O17 
tapped for the 3” pipe (12), through which the gas from the coal could 
pass out. The other was drilled and tapped for a thermometer well (8) 
and 4” pipe (11). The body of coal heated each time was therefore 24” 
in diameter, and it was necessary to have all the particles of coal in 
this mass at the same temperature. As coal is a very poor conductor of 
heat, it was decided to place iron disks (10) 4” apart throughout the entire 
length of the cartridge. These disks were large enough to touch the iron 
cartridge all around, thus taking on its temperature, and were drilled 
full of small holes to allow the gas to pass through them. As they were 
+” apart, the heat had to be conducted through only 4” of coal. This fact, 
together with the very slow rate of heating employed, led us to expect the 
temperature throughout to be the same, within very close limits. The 
temperature was read at the very center of the coal body, by means of a 
thermo couple, the end of which extended down to and touched the end 
of the thermometer well (8). 
3. Means of measuring the temperature. The thermo couples used 
were made of iron and nichrome wires welded together in an electric 
are. The couples were used with a Brown millivolt meter, with a resist- 
ance of 85 ohms. It was carefully calibrated and checked at the time the 
experiments were completed. The couples were correct to within 10° C. 
throughout the range of temperatures here reported. The couples were 
left in place throughout each entire test, and the temperature readings 
made whenever desired. 
4. To keep the gases swept out as formed. To sweep the gases out as 
they were formed gas free from tar was forced into the cartridge under 
pressure, through the 4” pipe (11), and allowed to pass out through (12) 
in a constant stream. The pressure of this gas inside the cartridge was 
measured by a mercury manometer, and was kept at about 24” of mer- 
cury. This gas could not be allowed to contain any O,, as it might then 
burn the coal or tar vapors at the higher temperatures, so air with the 
O, burned out was used. The arrangement of the apparatus as used is 
shown in Fig. 2. The air was burned in a small furnace made of a piece 
of 6” pipe about 2’ 6” long. This pipe had grates at the bottom and 
a coupling and plug at the top. The coupling was drilled and tapped #4” 
pipe size on one side, and connected to a large coil of +” pipe which rested 
in a tank of cold water. The small furnace thus constructed was filled 
with an anthracite fire and the plug at the top put in. The air pump 
then pumped air through the furnace and cooling coil and compressed 
