Contemporary Advances in Physics, XX 

 Ionization of Gases by Light 



By KARL K. DARROW 



The subject of this article, the ionization of gases by ultraviolet light, 

 is a narrow but singularly inviting department of modern physics. The 

 obstacles to experiment are so great that they are only now being overcome 

 by the latest improvements in laboratory technique; nearly all the valu- 

 able data are of recent acquisition, and the period of discoveries is not yet 

 past. Some of the results afford excellent confirmation of present atomic 

 theory; others are still obscure and challenging. 



THE subject of this article is more narrowly restricted than those 

 of many others of the series. It is narrower even than the title 

 might imply; for by "ionization" I mean for the present only the 

 detachment of the most loosely bound electron of a molecule or an 

 atom, and by "light" only the waves of the visible spectrum and that 

 adjoining range of wavelengths to which the name of ultraviolet is 

 customarily confined. Either of these limitations is implicit in the 

 other; for though most molecules are fashioned with electrons bound 

 with varying degrees of tightness, and the removal of any one thereof 

 is an act of ionization, it is beyond the power of such light- waves to 

 abstract any except the loosest. Perhaps it will be found instructive 

 if for so definite and circumscribed a problem I relate the methods 

 of experiment, the data of the experiments, the simple theory, and 

 the artifices which have been conceived to reconcile the theory and the 

 data, sometimes with success and at other times in vain. 



Most of the really valuable data are of recent acquisition, for there 

 are difficulties hampering the attack upon the problem, which the 

 progress of laboratory technique is only gradually clearing away. 

 Consider, for instance, the question of providing the light. It is de- 

 sirable to be able to illuminate the gas with monochromatic light of 

 any wavelength, photons of any energy. When ionization by electrons 

 is being studied, one varies the energy and the wavelength at will by 

 varying the voltage impressed on the electrons. With light, this is not 

 within our power ; one has to take the quanta as they are supplied by a 

 luminous source. 



If the spectrum of the source consists of bright lines widely separated, 

 the ionization which any of them alone produces may be measured, 

 and the energy of the quanta is very narrowly defined. On the 

 curve of ionization vs. frequency, then, every spectrum line supplies an 



341 



23 



