close intervals'. 



2 Baker (i960) has just called attention to the shoaling of young squid {Ommastrephes pteropus Steenstrup) at the surface 

 at night, noting that ' those tending to form shoals were mainly in the smaller size groups and their size did not appear to vary 

 greatly within any one shoal'. See also Zelikman (i960) on the swarming of euphausians in the Barents Sea. 



