September 11, 1914.

THE COLLIERY GUARDIAN

565

difficulty of dealing with spontaneous heating lies. If
the coal became readily saturated with oxygen, by a
process of weathering before the coal was stored or before
the goaves were closed, the possibility of spontaneous
heating could be removed. Of course, the coarser the
coal is, the slower is the absorption of oxygen, and the
longer the weathering necessary to render it safe from
further oxidation. The weathering of lump coal which
may be afterwards crushed in stacking or in goaves
would take an indefinitely long time, and would be
quite impossible in practice.

Part III.—The Thermal Value of the Absorption.

The absorption of oxygen by coal is known to produce
heat, and there is a distinct probability that the heat so
produced is a sufficient cause of spontaneous combus-
tion. Lamplough and Hill* were the first to measure
this quantity approximately, and found that the absorp-
tion of 1 cu. cm. of oxygen by coal is accompanied by
the evolution of 3-3 calories on an average. They
investigated several samples of coal, some bearing
pyrite and some not, and found within the limits of
their experimental error that the heat evolution is the
same whether coal or pyrite is undergoing oxidation.

In connection with some experiments made at the
Doncaster Coal Owners’ Laboratory, the thermal value
as determined by Lamplough and Hill was used in cal-
culations, but gave results not in agreement with those
of the experiments then made. The thermal value has,
therefore, been redetermined, and a number consider-

per cent, of oxygen. The quantity of oxygen not
absorbed by the coal was therefore :—

79 x 0-51 = 0-4 cu. cm.

Subtracting this from the volume of oxygen admitted
originally to the apparatus, the volume of oxygen
absorbed by the coal is given. The removal of the last
traces of oxygen from the reaction flask at a very low
pressure, in the way described, makes it certain that the
oxygen absorption measured is a true absorption and
not adsorption.

In order to measure accurately the rise in tempera-
ture, precautions must be adopted to prevent loss of
heat during the reaction. This may be done to a large
extent by employing a vacuum flask to contain the coal.
Even then there is a fairly considerable loss of heat
when the contents of the flask are much hotter than the
surroundings. The vacuum flask was therefore
immersed in a bath, and fitted with a simple device
which maintained the bath at a temperature about
0-2 deg. Cent, lower than that inside the flask. The
loss of heat is in this way reduced to a minimum. The
regulator is shown in fig. 2. A and B are two similar
glass bulbs connected through the tap C and the
U-tube D. The U-tube contains a mercury thread, so
that when the tap C is closed, any variation of pressure
in A or B moves the thread in the direction of the bulb
at the lower pressure (or hence the lower temperature).
When the pressures in A and B vary’equally, the thread,
of course, remains stationary. Platinum wires are

The temperatures were taken on thermometers divided
into deg. Cent. The specific heat of the coal and
the thermal equivalent of the flask, thermometer, etc.,
were determined by the method of mixture.

Preliminary experiments carried out by the method
described above gave a value of about 1-8 calories for
the absorption of 1 cu. cm. of oxygen. Three accurate
determinations were then made, with the following
results :—	_	,

(a)	With Hard Coal.

(1)	Initial temperature, 55 degs. Cent.; rise in
temperature, 5-29 degs. Cent.; 2-1 calories per cu. cm.

(2)	Rise in temperature, 3-50 degs. Cent.; 2-1 calories
per cu. cm.

(b)	With Soft Coal.

(3)	Initial temperature, 40 degs. Cent.; rise in
temperature, 3-11 degs. Cent.; 2-1 calories per cu. cm.

The coal used in the above-mentioned experiments
had been heated in vacuo on a water bath for three
hours, in order to free it from moisture and gas. It
contained no pyrite.

Some time has been expended in seeking the reason
for the considerable differences in the values now found
and those obtained by Lamplough and Hill. The
elaboration of the apparatus used by them, and the con-
sequent corrections to allow for the effect of each part,
make the cause of the difference somewhat difficult to
discover. Table II. is a re-arrangement of their
“ Summary of Results.”

Fig. 1 -Apparatus used for Determining the Absorption
of Oxygen by Coal.

