EFFECTS OF DRYING REAGENTS. 17 
As might be expected, the rapidity of gain or loss is greatest in the 
1-gram sample and least in the sample kept in an open bottle. A com- 
parison of these losses after seventy -two days on the undried samples 
shows, however, that even in that time the loss on the 1-gram sample 
is decidedly in excess of that on the 10-gram sample, and about twice 
that on the sample in the open bottle. This result would lead to the 
inference that the change in weight is not due entirely to moisture 
losses, but is influenced more or less by oxidation changes. That such 
is the case is shown by the results of tests of the gain or loss in weight, 
from time to time, of fine samples kept in tightly stoppered bottles, 
moisture determinations being made at the same time to find out 
whether the amount of moisture present in the sample varies with 
the change in weight. These tests are described under "Alteration 
of weight of samples of coal when kept in a finely powdered state," 
pages 19-22. 
COMPARISON OF EFFECTS OF DIFFERENT DRYING REAGENTS USED IN 
THE DESICCATORS EMPLOYED IN MOISTURE DETERMINATIONS. 
In the determinations of moisture made during the Louisiana Pur- 
chase Exposition and during the earlier part of 1905 duplicate results 
often were not as close as was desirable. This lack of agreement was 
found to be due, in part at least, to the use of calcium chloride as a 
drying reagent in the desiccators in which the fine samples of coal 
were allowed to cool after drying at 105° C, as may be seen from 
inspection of the following table giving the values obtained over con- 
centrated sulphuric acid, fused calcium chloride, and granular calcium 
chloride. 
Certain values from samples allowed to stand overnight, which are 
marked with a star (*), were on different gram portions of the sample. 
The values given for the other determinations, weighed as soon as 
cooled and weighed after standing in the desiccator overnight, are on 
the same weighed-out portions of the sample, which, as soon as weighed 
the first time, were put back in the desiccator and allowed to stand 
until the next day and then again weighed. The values over sulphuric 
acid are, as a rule, about one-tenth of 1 per cent lower on the sample 
after standing overnight, but this result can probably be accounted 
for by the small amount of moisture which might be taken up by the 
sample during the time required for making the first weighing. 
6128— Bull. 323—07 3 
