596 Original Articles. [Oct., 
perature of 1,000° Fahrenheit, the sides of which are composed of 
every variety of substance from polished metal to lamp-black. Let 
us place in this room a suitable instrument for measuring temperature 
call it a thermometer—it is clear that in whatever part of the room 
we place this instrument it will always denote 1,000", since all the 
walls are of that temperature, Let us bring our thermometer near to 
the polished metal, it will still denote 1,000°. Now the rays which 
reach the thermometer from the polished metal are twofold. First, 
there are those given out by the metal itself since it is red hot; and 
secondly, there are those which it reflects from the lamp-black surface 
beside it. But this twofold supply of heat from the ‘polished metal, 
partly given out and partly reflected, will be too much for the ther- 
mometer, unless it happens that the metal, in virtue of the heat 
which it reflects, gives out so much less on its own account. This is 
found to be the case, and it is very easy to convince ourselves by 
experiment that a reflecting body when heated gives out little light. 
Let us take a piece of platinum, partly polished and partly tarnished, 
heat it to a white heat and then immediately examine it in the dark. 
We shall find that the polished portion is much less luminous than 
the tarnished. Or, in like manner, let us heat some lead or tin to a 
good red heat, and then take the vessel containing it into the dark; 
if we skim its surface with a red-hot iron spoon we shall find that the 
heated metal underneath is much less luminous than the dross. We 
may vary the experiment by heating a piece of porcelain of a black 
and white pattern, and examining it in the same manner, when we 
shall find that the black is much more luminous than the white, the 
reversal of the pattern producing a very curious effect. 
But let us return again to our chamber, and now suspend in a 
vertical position, near one of the walls, a sheet of transparent colour- 
less glass, leaving it, of course, for a sufficient length of time to 
enable it to acquire the temperature of the chamber. Let us place 
our thermometer in front of this plate. Now, the plate being trans- 
parent will permit to pass through its substance all the red heat from 
the wall behind it: if, in addition to this, it gave out a great deal on 
its own account, we should have, just as in the previous case, a 
twofold supply of heat falling upon the thermometer, so that its 
temperature would rise above that of the chamber. Since this cannot 
be the case, it follows that transparent glass will give out very little 
red light on its own account when heated to redness. We may verify 
this for ourselves by taking various pieces of glass, some colourless, 
and others more or less coloured; let us heat them to a good red 
heat, and examine them in the dark. It will then be found that a 
colourless piece gives out very little light, while a highly-coloured or 
opaque specimen gives out verymuch. The law may be stated thus :— 
a body which absorbs much heat will also give out much on its 
own account when heated. 
But one important fact yet remains behind. Let us revert to the 
plate of glass suspended in the red-hot chamber, and suppose this 
plate to be of such a colour as to stop all the red rays that reach it 
from the wall behind. We have already seen that, if it stops a good 
