28 EXPERIMENTAL WORK IN FUEL TESTING. 
An examination of the figures in the foregoing table shows the very 
rapid increase in impurity of the coal with increase in specific gravity. 
The finer the coal is crushed the more complete would be the mechan- 
ical separation of the heavy impurities from the coal, but the behavior 
in this respect evidently varies greatly with different coals. In the 
Brazil coal the sulphur is nearly the same in all the coal under i-inch 
size and below 1.65 specific gravity, while the ash increases both with 
size and with the specific gravity. In the Maryland coal both sulphur 
and ash increase with specific gravity, and apparently without very 
much reference to size of product. Practically the same is true of the 
Pennsylvania No. 6 coal and of the West Virginia coal. The record 
of the Wyoming coal is not sufficiently complete for comparison. In 
all but the Brazil coal there is a marked tendency to increase of ash 
and sulphur in the fine dust, suggesting that the impurities are liber- 
ated in crushing and are more brittle than the mass of the coal. 
Noticeably different 'compositions in ash and sulphur of the portions 
of the various coals below 1.35 specific gravity, may indicate either 
actual variations in the specific gravity of the coal proper or variations 
in the nature of the combination of the sulphur and mineral matter. 
Further experiments along these lines seem very desirable. 
LABORATORY METHODS OF DETERMINING ADAPTABILITY OF COALS TO 
IMPROVEMENT BY WASHING. 
The laboratory method of testing consists in floating the sample 
upon solutions of different density and thereby separating the coal 
into portions of different specific gravity, the amounts of these por- 
tions being determined and each portion analyzed separately for ash 
and sulphur, as noted under the last heading. The samples used were 
crushed to one-half inch and finer before testing, and for convenience 
in handling were divided by sifting into two portions, one one-half 
inch to one-fortieth inch, the other one-fortieth inch and finer. On 
some of the earlier samples the division was made at one-twentieth 
inch. The solutions used for washing the samples were a calcium- 
chloride solution of 1.35 specific gravity and a zinc-chloride solution 
of 1.65 specific gravity. Solutions of 1.45 and 1.90 specific gravity 
were also used occasionally. The clean coal, low in ash, floats upon a 
solution of 1.35 specific gravity. Moderately high-ash coal sinks in 
a solution of this gravity, but floats upon a solution of 1.45 specific 
gravity. Coal very high in ash is heavier than 1.45, but floats upon a 
solution of about 1.65 specific gravity, while the slate and p} r rites 
sink in solution of this gravity. The results obtained by washing the 
samples upon these different solutions and the analysis of the resultant 
products, together with the analysis of the original sample and the 
analysis of samples obtained from actual washing tests made at the 
washery connected with the fuel-testing plant upon 5- to 6-ton lots of 
