The Question of Valency in Gaseous Ionization. 547 



Consider the similarity between the general equation for 

 the grating 



n\ — 2b cos 6 sin 8 



(in which h is the grating space, 26 the angle between colli- 

 mators, and 8 the angular deviation from the position of the 

 central image), and the equation for the retardation 



p = 2d cos 6 sin (23° -8). 



If d were equal to b, as is approximately the case, the 

 equations would be identical, if 8 were reckoned from the 

 point 23° out from the central image, instead of directly 

 from the central image. We may now say that an echelette 

 grating, in which these two effects are combined, behaves 

 from the standpoint of the distribution of the energy in 

 every respect similarly to the ordinary grating; the difference 

 being that, whereas in the grating the energy in the various 

 spectra on either side of the central image falls off in certain 

 ratios, in the echelette grating the point of reference is Ihe 

 oblique image, and the same ratios should hold for the energy 

 in the spectra symmetrically disposed about the oblique 

 image. The absent orders in the echelette spectra are 

 exactly paralleled by the absent orders with a flat grating, 

 any arbitrary portion of whose element is reflecting : the 

 remainder throwing off no energy. 



The arrangement of the fixed collimators was adopted by 

 the authors as the most logical one, and this is always the 

 easiest way to mount the grating for use in the infra-red 

 spectrum. 



To carry this work still farther, and make it productive 

 of still more practical results, this type of grating should be 

 thoroughly investigated in the visible region of the spectrum; 

 it is the intention of the authors to do this as soon as gratings 

 good enough for use in this region can be obtained. 



LI. On the Question of Valency in Gaseous Ionization. 



To the Editors of the Philosophical Magazine. 

 Dear StRS, — 



IN a paper entitled as above, Millikan and Fletcher * have 

 come to the conclusion that when a gas is being ionized 

 by means of radium or Rontgen rays, never more than one 

 electron is driven away from the molecule, so that there 



* R. A. Millikan and Harvey Fletcher. Phil. Mag. TGI vol. xxi. p. 753 



(1911). ' F 



