H. Sxey.—Smokeless and Sel/-feeding Furnace. 29 
then, in four minutes, it sank to 64°, and in six minutes to 63°. This indicates 
a loss of heat corresponding to only 14? F. by the interposition of the sheet of 
oxidized iron. This material was proved to heat very rapidly, and it also cools 
rapidly if a current of cooler air is passed over it. 
From these considerations it is evident that the heat in the flattened 
tubes*BB is continually radiating as long as there is a current of cooler air 
flowing through the other tube aa. Let us assume that the gases in both 
tubes are similar in quantity and properties, then it follows that all the air 
required by the furnace can be formed into a hot blast, having a temperature 
of at least 300? F. This is on the assumption that the fresh air is made no 
hotter than it would be if actually mixed with all the evolved gases ; but it 
will be seen that, by a proper arrangement and selection of material for the 
tubes, the air will have been raised to nearly 300? when it has traversed 
only half the length of the tube, or at A? ; consequently, as it goes onward to a 
still hotter part, near the tube В, it is continually acquiring fresh accessions of 
heat until it reaches that part of the thermo-convector where the temperature 
is, as we have seen, 600°; and similarly it may be shown that the evolved 
gases are cooled down to near 300? when they reach в?, and as they pass on 
they are rapidly cooled by imparting heat to the incoming current of cooler air, 
In the above apparatus, conduction and surface radiation only are alluded to ; 
but let us consider if transmission through a diathermanous medium could not 
be employed to advantage. We are indebted to Melloni for the discovery of 
the almost perfect transparency of rock-salt for all kinds of radiant heat. It, 
moreover, does not appear to suffer by a heat approaching to redness. Таш 
unable to ascertain if there is any difficulty in procuring it in large pieces, but 
as optical perfection would not be necessary, and plenty can doubtless be 
procured sufficiently transparent, small panes of this substance could be 
inserted at the top of each convolution of the air tube A. Then the hot gases 
from the furnace will instantaneously radiate heat into the air tube, and 
because the heat rays impinge on the surface of the sheet of oxidized iron at 
the bottom of the air tube, therefore they are immediately arrested and impart 
their vibrations to the contents of the air tube. Thus, as glass in our 
windows transmits all the rays of light, so do these plates of rock-salt fori, 
ot doors only, but windows for radiant heat. 
It is necessary for us now to consider the radiative property of the evolved 
gases in B, and the absorbent property of the air in the tube A (premising that 
both these properties are always possessed equally by the same body). First, 
with reference to the evolved gases, Tyndall has shown that they possess these 
qualities in an eminent degree, because they are compound gases; but this 
cannot be said of the air in the other tube ; indeed, if this air was quite dry 
and pure, all the radiant heat would merely pass through it without heating it at 
