Resonance Potentials for Electrons in Metallic Vapours. 41 



subordinate series of single lines. Thallium, group III., 

 behaves in a still different manner, as will appear below. 



The following table shows the critical wave-lengths deter- 

 mining resonance and ionization for the metals of group II. 



Table IV. 



Metal. l'5S-2p 2 . 1'5S. 



Mg 4571-38A 1621'7A 



Zn 3075-99 1319-95 



Ccl 326017 • 137869 



Hg 2536-72 1187-96 



The ionization potential is determined for all of these metals 

 by the lines in the third column, and the resonance potential 

 by the lines in the second column, if A, = 45 71 were the 

 single line spectrum of magnesium, the second column would 

 .give correctly the single line spectrum of these elements 

 of the same family. The experiments of Prof. McLennan 

 do not permit the above classification of the elements of the 

 same group as regards the single line spectrum, although 

 the classification is correct as regards ionization and resonance 

 potentials. 



When the potential difference between the cathode and 

 net was equal to or greater than the ionization potential, 

 about 7*8 volts, a strong arc was formed, and the entire 

 ionization-chamber was rilled with a green glow. A faint 

 blue-green glow, probably \=4571, could be detected at 

 potentials as low as 5*5 volts. This fact is a strong argument 

 that the line 2852 is not the single line spectrum of magnesium ; 

 for if so, one would not expect the line 4571 to appear below 

 7*8 volts. On the other hand, if the line 4571 were the single 

 line spectrum, one would not expect the line 2852 to appear 

 below 7*8 volts. McLennan observed the latter line at 

 5*9 volts. A possible explanation of these contradictory 

 results would be had if two resonance potentials existed, one 

 corresponding to \ = 4571 and the other to \ = 2852. We 

 have no experimental evidence of this fact. We would 

 further expect that if two resonance potentials existed for 

 magnesium, we would have two resonance potentials for 

 cadmium, mercury, and zinc. Experimental evidence here 

 again is to the contrary, although in the case of zinc 

 peculiar maxima were found in the total current curves, the 

 exact meaning of which has not been determined. 



If, referring to the values for magnesium in Table IV., we 

 apply the quantum relation hv = eV> using h = 6*547 . 10" 27 

 <erg sec, we obtain the theoretical values of 2*70 volts for 



