220 Mr. B. Stewart on the Radiation and Absorption of Gases. 



tional to its radiative power,, it would seem very probable that 

 for many gases and vapours a sensible or even a large proportion 

 of the heat absorbed is that which corresponds to a few definite 

 and disconnected wave-lengths. 



To take a particular case — suppose that we wished to ascertain 

 the absorptive power of sodium-vapour for the heat correspond- 

 ing to that portion of the spectrum around the double line D. 

 If the spectrum be a continuous one, we are quite certain that 

 the sodium-vapour will select the two rays forming the well- 

 known double line, which will be absorbed with great avidity, 

 while the remainder of the heat will be allowed to pass. Also a 

 small quantity of sodium-vapour would be sufficient to do this, 

 and a large quantity would do no more. Perhaps a foot of 

 vapour would absorb no more than an inch, at least sensibly no 

 more. If, therefore, we wish to ascertain the absorptive power 

 of sodium-vapour, we must not compare one foot of vapour with 

 one inch, but one foot or one inch of vapour with a total absence 

 of all vapour. 



The same line of reasoning must, I imagine, be extended to 

 all gases and vapours ; nor do I think that the absorptive power 

 of any gas for radiant heat can be considered as determined until 

 the gas which is interposed as a screen between the source of 

 heat and the instrument which measures the radiation be com- 

 pared with a total absence of. all such gas : at any rate that 

 tenuity ought to be reached for which an increase of gas pro- 

 duces an increase of absorption. 



I wou'ld next remark that there is probably a connexion, of 

 the precise nature of which we are ignorant, between the kind of 

 heat radiated by gases and vapours, and the boiling-points* of the 

 liquids which give rise to these. 



To make an extreme supposition, let us suppose that sodium- 

 vapour does not radiate any heat of a lower refrangibility than 

 the double line D. If we suppose at the same time that heat of 

 this refrangibility is entirely absent in the radiation of low tem- 

 peratures, then we cannot imagine sodium-vapour to exist at 

 these temperatures. If these suppositions be considered extreme, 

 nevertheless they suggest to the mind the probability of a con- 

 nexion of the kind indicated above, and lead us to imagine that 

 gases which radiate heat of a high refrangibility belong to sub- 

 stances which have high boiling-points. 



Now, of all known substances, oxygen, hydrogen, and nitrogen 

 have the lowest boiling-points, indeed we have not yet been able 

 to liquefy these gases ; so that, according to the above hypo- 

 thesis, we should associate the description of heat radiated by 



* The word " boiling-point " is here used as a single word which suffici- 

 ently represents the meaning to be conveyed. 



