148 GEOLOGY OF OHIO. 
worst part of the seams would often be more valuable than a knowledge 
of the best. The average is what is demanded for practical guidance. 
Two plans have been followed in selecting samples for the analyses 
to be here reported. In one of them, the aim has been to secure a 
representation of the seam just as it is mined, by taking a cubic inch or 
less of coal from every inch of height of the seam, the partings and 
other rejected portions being, of course, omitted. 
It can be urged against this method that the seam is sampled at 
but one point, and that but one room is in reality tested. This is true, 
and analysis of the coal from other rooms of the mine will often vary 
considerably from the results obtained, but the aim has been to take 
the samples from rooms where the coal is approved, and as a conse- 
quence, the figures published will be somewhat more favorable than the 
total output of the mine would deserve. 
In one instance, an unexpected result was reached. Three sets of 
samples were taken, one of the whole seam, sampled as above described, 
one of the general output by the method next to be given, and one of a 
block selected by the manager as the best coal of the mine. The first 
set of samples showed the best coal, and the selected block was the 
poorest of the three. In selecting the samples for the first result, all 
the slaty partings that were said to be rejected by the miners, were 
thrown out, but it is probable that in the general output a good many 
remain. The inferiority of the selected sample only shows how little 
reliance can be placed on even a well-trained eye in estimating the 
nicer points of quality in a sample of coal. 
The second plan was to take samples from the bank cars as the 
coal came out, or from the loaded railway cars, the specimens being 
taken from a large number of blocks, never from less than 50 blocks, 
and often from as many as 100. This method was counted on the whole 
most satisfactory, and was used in a great majority of instances. 
In any case, to fairly determine the average composition of a 
particular coal seam, or of a particular mine is a more laborious under- 
taking than we have been wont to consider it. A single analysis of 
any sort does not go far towards determining it, but least of all is it 
trustworthy when the sample analyzed is the best that the mine pro- 
duces. | 
A necessary result of these methods of selecting the samples has 
been to show a considerably larger proportion of ash and of sulphur in 
