usually within an hour. Apparently some of the floats, 

 though severed from the remainder of the colony, still 

 possessed enough substrate to permit secretion of enough 

 additional gas for reflotation. 



In neither of the above cases was all gas removed 

 from the pneumatophores. Jacobs cites no data but presents 

 a figure which shows about half the gas being expelled from 

 the siphonophore float. My vacuum extraction method 

 generally left a residual bubble of gas within the pneumato- 

 phores which was insufficient to provide flotation. 



A single direct observation on rate of gas produc- 

 tion was performed using the pneumatophore in Experiment 

 6. After measuring the float and enclosed gas, with an 

 ocular micrometer at 27x, a small bubble of gas was gently 

 extruded by mild pressure. The float was then placed in a 

 1-cc syringe with no gas phase and immersed in a water 

 bath at 7°C. The extruded bubble was measured for volume 

 and the contained gases analyzed by an ultra-micro tech- 

 nique 13 (Appendix A, fig. A2). 



At the end of the four-hour test period the syringe 

 was removed from the bath. The dissolved gases in the 

 1-cc of water were extracted and analyzed, and the remain- 

 der of gas within the float was extruded, measured and 

 analyzed. The results and a protocol for the rate calcula- 

 tions appear in Appendix C. 



Secretion of additional carbon monoxide is immedi- 

 ately evident from the 20 percent increase in that gas within 

 the float. By allowing for the total CO present at the start, 

 and adding the amounts of additional CO encountered within 

 the float and dissolved in the water in the syringe, a total 

 production figure for additional CO of 0. 31 mm 3 was found. 

 Thus, the pneumatophore produced approximately 0. 08 mm 3 

 of carbon monoxide per hour if production rate was linear 

 for this period. This equals a total volume somewhat in 

 excess of the starting volume and indicates that a float this 

 size may easily fill itself in a matter of a few hours. How- 

 ever, this still appears slow compared to the observations 

 noted above, even though the total gas phase was not removed 

 from the pneumatophore under observation. In addition, it 

 is not certain whether all of the gas was produced continu- 

 ously or in a burst during some period of the experiment. 



21 



