SECT. 4] SOUND SCATTERING BY MARINE ORGANISMS 533 



When the individual scatterers are not distributed at random, such as might 

 happen in schools offish, the possibility exists that extra strong back-scattering 

 may occur at selective frequencies which could depend on the spacing of the 

 scatterers. Such circumstances may explain the observations of very low- 

 frequency scattering that have occasionally been reported (e.g. Burling, 1956) 

 and perhaps also those of Ellis and Winterhalter (1956). The present authors 

 have recorded strong scattering in the shallow water of Vineyard Sound, using 

 the seismic profiler (to be described by Hersey in Volume 3) with a narrow band- 

 pass filter setting centered at 250 c/s. Simultaneously we observed the same 

 scattering group with a 40 kc/s echo-sounder. Although we did not sample this 

 scattering group, we have found on other occasions that ones of similar appear- 

 ance on the echo-sounder record are schools of fishes of a foot or two in length. 



6. What is "The Deep Scattering Layer" ? 



We cannot avoid a summary statement giving some tentative answer to the 

 now time-worn question, "What is 'the deep scattering layer'?" In proposing 

 that bathypelagic fishes with gas-filled swim-bladders might be the source of 

 the observed sound-scattering, Marshall (1951) lists four requirements which 

 an organism must meet before it can be considered seriously in this role. 

 (1) The organism must be a widely distributed one. (2) The organism must be 

 shown to exist in concentrations at the appropriate depth during the daytime. 



(3) The organism must show pronounced powers of diurnal vertical migration. 



(4) The organism must have the ability to refiect sound. Specifically, the product 

 of the organism's scattering cross-section and its population density must lead 

 to reverberation levels in agreement with those observed. 



Numerous studies, but particularly a number cited here made in connection 

 with the scattering-layer problem itself, show that several types of animals 

 fulfill the first three requirements. Conspicuous among these animals are 

 bathypelagic fishes, euphausids, and other crustaceans — those of the sergestid 

 type, for instance. Probably many other animals, about which good information 

 is presently lacking, will also be found to meet these requirements — squids, 

 for example. 



The acoustic scattering cross-section of fishes, crustaceans, squids and other 

 sorts of "fleshy" animals without air-bladders are subject to estimate well 

 within a factor of 10 for cases where some such animal of similar size-to-wave- 

 length ratio has been measured. This includes animals from about 2 to 30 cm in 

 length at frequencies from 10 to 30 kc/s. The acoustic scattering cross-sections 

 of fishes or other animals ^ containing gas bubbles are subject to accurate 

 estimate (within about 20%) if the size and shape of the gas bubble at one 

 atmosphere of pressure are known. 



1 In addition to fishes, a few other bathypelagic animals enclose gas bubbles ; for 

 instance, certain siphonoiDhores and the cephalopods, Spirula and Nautilus. These animals 

 inhabit that part of the water column in which deep scattering layers are observed but 

 very little is known about their distribution, vertical migrations and abundance. 



