670 BELL SYSTEM TECHNICAL JOURNAL 



plored is very little, compared with what remains. As for the dis- 

 tribution-in-energy, the first step in determining it was taken some 

 sixteen years ago. It was indeed a great step ; for it led to the discovery 

 of the process of excitation, the transfer of energy from moving electrons 

 to atoms which shift these latter from their normal condition into one 

 or another of their "excited" states.^ But it was only a beginning; 

 the method had to be much modified and refined, to make it capable of 

 finding the answers to such questions as I have phrased above; and 

 the modifications were scarcely even imagined as lately as four years 

 ago. Hence the reader must not expect to be introduced to a very 

 great body of systematized knowledge. As for experiments on protons 

 and on other kinds of charged atoms, they too are all extremely recent. 



1 begin with the experiments on interception of electrons. 

 Suppose then that a beam of electrons is sent across a tube, having 



first been limited by a sequence of slits or holes so that it has a definite 

 contour, like the beam of a searchlight, which it retains all the way 

 across the tube if there is vacuum. Further, suppose that on the far 

 side of the tube there is a collector just large enough to swallow up the 

 entire beam so long as it does not spread, but no larger; or alternatively, 

 a collector covered by a screen pierced with a hole just large enough 

 to let the beam, or a fraction of it, pass through. Even so, the result 

 may depend on the diameter of the beam in a way which the reader will 

 see for himself later on; it is best to think of a very narrow pencil of 

 corpuscles.^ 



When a gas is introduced into the tube, the current into the collector 

 will decrease. The decrease will be proportional to the density of the 

 gas, so long as this is not too great; and it will be possible to define a 

 "cross-section of the molecule for interception of electrons" in the same 

 general way as is the custom in many other fields. Which is to say: 

 denote by dx the distance which the beam traverses through the gas; 

 by Q the number of electrons which enter the gas per unit time, hence 

 by Qe the amount of charge which in unit time would arrive at the 

 collector were the gas away; by Re the amount which in unit time does 

 actually arrive at the collector; by N the number of molecules per unit 

 volume: then the cross-section in question — call it c — will be defined 



2 That is to say, the discovery by experiment; it had been predicted by Bohr (for 

 the history of these matters, see for instance my "Introduction to Contemporary 

 Physics," Chapter yill). 



■• Of course this is an ideal which can never be perfectly realized. No matter how 

 many diaphragms may be set up in a row to narrow and sharpen the beam, there will 

 always be transverse motions of the electrons, relatively more important the smaller 

 the forward velocity is made. Moreover the mutual repulsion of the corpuscles will 

 tend to widen out the beam by driving its members apart. This is one of the reasons 

 why experiments in this field were first performed on fast electrons, then extended to 

 smaller and ever smaller velocities as time went on and technique was improved. 



