288 Dr. Norman Campbell on the 



that the apparent ionization potential of the metal is pre- 

 cisely that of the gas which would be expected to be present, 

 namely hydrogen. It is known that the two other forms of 

 current which can be made to leave a metal surface, the 

 photoelectric current and the thermionic current, are pro- 

 foundly influenced by a treatment o£ the metal which is 

 usually regarded as tending to free it from gases, and espe- 

 cially hydrogen. The earlier observers of the ionization 

 potentials of metals did not subject them to any such 

 treatment, and it is possible that the true ionization potential 

 of metals has never been observed. 



But if the true ionization potential could be observed, 

 some light might be thrown on the most difficult problem of 

 modern physics," the photo-electric effect. If the photo- 

 electric effect represents an action of the light on the metal 

 itself, then we should expect perhaps that metals should be 

 more easily ionized than gases ; they should have a lower 

 ionization potential, and probably yield more electrons for a 

 given beam of incident rays. For the photoelectric effect 

 certainly occurs with metal in the presence of light of a 

 wave-length much longer than that which affects gases; 

 according to all prevalent theory a longer wave-length is 

 associated with the absorption of a smaller energy in the 

 individual atomic reaction. On the other hand, the argu- 

 ments of Fredenhagen and the experiments of Kiistner * 

 have been advanced to prove that a pure metal surface does 

 not exhibit any photoelectric effect at all ; they would 

 suggest that metals should be less easily ionized than gases. 



At any rate, it appeared that the investigation of the effect 

 on another phenomenon of the treatment which influences so 

 profoundly these emissions of electrons from metal surfaces 

 would probably lead to some interesting results. The preli- 

 minary results of such an investigation are here set forth. 

 As might be expected, the matter has proved somewhat 

 complex, but certain facts seem indubitable and of some 

 importance. 



3. The apparatus employed is shown diagrammatically in 

 fig. 1. A and B are brass cylinders ; W a strip of platinum 

 10 x 3 mm. coated on the upper side with lime and supported 

 by the wires c, d which lead in the heating current ; P is a 

 strip of the same metal, 15x15 mm. also supported by 

 leads a, b. A, B, P, W are all insulated from each other 

 and from earth, and could be connected either to a battery 



* Fredenhagen, Phys. Zeit. xv. p. 65 (1914) ; Kiistner, ibid. xv. p. 68 

 (1914). 



