382 On Ionization and Wireless Telegraphy. 



appreciable in air at ordinary pressures and temperatures 

 when under electric alternations of the right order of 

 frequency. And this was what was done. If the con- 

 ductivity had been inappreciable, it would have been harder 

 to believe in any increase o£ wave-velocity, since the same 

 theory involves both effects. As the conductivity proved 

 appreciable, it seems (to some minds at any rate) easier to 

 believe in the increased wave-velocity under the appropriate 

 conditions. And this is all that was claimed. 



The desirability of using air at low pressures had not- 

 escaped the notice of the experimenters, and was among the 

 matters discussed by them before starting the work *. The 

 line of attack then made was that which seemed best for 

 a start, but was undertaken with the full knowledge that 

 it by no means represented the conditions supposed to obtain 

 in the higher regions of the atmosphere. 



Any one experimental method in the laboratory will 

 probably fall short of full correspondence with Nature 

 in many respects, and can but throw a side light on the 

 issue. Thus the theory of Eccles supposes long-distance 

 wave propagation in air at low pressures and temperatures 

 and ionized by the sun ; whereas the experiments cited 

 deal only with oscillations of like frequency in air under 

 ordinary pressures and temperatures and ionized by radium 

 or X-rays. Hence, the conditions of the experiments (as 

 clearly shown in the paper) differ in three respects at least 

 from those supposed to hold in those regions of the atmo- 

 sphere to which the theory applies. Accordingly, any 

 statement as to the experiments yielding some support to 

 the theory should be taken in the light of the context. If 

 isolated from the context such a statement might well 

 appear misleading. 



Probably no one piece of work can either wholly confirm 

 or disprove such a theory as that of Eccles, but can only 

 form one link in a chain needing other links for its com- 

 pletion as a conclusive confirmation or disproof. 



In the meantime, if a moderate ionization produces a 

 detectable conductivity in air at ordinary pressures and 

 temperatures under electric oscillations of two million per 

 second, it may well seem less difficult to suppose that the 

 special effects needed by the theory of Eccles for ionic 

 refraction are forthcoming under the conditions obtaining in 

 the atmospheric regions in question. 



Accordingly, in this respect and to this extent, the theory 

 seems favoured rather than disfavoured by the experiments^ 



Nottingham, January 16, 1914. 



* See p. 563, Phil. Mag-. Oct. 1913. 



