Stonry—Of the Kinetic Theory of Gas. vl 
encounter between two molecules of the same kind. A suggestive 
experiment was made in the laboratory of the Royal Dublin 
Society, when Professor Emerson Reynolds, F.R.S., and the 
present author were engaged in examining the spectra of coloured 
vapours (‘ Philosophical Magazine,” July, 1871, p. 41). We had 
examined the splendid absorption spectrum of chlorochromic anhy- 
dride mixed with air. Here the encounters that the molecules of 
the vapour met with were most of them encounters with molecules 
of air, the minority only being encounters between molecules of 
the vapour. If the vessel containing the vapour be freed from 
air, the encounters between molecules of the vapour will be present 
in the same number as before, while all encounters between air and 
vapour will be absent. Hence we argued (for at that time we 
supposed the motions in the molecules to have been evoked by the 
undulation in the ether), that the molecules being less knocked 
about should produce a spectrum of lines that would be less diffuse. 
On trying the experiment, however, we found that the spectrum 
was sensibly the same as before. From this observation it would 
appear that the motions to which the spectrum is due are excited 
by the encounters between molecules of the vapour, and are unin- 
fluenced by encounters between air molecules and vapour molecules. 
Hence they are due to something different from the mere kinetic 
energy of the collisions ; and this something may be, and probably 
is, that during the struggle between molecules of the same kind, 
a struggle which is a protracted struggle from the molecular 
standpoint, there occurs either always, or now and then, an inter- 
change of some of the chemical atoms constituting the molecules. 
This, which is equivalent to two chemical decompositions, followed 
by two equivalent chemical combinations, must set the electrons 
concerned into a state of more or less activity. 
1 The amplitude of the BJ motions which the undulation in the ether, if acting 
alone, can develop is apparently too small ; and this is probably because, at this small 
amplitude, the transference of energy from Bé motions to Ba and A motions is incon- 
spicuous. Where this is the case the etherial undulation acting alone upon the gas 
cannot suffer any sensible amount of absorption. But the conditions are altogether 
different if some other agency produces an amplitude which can freely part with its 
energy by conduction, and which at the same time can be acted on by the ether in the 
direction which tends to keep the amplitude up. Under these circumstances there will 
be active withdrawal of energy from the ether, and an absorption spectrum will result. 
This seems to be what happens in coloured gases. 
