78 



BELL SYSTEM TECHNICAL JOURNAL 



the scattered quantum in its turn is to be in the visible spectrum, its 

 wave-length must be less than some 8000 A., its energy more than 

 roughly 1.5 equivalent volts. The primary corpuscle of light must 

 therefore not cede to the molecule or atom more than (3 — 1.5) or 1.5 

 equivalent volts of energy — the material particle must therefore be 

 able to receive energy in quantities less than this, quantities preferably 

 which are small fractions of an equivalent volt ; its excited states should 

 differ from the normal state by energy-differences of this order; and 

 molecules satisfy this condition. ^^ 



En somme, then, the scattering of light with change of frequency was 

 never discovered in all the abundant work on the common monatomic 



Fig. 3— At top, spectrum of primary light, reduced by a filter almost to a single 

 strong line; below, that of light scattered by benzene, showing shifted lines due to 

 that single line (4358). (C. V. Raman; Indian Journal of Physics.) 



gases, because: first, the atoms were too few in unit volume to cause 

 much scattering; second, they could not be squeezed together without 

 destroying their power of entering distinct excited states; third, the 

 excited states had energy-values so high, that a quantum of the visible 

 or the near ultra-violet spectrum striking a normal atom would simply 

 not have had energy enough to transfer it into one of them. It was 

 discovered by Raman in the way in which he discovered it, because he 

 was working with molecular liquids where: first, the molecules were 

 numerous in unit volume and scattering was frequent; second, in spite 

 of being squeezed together the molecules retained their power of enter- 

 ic Atoms however do not always satisfy it; in particular, those of the noble gases 

 and of mercury, the substances most often used in optical experiments on monatornic 

 gases, possess no excited states differing from the normal state by less than four equiv- 

 alent volts — another reason why the Raman effect was not sooner discovered. The 

 more massive of the alkali metals have excited states superior to the normal by about 

 1.5 equivalent volts, while metals of the third column of the periodic table would be 

 very favorable. 



