42 
EXPERIMENTAL WORK IN FUEL TESTING. 
In order to see if mechanical losses might account for the differences 
on the bituminous coals, determinations for the ash after the determi- 
nation of the volatile matter were made on one of the Pennsylvania 
samples. The average results for ash on the samples by the two 
methods are: Official method, 12.56; preliminary heating, 12.53. 
These results indicate no mechanical loss whatever, and in none of the 
samples except the lignites were visible solid carbon particles driven 
off in the form of sparks, and the differences must be ascribed to the 
different breaking down of the hydrocarbon compounds by the differ- 
ence in heat treatment. 
Comparisons of the results of volatile matter on a great number of 
samples differing from one another in moisture content indicate that 
the presence of loosely held moisture in the sample causes a higher 
volatile result. In order to obtain more definite data on this question, 
three samples low in moisture and representing widely different kinds 
of coal were selected for a series of determinations. The effect of 
loosely held moisture upon the determinations for volatile matter in 
each of these samples was determined by adding definite amounts of 
water to the sample after weighing out. The water was thoroughly 
mixed with the sample by means of a fine platinum wire and the vola- 
tile determination then made in the usual manner according to the 
official method. The first sample selected was a sample of Pennsyl- 
vania coal. The proximate analysis of the air-dried sample and the 
results for volatile matter determined in the presence of additional 
moisture are shown in the next table. Air-dried samples of Illinois 
coal and Arkansas lignite were treated in the same way, with the results 
also shown in the table. 
Determinations of volatile matter (percentages) si howing effect of loosely held moisture. 
Proximate analysis (official 
method). 
Volatile -matter determinations in the 
presence of stated amounts of added 
moisture. 
Fuel treated. 
Mois- 
ture. 
Vola- 
tile 
mat- 
ter. 
Fixed 
car- 
bon. 
Ash. 
Sul- 
phur. 
0.03 
gram. 
0.05 
gram. 
0.1 
gram. 
0.15 
gram. 
0.2 
gram. 
0.3 
gram. 
Pennsylvania coal 
Illinois coal 
Arkansas lignite 
1.05 
2.35 
10. 85 
33.00 
39.35 
38.50 
53.30 
44.65 
31.40 
12. 55 
13.65 
19.25 
1.75 
'"."83" 
"39.'60" 
33.60 
39.30 
40.35 
33.70 
40.00 
41.20 
33.80 
40. 05 
40.90 
34.10 
39.75 
44.90 
33. 90 
Without exception these results show that the presence of loosely 
held moisture in the sample increases the value obtained for the vola- 
tile combustible matter. The average increase for the Pennsylvania 
sample is about 0.7 per cent. On the Illinois sample the increase for 
volatile matter is 0.4 per cent. On the Arkansas lignite the increase 
is 3.3 per cent. 
To see what effect this loosely held moisture might have on the vola- 
tile determinations where the sample was first subjected to four 
