674: Prof. 0. W. Richardson on the 



Na + , indicating that the large ionization usually observed 

 with this substance is due to contamination with the salts of! 

 the alkali metals. A specimen of calcium sulphate, which 

 he examined in a similar way, gave values of e/m corre- 

 sponding to Ca + in air at all pressures below about 0*1 milli- 

 metre. The changes in the apparent value of e/m which 

 occurred at still higher pressures were such as would be 

 expected to arise from the interference of the gas molecules 

 with the free morion of the ions, and they do not indicate 

 any change in the nature of the ions themselves, at low 

 pressures, as the pressure of the gas is altered. This is true 

 for such very different gases as hydrogen, air, and carbon 

 dioxide. A full account of Mr. Davisson's work will be 

 published shortly, but it may be permissible to state that he 

 has devised a number of other experiments which fully con- 

 firm the results of those which have been described. The 

 evidence in fact leaves no doubt that only a negligible 

 fraction, if any, of the positive thermionic current from salts 

 is carried by ions which are not metallic atoms. 



I am aware that this is a standpoint which is diametrically 

 opposed to the views which have been expressed by a number 

 of recent investigators of the subject*. The direct evidence 

 to the contrary is, however, quite meagre. Hortonj has 

 already pointed out that the hypothesis of neutral doublets 

 put forward by Garrett J fails to account for one of the most 

 prominent figures of the case, which is, that these substances 

 emit a large positive iouization when the negative emission 

 is too small to measure. Garrett has also adduced direct 

 evidence that a considerable part of the emission from 

 aluminium phosphate is carried by hydrogen atoms. 

 Davisson has looked for this phenomenon, using a method 

 which I believe to be more reliable, and has been unable to 

 confirm it. 



Apart from the evidence brought forward by Garrett to 

 support his hypothesis of doublets and his experiment on the 

 magnetic deflexion of the positive ions, the remaining evidence 

 as to their nature is of an indirect character. The bulk of it 

 relies on the effect of various agents, chiefly gaseous, in 

 increasing the quantity of the emission. In the light of ex- 

 periments which will be described below, it would seem to 

 require great boldness to draw any conclusions as to the 



* Cf. Garrett, Phil. Mag-. [6] vol. xx. p. 573 (1910). Horton, Camb. 

 Phil. Proc. vol. xvi. p. 89 (1910) ; Proc. Koy. Soc A. vol. lxxxiv. p. 433 

 (1910). W. Wilson, Phil. Mag-. [6] vol. xxi. p. 634 (1911). 



t Camb. Phil, Proc. vol, xvi. p, 89 (1910). 



% Loc. cit. 



