PHOTONS AND ELECTRONS 27 



increase of E with increasing U becomes sharply greater when U reaches 

 hvK. If we hold U constant, divide the X-ray spectrum into consecutive 

 small equal ranges of wave-length, measure E for each of these, and plot 

 E as function of wave-length, we obtain the distribution-in-energy curve 

 of the continuum : this has a maximum at a wave-length about half again 

 as great as that of the upper frequency limit, and after passing the 

 maximum, it drops sharply to the axis of abscissae at X = ch/U. The 

 general shape of the curve varies little from one element to another, and 

 its upper end is the same for all elements, as we have just seen; but its 

 height and the area under it — the total energy radiated in the form of 

 X-rays — increase rapidly in passing in the direction of increasing atomic 

 weight along the table of the elements. At best, however, the total 

 energy radiated in X-rays is less than 1 per cent of what the impinging 

 electrons possess; the rest appears as optical radiation or as heat. 



X-rays may be generated by projecting another beam of X-rays 

 against a piece of matter; when so produced, they are known as "fluores- 

 cent X-rays." They comprise only the spectrum lines of the elements 

 present in the irradiated mass, without a continuimi (except for such a 

 feeble one as may be produced by electrons released within the matter 

 and stopped before they emerge from it) . This is desirable in many ways ; 

 but unfortunately fluorescent X-rays are always far weaker than the 

 beams which may be generated by strong electron bombardment. The 

 law connecting the lines which are produced with the frequency of the 

 irradiating rays may be deduced by the reader from what has already 

 been said. 



An effect alternative to the production of X-ray spectrum lines is of 

 much theoretical interest. Imagine an atom from which a K electron has 

 been expelled, leaving behind a vacancy which is presently filled by an 

 electron dropping into it from the L„ shell. As I have already said, it is 

 common for this second process to be accompanied by the emission of 

 a photon, bearing off with it the energy {hvK — hvLn). On the other 

 hand, it is possible for extra electrons to be emitted in place of the 

 photon, precisely as if this corpuscle of light had been reabsorbed in the 

 very atom whence it was about to emerge, and had employed its energy 

 in ejecting, say, an M^ electron; we find electrons coming out of the 

 substance possessing the kinetic energy (hvK — hvLu — hvui), which 

 when added to the energy hvui required to extract the Mi electron is none 

 other than the energy of the vanished photon. It is possible for two or 

 three electrons to be ejected simultaneously in this fashion. The effect 

 is known as the Auger effect or as "internal conversion." Whether one 

 should imagine the photon as enjoying a temporary though very brief 

 existence between the inward falling of the electron which starts the 

 process, and the outward springing of the observed emerging electrons, or 



