536 HERSEY AND BACKUS [CHAP. 13 



[FIFo= (P/Po)^''^] has been observed during both a sunrise and a sunset migra- 

 tion. This seems to imply that retention of the swim-bladder is worthwhile 

 even though the fish is at neutral buoyancy only at the top of its depth range. 

 The relationship that would obtain were the swim-bladder contents being 

 adjusted to maintain neutral buoyancy [P/Po = (P/Po)'"^] has been observed 

 during a sunset migration. Thus it seems that some bathy pelagic fish is able to 

 absorb swim-bladder gas at a rate which maintains its neutral buoyancy 

 during an ascent of 200 m in about 1 h. It is probable (though we have not yet 

 demonstrated it) that this animal can effect the converse operation during the 

 sunrise descent. Since some of the steps in the arguments that these per- 

 formances are physiologically improbable are extrapolations from the func- 

 tionings of surface-living fishes, it seems proper to us that the problem now be 

 returned to physiologists for further consideration. 



D. Deep Scattering Layers and Energy Distribution in the Ocean 



Scattering-layer study suggests that a very sizeable portion of the macro- 

 scopic animals in the upper 1000 m of the water column engage in an extended 

 diurnal vertical migration. The effect of these migrations in distributing energy 

 downward from the euphotic zone throughout the upper 1000 m must be 

 considerable and has not been generally appreciated. As the animals in these 

 layers are identified and their populations are measured, estimates of this 

 energy flow can be made. Since the speed and direction of currents may vary 

 widely throughout the upper 1000 m of the water column, the horizontal 

 transports of energy due to these migrations may also be important considera- 

 tions, 



E. Extending Acoustical Observations 



At present there are instrumental limitations (especially in transducers) 

 which have restricted studies of the frequency dependence of sound-scattering 

 in the sea to the band from 1 or 2 kc/s to around 30 kc/s. Observations made in 

 this band indicate that these studies could be pursued profitably down to the 

 region of a few hundreds of cycles and up to the neighborhood of 100 kc/s. The 

 most satisfactory means of generating broad-band sounds at present is with 

 small explosive charges. It is evident, however, that many studies of the 

 frequency-dependence of sound-scattering could be better done with devices 

 generating complex sound signals repetitively, such as those recently come to 

 use in seismic studies (see Volume 3). It only remains to make these devices 

 efficient enough sound-generators so that they may be used with such relatively 

 weak reflectors as scattering layers. 



Broad-spectrum scattering observations could be extended to many parts of 

 the oceans and the generalizations drawn from comparing them should be 

 useful in delimiting the faunal provinces of the upper 1000 m of the high 

 seas. 



