ON THE RADIATION OF HEAT. 
167 
In considering the air as the medium of the transfer of heat, he 
supposes that the hot or cold surface of the canister heats or 
cools, and consequently causes an expansion or contraction of, 
the adjoining atmosphere;—that the first layer of air thus ex¬ 
panded or contracted presses on or is compressed by the portion 
before it;—that this process is renewed in a rapid succession, 
and that an undulatory motion of the heated or chilled air is thus 
propagated to the mirror and thence reflected to the focus; and 
each pulsation being accompanied by a discharge of heat from 
the portion of air at the higher temperature, that the heating or 
cooling effect is conveyed to the thermometer simultaneously 
with the progress of the undulation. I may here remark that it 
appears not improbable that such expansions and contractions 
should take place ; and when we take into account the change in 
the capacity of air for heat produced by expansion or compres¬ 
sion, we have at least a plausible reason for the transfer of heat 
which he supposes to accompany the aerial undulation. 
I do not wish to be understood as adopting Leslie’s views, 
but I conceive them deserving of further investigation, and the 
only conclusion I wish at present to draw from the previous ex¬ 
periments is, that they are only explicable on an undulatory 
theory, and consequently, if air be not the medium of the trans¬ 
fer, that they furnish an additional and perhaps conclusive argu¬ 
ment in favour of the undulatory theory of light. 
I believe no cause has been attempted to be assigned for the 
difference in the radiating powers of surfaces except by Leslie, 
who supposes it to arise from the different distances of the atmo¬ 
spheric boundary; I may, perhaps, therefore, be allowed to refer 
to a known property of bodies which probably ought to lead us 
to anticipate such results, viz. their different capacities for heat . 
If two surfaces* are of the same elevated temperature and placed 
in the same medium, they may be considered as having the same 
tendency to attain the temperature of that medium, and may con¬ 
sequently be expected to give off the same portion of their ex¬ 
cess of temperature , and consequently quantities of heat propor¬ 
tional to their capacities. If we look to classes of bodies I be¬ 
lieve they will be found to be in accordance with the cause I 
have just ventured to assign. 
Having recently received one of Melloni’s thermomultipliers, 
I have made a few experiments with reference to the transmis¬ 
sion of heat (from boiling water) through crystals (of rock salt, 
alum, and rock crystal), which were sent with the instrument. 
I found the diathermancy of rock salt very marked, though not 
* Taking the term in the physical sense of having some definite thickness, 
which may be different, in different substances. 
