302 BELL SYSTEM TECHNICAL JOLRXAL 



follow their own later history, or the adventures of a Ijeam of radia- 

 tion after it sinks into a metal. We have inferred that the electrons 

 which collide with a piece of tungsten and disappear into it transfer 

 their energ>- to X-rays, but the inference lacked the final support 

 which would have been afforded by a demonstration of these very 

 electrons, still personally present after the collision but deprived of 

 their energ>'. Now when electrons are fired against mercury atoms, 

 this demonstration is possible, and the results are very gratifying. 

 I have already several times had occasion to remark, in this series of 

 articles, that when an electron strikes a free atom of mercury, the 

 result of the encounter is very different, according as its energy of 

 motion was initially less than some 4.9 equivalent volts, or greater. 

 In the former case, it rebounds as from an elastic wall, having lost 

 only a very minute fraction of its energ>', and this fraction spent in 

 communicating motion to the atom; but in the latter case, it may and 

 often does lose 4.9 equivalent volts of its energ>' eti bloc, in a single 

 piece as it were, retaining only the e.xcess of its original energy over 

 and above this amount. Thus if electrons of an energy of 4.8 equiva- 

 lent volts are shot into a thin stratum of mercury vapor, nothing 

 but electrons of that energy arrives at the far side; but if electrons 

 of an only slightly greater energj', say 5.0 equivalent volts, are fired 

 into the stratum, those which arrive at the far side will be a mixture 

 of electrons of that energy, and very slow ones. The very slow 

 ones can be detected by appropriate means, and the particular \alue 

 of the energy' of the bombarding electrons, at which some of llicni 

 are for the first time transformed into these very slow ones, can be 

 determined. Once more we meet that question as to whether the 

 transformation does make its first appearance suddenly, but in this 

 case the indications that it does are rather precise and easy to read. 

 Furthermore it is possible to measure the energy of the slow electrons, 

 and one finds that it is equal to the initial energy of the electrons, 

 minus the amount 4.9 equivalent volts. (These measurements are 

 not so e.xact as is desirable, and it is to be Impi'd ilial sdmobody will 

 take up the task of perfecting them.) 



We, therefore, see both aspects of the transaiiicui wliich occurs 

 when an electron whereof the energy is 4.9 equivalent \olis, or greater, 

 strikes a mercury atom. It loses 4.9 equivalent volts of energy, and 

 we measure the loss; the atom sends forth radiation of a certain 

 frequency, and no other; the atom does not send forth even this 

 frequency of radiation, if none of the electrons fired against it has at 

 least so much energy. We have already compared the energy trans- 

 ferred with the frequency radiated, and as in the case of X-rays 



