CONTEMPORARY ADVANCES IN PHYSICS 355 



curves (obtained of course by applying the densitometer to the spec- 

 trogram) with sodium. On the other hand, the experiences of Ditch- 

 burn with potassium are not encouraging. Not only did he have to 

 shoot a jet of rapidly-distilling vapor across the beam of light, but he 

 was obliged to swamp it in a vast excess of nitrogen — partly to keep the 

 metal from boiling away in a rush, partly it seems to prevent the vapor 

 from attacking the quartz windows. The curves are very crinkly, 

 and it is difficult to tell what share of the absorption should be credited 

 to molecules and what to atoms. 



Nevertheless Ditchburn was able to deduce a value of the coefficient 

 k having the same order of magnitude — 10~^^ — as those which Mohler 

 and Boeckner had obtained with caesium and with rubidium when 

 they were measuring, not the disappearance of photons from the beam, 

 but the advent of ions in the gas. Mohler and Boeckner themselves 

 observed the absorption of light in caesium, and they found for k the 

 value 4-10~^^, — a good agreement, but they qualify it with the words 

 "subject to great uncertainty because of the low value of the total 

 absorption." Let it be pointed out in closing, that agreements such 

 as these are proof that in this region of the spectrum, photons ionize 

 when they are absorbed, and absorption is due to ionization. To 

 physicists familiar with the new atomic theories, this seems self- 

 evident, and scarcely worth the proving; but it is not self-evident, 

 and there was a time, not many years ago, when such a proof would 

 have been a sensational event. 



