6. Measured rates of diffusive loss of carbon monoxide 

 from siphonophore floats, when used to calculate diffusion 

 constants for the float walls, give values very close to those 

 determined for chitin, indicating that it is most likely this 

 material which provides the diffusion barrier in the float 

 wall. 



7. Calculations of the energy required to perform the 

 physical work of countering hydrostatic pressures at the 

 daytime depths of the DSL indicate that the siphonophore 

 may very well require more than the observed one hour 

 needed for refilling the float at a pressure of one atmosphere. 

 This is not surprising since the volumes of gas which must 

 be produced are 30 to 40 times as great. 



8. It should be noted that work estimates have been 

 based upon what may actually be "basal" or nonsecretory 

 levels of oxygen consumption. Observations from a single 

 experiment indicate the possibility that energy production 

 from aerobic respiration may rise to at least double the 

 "basal" level while CO is being produced. The time required 

 for refilling the float at high pressures may thus be some- 

 what less than calculations indicate. 



9. Comparisons of data obtained from intact siphono- 

 phores and from automatized floats indicate the validity of 

 results obtained from tests performed upon the latter. 



30 



