442 FOOTE AND MOHLER: IONIZATION 



These data, plotted logarithmically, give a straight line, the 

 slope of which determines 5. The following values expressed 

 in kg. cal. were used in the above computations. 



Landolt-Bornstein-Meyerhoffer Tab. 

 From data in Kaye and Laby 

 Pier 



Mohler and Foote, et al. 



Bom (confirmed experimentally by Foote and 

 Mohler) 



From data of Stelzner and Niederschulte 

 From data of Stelzner and Niederschulte 

 Landolt-Bornstein-Meyerhoffer Tab. 

 Kaye and Laby, actually 0.6 

 On substituting these data in equations (8), (9) and (10) we 

 obtain : 



y'Hgci = 203 kg. cal. =0= 8.8 volts (11) 



Jugch = yngci - 42 =^ Uugci - 1.8) volts (12) 



/Hgci. = /'ng - 79 - (123451'. 10-8 + 7.0) volts (13) 



where 



V = i.5g», since J'^g = Jug + i2345j'.io-^ 

 The value of the spectral frequency v = 1.5^ for mercury 

 does not appear in the literature, although Sommerfeld^ has 

 determined this frequency for zinc as 159000. It represents 

 the frequency of the quantum involved in the removal of the 

 second electron from the metal ion. 



It would therefore appear that mercurous chloride may be 

 ionized by dissociation into (Hg)+ and (Cl)~ at 8.8 volts. A 

 second type of ionization, Jugch into (HgCl)"*" may occur, but 

 the value cannot be computed by the methods outlined. This 

 would require a knowledge of spectral series of HgCl, which at 

 present is not available. 



The ionization of mercuric chloride into (Hg)"*"^ and 2 (CI)" 

 requires an amount of work eV where V = (123451'. io~^ + 

 7.0) volts and v = 15^, but this type of ionization could not 

 exist, in ordinary arc phenomena, except by a two-stage process, 

 and then but rarely. A far more probable type of ionization is 



'• Referred to by Born. Zeit. f. Phj'sik 1: 252. 1920. 



