secondary acoustic targets are certain to pass through a 

 size permitting a brief period of resonance in response 

 to the sonar frequency in use. The additional sonic targets 

 then, multiplying the regional populations of scatterers at 

 any given moment and with some of them in resonance, 

 must contribute to the increase in scattering intensity 

 which is so routinely seen on precision depth recorders 

 during the crepuscular rise of the DSL. 



Conversely, when the scattering layer migrates 

 downward in the early morning hours it is often seen to 

 diminish in intensity, and occasionally to fade out entirely 

 for some period of time. A major cause of this phenomenon, 

 of course, must be inverse-square loss resulting from in- 

 creased range between signal source and target. However, 

 it is not impossible, in the light of estimates of refilling 

 time for the float based upon physical work equations, that 

 the siphonophores undergo a period when, in fact, their 

 pneumatophores are not fully inflated and therefore respond 

 poorly as sound reflectors. The same postulate could also 

 be put forth in the case of fishes associated with the DSL 

 which possess swim bladders. 



Maximum possible rates of gas secretion still re- 

 main a mystery, but on the basis of diffusion studies it is 

 reasonable to assume that some production of carbon monox- 

 ide must go on at all times simply to counter diffusive loss. 

 If this is true, then the lack of evidence for CO production 

 by most of the floats presented in tables 1 and 3 suggests 

 the unavailability of a suitable substrate for gas secretion. 

 This means that the figures presented for rate of oxygen 

 consumption^ although evidencing considerable spread, are 

 representative (under the circumstances of probable contin- 

 uous decline in float condition) of a "basal" or nonsecretory 

 respiratory rate. In the sea the siphonophores may be non- 

 secretory (although possibly still resisting diffusive loss) 

 only during the vertical ascent and while in the ascended 

 layer. It may well be that during the remainder of their 

 existence they possess a respiratory rate more closely 

 approximating or very likely surpassing that of the pneu- 

 matophore in Experiment 6. 



The miscellaneous observations of rapid filling 

 times for siphonophore floats, made by various workers 

 during the last 100 years, have necessarily all been made 



28 



