SOME CONTEMfOK.lRV AIHASCI.S IX rilYSUS III 295 



its c(iiiivali-tu \v;ivf-trains an- on i\\v vvhcilr of a Iowit fri'<|iii'iuy than 

 ilu- most iiiti'iisi- of thi- \va\e-irains wliicli lonstituti- it wlu-ri- it is 

 narrow and hinh- By i-xaininin^j and rcsol\inj; tlu- X-rass radiated 

 frotn .1 t.ir^;i't , at \arious inclinations to tlu- diriTlion of tlu- Iximhardinn 

 rli-itrons, this was vorilicd -Acrilk-d in part, not altont-'llur. Thi' X- 

 ra\s radiated nearly towards the .source of tlie electron-stream inciiiHe a 



/ 



otrrcRCNCf or pottntui Am.co to tubc in volts 



Fig. 4 — Curves ("isochromatics") each representing the intensity of X-radiation 



of a very narrow range of frequencies, plotteil versus the energy of the bomljartling 



electrons. (Duane & Hunt, Physical Review) 



lesser proportion of high-frequency wave-trains, the\' are softer as 

 the phrase is, than the X-ra\-s radiated nearly along the prolongation 

 of the electron-stream. In the spectrum of each of these beams of 

 X-rays, there is a wave length where the density of radiant energ>- 

 attains a ma.ximiim, and this wave length is longer in the former 

 beam than in the latter one. .So much is implied in tiie classical 

 theoi>'. 



But it is nowhere implied in the cl.issical theor\' that the spectrum 

 of an X-ray beam, produced when electrons of a constant energy 

 rain down upon a metal, should extend upwards only to a certain 



