Barss — Ionization by Collision in Gases and Vapors. 241 



Investigations were next made with sulphur dioxide and 

 alcohol vapors for two different pressures in each. The results 

 were of the same general nature as those obtained in air, thus 

 showing that the positive ion in vapors plays a very small part 

 in ionization by collision until the sparking potential is almost 

 reached. 



So far, the anomaly presented by the behavior of vapors has 

 not been definitely explained. It may be due to some instabil- 

 ity in the molecular structure of the vapors when they are acted 

 upon by alpha radiation ; in other words, the alpha radiation 

 may be effective in producing a chemical change. 



It is more probable that the anomaly may be explained by 

 supposing the formation of aggregate molecules which act in 

 the capacity of a thin metal wire 1 in carrying a conduction cur- 

 rent through the vapor. The continuity of these aggregate 

 molecules would be more stable when the plates are closer 

 together than when they are farther removed ; and hence we 

 would expect the current to increase as the distance between the 

 plates becomes less. This fact experiment shows to be true. It 

 is also possible that the formation of these aggregate molecules 

 will be a function of the pressure, the number formed decreas- 

 ing as the pressure becomes less until finally a pressure is reached 

 when the current is not increased by this conduction effect, but 

 is entirely due to ionization proper. 



Summary. 



The collision constants are shown to be independent of the 

 nature of the ionizing agent ; therefore the negative ions gener- 

 ated in a gas by collisions with alpha particles are identical 

 with those generated by Rontgen rays and ultra-violet light. 



The theory of ionization by collision as developed for gases 

 holds for vapors ionized by alpha particles provided the pres- 

 sure is not too great. As the pressure is increased, evidence of 

 irregularity is obtained which seems to point to some instability 

 in the molecular structure of vapors, either due to the forma- 

 tion of aggregate molecules or to some chemical change. 



In vapors, as in gases, the negative ion alone is effective in 

 producing new ions by collisions with neutral molecules, until 

 the sparking potential is approached. 



In concluding this paper, I wish to thank Professor E. M. 

 "Wellisch of the Sloane Laboratory for his many valuable sug- 

 gestions and for his kindly interest throughout the experiment. 



