550 The Question of Valency in Gaseous Ionization. 



they may altogether make the number appear several times 

 too small. 



Nevertheless, it may certainly be said that in the experi- 

 ments o£ Millikan and Fletcher only very few doubly- 

 charged ions existed. As this result differs, at least quanti- 

 tatively, from ours, and ours again from Townsend's, the 

 conclusion seems justified that the conditions for the gene- 

 ration of plurivalent ions are not yet sufficiently known. 

 To throw some more light on them, we think it necessary to 

 vary the experimental conditions widely. One of the points 

 to be considered, as has been shown by Townsend, is the 

 secondary radiator. In our experiments a secondary radiation 

 may have come from lead, wood, aluminium, brass, or 

 paraffin. This may explain the discrepancy between Towns- 

 end's and our experiments. It is not at all impossible that 

 it may be necessary to use a homogeneous radiation of distinct 

 frequency, such as is, according to Barkla and Sadler, 

 originated in the secondary and tertiary Rontgen radiation. 

 It may have happened that such a radiation has been gene- 

 rated in different amounts in the experiments of the different 

 authors. Even the secondary cathode radiation may prove 

 itself important. We want also to point to the fact that 

 certain impurities in the gas, such as ozone or nitric oxides 

 or other electronegative substances, seem to facilitate the 

 detachment of an electron from a molecule, as is shown by 

 some experiments now being carried on by one of us. 



Finally, we want also to express the opinion that the results 

 of Millikan and Fletcher still leave the question an open one, 

 the more so since in other parts of physics observations of 

 multiply-charged positive ions are actually beginning to 

 accumulate. Thus the double charge of a particles is now 

 beyond any doubt : Gehreke and Reichenheim * have mea- 

 sured double charges in the anode rays of strontium ; and in 

 the Kanalstrahlen, J. J. Thomson f, v. Dechend & Hammer J, 

 and others have observed various atoms and molecules with 

 multiple charges. For example, J. J. Thomson observed 

 C++, 0+++, N++ 5> He++, Hg++, and Hg+++. There is 

 evidently no relation to valency. Also, in these cases 

 impurities seem to be important, as will be shown soon by 

 one of us. Farther, we have no reason to assume that there 

 should be any great difference between the ionization in an 



* E. Gehreke and 0. Beichenheim, Verh. d. JD. Phys. Ges. vol. ix. 

 p. 373 (1907). 



t J. J. Thomson, Jahrb. d. Bad. u. Elektr. vol. viii. p. 2 (1911). 



I v. Dechend and Hammer, Jahrb. d. Bad. u. Elektr. vol. viii. p. 1 

 (1911). 



