PHOTONS AND ELECTRONS 35 



these photons nor be excited by them. Yet it is observed that when 

 exposed to the irradiation, the mixture of vapors emits not only the Hne 

 2536 A of mercury, but several lines of the spectrum of thallium ! and the 

 like effect occurs when the vapor mixed with the mercury, instead of 

 being thallium, is any one of a number of different elements (for instance, 

 cadmium, silver, lead, bismuth, indium, zinc). 



It appears therefore that there is a mechanism of direct transfer of 

 energy between atom and atom, not involving the emission of a photon 

 from one and its absorption by the other. This process occurs at colli- 

 sions, and those at which it occurs are called "collisions of the second 

 kind," a name very much open to criticism for its total lack of descriptive 

 quality! In the cases which I have just mentioned, some of the energy 

 absorbed from the incident hght is reradiated in the form of light of 

 different (usually lower) frequency, so that these may be considered as 

 additional cases of the production of light by light. Another case of this 

 type is displayed by sodium: here we have a normal state N and two 

 excited states Ai, A2 very close together, the transitions between N and 

 A I and those between N and A 2 corresponding to the famous yellow 

 D hues of sodium; when rarefied sodium vapor is irradiated by either 

 D line, photons of the same line only are emitted, but when the same 

 vapor is mixed either with hydrogen or with a much larger quantity of 

 sodium, the other Une appears, proving that collisions between two 

 sodium atoms or between one sodium atom and a hydrogen molecule are 

 capable of transferring one of the atoms from the state Ai to the state A 2, 

 or vice versa! 



Even in these cases, not all of the energy given up by the primarily 

 excited atom reappears as light. Sometimes none of it reappears as light, 

 and we have, not a special case of the production of light by light, but a 

 special case of the absorption of light with its total conversion into some 

 other form of energy. Thus when a mixture of mercury vapor and 

 hydrogen is irradiated with 2536 A, the excited mercury atoms are able to 

 dissociate the hydrogen molecules ; this is especially interesting, since light 

 of this wave-length is not able to dissociate hydrogen molecules without 

 the aid of such an intermediary. (There are, however, numerous cases of 

 molecules which can be dissociated by direct absorption of light 12). 

 Again, the excited atoms may ionize atoms of the other kind present in 

 the mixture. There are also many chemical changes which may be 

 brought about by excited atoms, though the light which excites them 

 cannot by itself bring on the changes which after excitation they are 

 capable of producing; but this idea is no novelty to anyone acquainted 

 with photochemistry. 



12 See, for instance, an article of mine in Chemical Reviews 5: 451-466. 1928; 

 the subject has, however, been much developed since then. 