Fig 2 -Temperature Regulator for Calorimeter Jacket

2.2001

2,000-1

1.900

1,500-

1.AOO-

900

800

ZOO-

GOO

.500

400

300-

200

too

193

168

12Q

96

TIME IN HOURS.

m
iu
a:

z
in
o

IX

O 1.700-

UJ
2 1 600

S

Fig. 3.—Curves showing the RelXtion between the
Rate of Absorption of Oxygen and Time for

Hard Coal at Temperatures of 40°. 60°. 80°, 1^0°

120°. 140°. and 160° Cent, respectively

q

U1 1,300-
00
tr

z i,ioo-
UJ
o

ably lower than that obtained by the previous
investigators has been found.

In measuring the evolution of heat during the reaction
two quantities have to be known accurately : one is the
quantity of oxygen used, and the other the rise in
temperature. Every effort has been made in the present
investigation to keep the apparatus as simple and as
free from complications as possible, so that errors may
be easily detected, and a large number of “ corrections
for various effects avoided.

In order to put the measurement of the quantity of
oxygen absorbed beyond doubt, the simple and obvious
method has been adopted of admitting to the coal
in vacuo a volume of oxygen determined accurately
outside, and independently of, the reaction flask. The
absorption was allowed to continue until practically all
the oxygen had disappeared, as was shown by a con-
stant volume gauge on the apparatus. The small quan-
tity of oxygen remaining can be removed from the
reaction flask by admitting a known volume of nitrogen,
rapidly pumping this out again, and analysing the gas
so obtained. Thus, in one experiment, after the gauge
had shown that little oxygen was left, 80 cu. cm. of
nitrogen were admitted, and rapidly removed by a
Toplor pump. The volume of gas so obtained was
79 cu. cm., and analysis showed that it contained 0-51

* “ The Slow Combustion of Coal Dust and its Thermal
Value,” by F. E. E. Lamplough and A. Muriel Hill, Trans.
Inst. M.E., 1913, vol. xlv., p. 629. Colliery Guardian,
June 6, 1913, p. 1212.

fused through the glass at a and 5, and the length of the
mercury thread is so adjusted that when A and B are
at the same pressure the wire b just touches the mercury
surface. The method of using the apparatus is obvious :
ofie bulb A is placed in the reaction flask, the other B
hangs in the bath, the U-tube being well over the side
of the bath. The required difference of temperature (in
these experiments, 0-2 deg. Cent.) having been
obtained, between the contents of the reaction flask and
the bath, the tap C is closed, and an accumulator work-
ing a relay is connected to the wires a and b. An
electric lamp immersed in the bath is connected to the
relay. When the temperature of A rises, the mercury
thread rises and touches b, completes the circuit, and
lights the lamp in the bath. As soon as the tempera-
ture of the bath reaches the required point, the mercury
is forced down from b by the pressure in B, and the
lamp goes out. Hence, whether the temperature in the
reaction flask rises or falls, the temperature of the bath
varies by a similar degree, ensuring that the reaction
flask neither loses heat to, nor gains heat from, the
bath. The heat lost from such an apparatus is
negligible. For example, when the contents of the flask
were at 40 degs. Cent, in an atmosphere of nitrogen, the
cooling was only 0-5 deg. Cent, in 30 hours, when the
bath was kept 1 deg. Cent, lower than the contents of
the flask. As the experiments only last from two to
three hours, and the temperature difference was 0-2 deg.
Cent., it is obvious that no correction is needed for the
heat lost.

Table II.—Re-arrangement of the “ Summary of

Results ” in Lamplough and Hill’s Paper.*			
	Rate of absorp- tion.	Cubic ' centi- metres absorbed.	Calories per cubic centi- metre.
Slack			 		... 140	31	...	3’7
Bullhurst 		... 112	60	...	3’8
Anthracite			... 590	132	...	3’6
Hard coal No. 1			... 18’6	195	...	3’6
Top softs, containing 43	per		
cent, of pyrites	... 54’0	321	...	3’6
Top coal 	  Top softs :—	... '46’0	459	... 2’8
First part 		... 186’0	534	...	3’2
Second part 		—	... 1,041	...	3’0
Hard coal No. 2 		..’ 90’0	... 1,110	...	2’9
* “ Iron pyrites from coal”		is omitted.	

By arranging the results in this way, two points
become at once obvious, namely :—

(1)	The more rapidly, in general, that the coal
absorbs oxygen, the less is the heat evolved.

(2)	Coal which absorbs much oxygen, with one
exception, gives out much less heat in proportion than
coal which absorbs little oxygen. Further, the heat
evolved decreases all through the series as the quantity
of oxygen absorbed rises.

Both these conditions are, of course, exactly con-
trary to general experience. Of two similar reactions,