936 Dr. R. D. Kleeman on Disintegration of an Ion 



Indications as to the Nature of the Negative Ion Clusters in a 

 Gas and the Magnitude of their Mean Free Paths of 

 Disintegration. 



It will have been observed that the curves obtained on 

 plotting the logs of the currents against the electric field 

 show more or less marked kinks. They appear to indicate 

 that for the corresponding electric fields the average or mean 

 free path of dissociation of a cluster is equal to the distance 

 between gauze and plate. We have seen that the mean free 

 path of a cluster should increase with increase of electric field 

 till it reaches a maximum and then decrease. Therefore, 

 when the electric field is very small the average mean free 

 path will lie between gauze and plate. As the field is in- 

 creased the average path gradually becomes equal to the 

 distance between gauze and plate. Therefore a discontinuity 

 should occur in the current corresponding to this point. This 

 discontinuity is rendered less marked than it would otherwise 

 be by the fact that the clusters drawn through the gauze 

 are in different stages of disintegration through molecular 

 collision, and that the possible free paths of a cluster are not 

 equal to one another but are grouped about a mean free path. 

 The tangent to the current curve at this point would evidently 

 therefore undergo a sudden decrease in value with increase 

 of field. For larger fields than corresponding to this point 

 the mean free path is greater than the distance between 

 gauze and plate, the maximum path defined being in this 

 case greater than this distance. But if the field be 

 sufficiently increased, a point will be reached when the 

 average path is again equal to the distance between gauze 

 and plate. As before, this would be attended by a dis- 

 continuity in the current. Thus each kind of cluster drawn 

 through the gauze gives rise to two kinks in the logarithm 

 current curve. These kinks may be called first and second 

 order kinks respectively. From the discussion of the average 

 path, it appears that corresponding to a kink of the first order 

 the disintegration of clusters is principally due to bombard- 

 ment by neutral molecules, and corresponding to a kink of the 

 second order principally due to the action of the electric 

 field. The effect of ionization by collision is smaller, corre- 

 sponding to a kink of the first order rather than to a kink of 

 the second order, since the electric field is smaller in the former 

 case, and the former kink should therefore be less conspicuous 

 than the latter. The kinks produced by one kind of cluster 

 would very probably not coincide with those produced by 

 another kind. Thus experiments of the nature carried out 

 might furnish some information on the number of different 



