362 Professor Forses’s Researches on Heat. 
sult in my former paper, art. 45. Under any circumstances the 
experiment is a troublesome one, but I have succeeded in arrang- 
ing it in perhaps as satisfactory a way as it admits of being done. 
The great difficulty arises from the minuteness of the quantity 
of heat reflected, and consequently the large quantity absorbed 
by the plates, which complicates and obscures the effect. This is 
more particularly true with dark heat, which, at the same time, 
furnishes the most important case to be examined. The effect of 
the absorbed heat is to produce a powerful secondary radiation. 
28. My first inquiry on resuming the subject was to ascer- 
tain the relative order of several different substances as to their - 
power of reflecting heat. This was not proposed to be done with 
a view to a general inquiry into that important subject, which 
I reserved to another occasion, but simply to ascertain what re- 
flecting surfaces might be best employed in polarizing by reflec- 
tion. Several series of experiments gave the following arrange- 
ment of substances according to their power of reflecting heat, at 
an incidence of 45°, beginning with the most perfect reflector. 
Polished speculum metal. 
Mica, split by the hand into thin plates. 
Mica, split by heat (see art. 20). 
Thick plate of mica. 
Rock-salt, with a thin coating of varnish. 
Polished rock-salt. 
Glass. 
( Alum. 
The three last substances (so different in their diathermancy) 
were nearly equal in their reflective power for dark heat (from 
brass about 7000°). The above order did not, however, appear 
to be changed for heat from incandescent platinum, except that 
elass seemed to stand above alum and even salt. Ina general way, 
we may consider the measure of metallic reflection to be from two 
to three times as great as that from mica split by heat, which is 
