i64 DISCOVERY REPORTS 



ever densely crowded, would in fact act as scatterers of sound. Marshall (195 1), for instance, thinks 

 that bathypelagic fishes, with their strongly reflecting swim bladders, are the primary agents, adding 

 (Marshall, i960) that 'it seems likely that the most prominent of the fish sound-scatterers will prove 

 to be myctophids'. Nevertheless, concentrations of euphausians have been found associated with 

 scattering layers (Hersey and Moore, 1948; Saito and Mishima, 1953) and although small and 

 lacking air bubbles they could well it seems be interceptors of sound beams especially if gathered 

 tightly in swarms. The causes of scattering, however, remain obscure, and as Johnson (1957) 

 pointedly remarks, ' Improved techniques of sampling with controlable gear are needed to solve the 

 riddle'. Moore (1958) writes, 'From the physical aspects, fishes, with their highly reflecting swim 

 bladders, seem the most likely [agents], although from the biological point of view much of the 

 evidence points toward the euphausiids. . . . Whatever the source, these scattering layer records 

 provide very valuable information concerning the details of diurnal migration '. The converse, that 

 the absence of such records provides no such information, pointing it might be to an absence of 

 diurnal migration, could equally be true, provided the phenomenon of deep scattering is of animal origin. 



Vertical distribution at Station 1835 



Further evidence of the apparently incessant surface massing of the vast majority of the krill and 

 of the absence in this species of any regular vertical movement is provided by certain operations that 

 were conducted to investigate these phenomena in October 1936. The work to be described was 

 carried out at Station 1835 in the western part of the Weddell drift, a locality which in October 

 (p. 391, Fig. 133) is rich in swarms of 10 and 22 month old E. superba. The procedure was as follows: 



Beginning shortly after noon the vessel steamed seven miles along a south-west course towing one 

 after another an ascending series of twelve i-m. diameter horizontal subsurface closing nets, the 

 deepest at 250 m., the remainder rising step by step (Table 33) to a level 10 m. below the surface, 

 the depth in each instance being measured by Kelvin tube. The towings were of 20 min. duration, 

 each being accompanied by a simultaneous 20 min. horizontal towing (Table 34) with a net of the 

 same kind on the surface. These simultaneous south-west-going surface and subsurface towings 

 covered the hours of afternoon daylight and the twilight period of sunset and dusk. They will be 

 referred to as series (a). On completion of this part of the programme the vessel was put about and, 

 now in full darkness, steamed 4 miles back along the course she had taken, towing as before simul- 

 taneous surface and subsurface nets, the latter, however, not rising but descending step by step from 

 the 10 m. level to a depth of 128 m. The night towings will be referred to as series {b). 



In view of the extreme patchiness of the krill it is possible that had the ship continued along her 

 south-west course on completion of series {a) she might have passed from an area rich in swarms to 

 one where they were scarce or absent (or vice versa) and the resulting data, in so far as they could be 

 used to compare night conditions with day, would have been valueless. To circumvent such a possi- 

 bility, therefore, and to ensure that the population sampled by night was in fact the same as that 

 sampled by day, the ship was deliberately put about, covering in series {b) a substantial part of the 

 same water mass covered in series {a). 



For about a week prior to this investigation we had been striking dense swarms of young 10 month 

 old krill both by day and by night in the surface waters through which we passed. Because of this 

 certain precautions were taken in order to prevent, or at any rate reduce, contamination of the sub- 

 surface samples with krill that might possibly (p. 160) have come from the surface, the long established 

 practice, with its frequent warp stoppages, of fishing the nets in multiple flights, that is, several on the 

 same warp, being abandoned in favour of a subsurface series shot singly, the nets in each instance 

 being passed rapidly through the surface zone and paid away without interruption and at maximum 



