MOISTURE LOSSES DURING SAMPLING. 7 
analysis of this sample, and analysis of the portion of the sample 
reduced at once, are as follows: 
Per cent. 
Portion sampled at once 5. 07 
Portion sampled after standing five days 4. 58 
Loss by standing 49 
These coals dry down to about 1 per cent moisture, so that the 
loosely held and surface moisture amounted to about 4 per cent. 
From this experiment the moisture loss which occurs during the short 
time that the sample usually is allowed to stand in the laboratory 
before sampling would appear to be unimportant. 
INVESTIGATION OF MOISTURE LOSSES DURING SAMPLING. 
In preparing the samples for these tests the coarse sample (40 to 50 
pounds in weight), when received at the laboratory, was divided by 
quartering. One portion of 500 grams was at once ground without 
air drying in the ball mill. Another portion of about 5 to 10 pounds 
was air dried in the usual way before the final pulverization. The 
results (reduced to sample as received) in moisture and ash for the 
two portions are as follows: 
Moisture and ash determinations on air-dried samples and fresh samples of coal. 
Designation of sample. 
Regular sample 
(air dried). 
Fresh Sample (not 
air dried). 
Moisture. 
Ash. 
Moisture. 
Ash. 
i 13.54 
I 13. 72 
33.85 
5.19 
10.57 
15.68 
13.47 
13.80 
11.91 
10.74 
10. 32 
7.30 
14.01 
11.65 
15.59 
11.53 
11.74 
13.01 
a 9 36 
11.92 
33 00 
3 58 
9.95 
11.78 
12 43 
12. 58 
11.58 
11.37 
Texas No. 4 
10.40 
7.65 
Arkan sas No. 8 
14.16 
Indiana No. 12 
11.98 
Illinois No. 23 B 
16.11 
Illinois No. 23 A 
11.63 
Missouri No. 6 
11.15 
Illinois No. 22 A 
12. 57 
aThis sample was ground very fine in ball mill, which probably accounts for the loss being so large. 
The results obtained for moisture from the original samples that 
were ground down without preliminary air drying are without excep- 
tion decidedly lower than the results obtained from the regularly pre- 
pared air-dried samples when calculated to the sample as received, 
the greatest difference being over 4 per cent and the average difference 
over 1£ per cent. Furthermore, the fine sample, after being ground 
without previous air drying, as shown by experiments (see pp. 13-17), 
gives up moisture so readily as to indicate that alarge additional loss 
during the handling and weighing of the sample in the laboratory is 
almost certain. The consequent errors due to moisture losses are 
liable to be so large as to affect very seriously the accuracy of the 
results obtained. 
