288 



BELL SYSTEM TECHNICAL JOURNAL 



corner as the three here reproduced marks liie entree en schie of a 

 new ray or set of rays." 



But a determination of an excitation-potential is not a measurement 



ELECTRON EdCRCY VOLIS 



Fig. 9 — Breaks in a photoelectric-current curve indicating excitation-potentials- 

 (Philosophical Magazine.) 



of the wave-lengths of the excited rays; and while it is supposed that 

 excitation-potentials between 1000 volts and 100 volts are associated 

 with rays of wave-lengths between 12.-\ and 123.A, thi.s is merely a 

 supposition." We require a theoretical relation between excitalion- 



" It is clear from this account that the photosensitive disc might be replaced by a 

 photographic plate, on which the opacity due to the rays produced by consecutive 

 vahiesof I' could be measured; or by an ionization-chamher, in which the ioni/ation- 

 currents could be measured. It is ecpially clear that neither method would be 

 so suitable for detecting slight discontinuities in rate of increase of radiant energy 

 with increase of V. However, both methods are used at higher frequencies, where 

 by <lispersion of the waves a discontinuity in the intensity at a single wave-length is 

 made more conspicuous. 



'* This is the best place to remark that electrons of voltage I' bombarding a solid, 

 in addition to exciting (if V is high enough) rays characteristic of the bombarded 

 atoms, excite also a continuous spectrutn of rays of all frequencies up to a maximum 



