Pickwell. 



Finally, it is necessary to point out that in order for bubbles 

 released at depth to be comparable in size to those observed at am- 

 bient pressure (1 atm), an additional mass of gas must be present 

 within the bubble. This is almost certain to be the case since the 

 pneumatophore, in maintaining relatively constant volume, must 

 secrete additional gas. For example, at a hydrostatic pressure of 

 30 atmospheres (300 m depth), a float of 1 mm3 volume must secrete 

 a total of 31 mm-^ of CO to maintain this volume. The range in ini- 

 tial size of released bubbles is thus probably comparable to that given 

 in Figs. 3 and 4. 



SUMMARY 



1. Physonectid siphonophores are potentially responsible for res- 

 onant scattering from the DSL partly due to their gelatinous, gas- 

 filled floats, and partly due to bubbles expelled from these floats. 



2. Voluntary expulsion of bubbles by siphonophore floats has been 

 observed and photographed. Range in volume of expelled bubbles 

 was 0. 03 mm3 to 2. 51 mm^. Range in volume of pneumatophores 

 observed to expel bubbles was 0. 25 mm^ to 12. 55 mm^. 



3. Observations on siphonophores from a deep submersible vehicle 

 at the depths of the DSL revealed their pneumatophores to be inflated 

 at all times observed. They are thus continuously potential sound 

 scatterers. 



4. Range in theoretical resonant frequencies based on the initial 

 size of the expelled bubbles at 100 and 400 meters depth was 57 to 

 111 kc/ s for the smallest and 12 to 24 kc/ s for the largest. A simi- 

 lar range for bubble expelling pneumatophores was 27 to 50 kc/ s for 

 the smallest and 7 to 13 kc/ s for the largest when calculated on the 

 basis of float volume. 



387 



