ON THE THEORY OF EXCHANGES. 10/ 



Connexion between Radiation, Absorption, and Conduction," has given a very 

 lucid statement of a hypothesis of tliis kind, accompanied with a remarkable 

 experimental verification. 



On the supposition that an ether envelopes the molecules of matter (just 

 as the air surrounds the string of a musical instrument), the author points out 

 tiiat the reciprocity of absorption and radiation is a necessary mechanical 

 consequence of this theory, on the principle of the equality of action and 

 reaction. He then goes on to say, " the elementary gases which have been 

 examined all exhibit extremely feeble powers, both of absorption and radia- 

 tion, in comparison with the compound ones. In the former case we have 

 oscillating atoms, in the latter oscillating systems of atoms. Uniting tlie 

 atomic theory with the conception of an ether, it follows that the compound 

 molecule, which furnishes points d'appui to the ether, must be capable of 

 accepting and generating motion in a far greater degree than the single atom, 

 which we may figure to our minds as an oscillating sphere. Thus oxygen and 

 hydrogen, which taken separately or mixed mechanically produce a scarcely 

 sensible eff"ect, when united chemically to form oscillating systems, as in 

 aqueous vapour, produce a powerful effect. Thus also nitrogen and hydrogen, 

 which when separate or mixed produce but little action, when combined 

 to form ammonia produce a great action. So also nitrogen and oxygen, 

 which, as air, are feeble absorbents and radiators, when united to form 

 oscillating systems, as in nitrous oxide, are very powerful in both capacities." 



This great absorbing power which belongs to a compound molecule is a 

 very interesting result, and seems to be well explained by this hypothesis ; but 

 M'hether all compound gases without exception are more absorptive than their 

 components, in the absence of experimental evidence may, I think, admit of 

 being questioned. 



It has been shown in this Report that internal radiation follows immediately 

 from the theory of exchanges, and is independent of the distance from the 

 surface. In an uncrystallized medium, this radiation will, by the principle 

 of sufficient reason, be equal in all directions ; but here a question arises 

 which shapes itself thus : — Let us suppose a polished surface of indefinite 

 extent, bounding an uncrystallized medium of indefinite thickness; and placed 

 opposite to this surface and parallel to it let us imagine an indefinitely ex- 

 tended surface of lampblack ; and finally, let the whole arrangement be ke\it 

 at a constant temperature. Now we know the quantity of heat which radiates 

 from the lampblack in directions making different angles with the surface ; 

 and since the proportion of this heat which after striking the polished surface 

 penetrates it in a certain direction must be equal to the quantity of heat 

 which leaves this surface from the interior in the same direction, it can be 

 readily conceived how, by means of optical laws, we may be enabled to tell 

 the internal radiation, in different directions, of the solid to which this surface 

 belongs. It is remarkable that the internal radiation deduced by this method 

 for an uncrystallized body is equal in all directions — a result which we have 

 seen may also be arrived at by the principle of sufficient reason. 



In order to define internal radiation, let us conceive a square unit of sur- 

 face to be placed in the midst of a solid of indefinite thickness on all sides, 

 and consider the amount of radiant heat which passes across this square unit of 

 surface in unit of time, in directions very nearly perpendicular to the surface, 

 and comprehending an exceedingly small solid angle h<p. Call this heat Rt^, 

 then R may be viewed as the intensity of the radiation in this direction. 



Now if R denote the radiation of lampblack, and /.i the index of refraction 

 of an uncrystallized medium, it may be shown, that the internal radiation as 

 thus defined is equal to R/:^^ 



