4!)S 



Bulletin 167 



The rate of gas production in the closed arms of the Smith 

 tubes was followed by measuring with the Frost card. The 

 amount of carbon dioxid formed was determined approximately 

 by absorbing it within the tube with sodium hydroxid, the volume 

 being noted before and after absorption. The residue was sub- 

 jected to the flame test for hydrogen. A positive reaction in- 

 variably resulted, although the explosion was frequently so faint 

 as to arouse suspicion that the material tested might be a mixture 

 of gases rather than pure hydrogen. The tabulated results fol- 

 low. The figures express the percents of the closed arm of 

 Smith tubes which were filled with gas. 



