1913-] STEVENSON— FORMATION OF COAL BEDS. 101 



sarily on such analyses as can be reached ; but here, at the very out- 

 set of the inquiry, the worth of these analyses is a matter of doubt. 

 Formerly, samples were selected at random, fragments taken from a 

 heap were supposed to represent the average of the bed. For many 

 years, however, sampling in the United States has been on the com- 

 mercial basis and very frequently the effort has been to ascertain 

 the run of mine composition. Analyses of samples collected accord- 

 ing to the different methods are, evidently, not of equal worth for 

 comparison, though it must be conceded that in a very great propor- 

 tion of cases, results obtained by the old method are remarkably like 

 those obtained by the new. 



Aillied to this is the other query as to how much of the deposit is 

 to be considered in determining the impurity of the coal. There are 

 those who, in a discussion like this, would throw out of considera- 

 tion all partings, thick or thin, and would consider only the coal itself, 

 holding that partings, as interruptions in the process of formation, 

 have only indirect bearing on the subject. But others maintain that 

 no part of the bed should be neglected, as the deposit must be con- 

 sidered as a whole. There is some degree of propriety in each con- 

 tention. Study of the coal itself gives a nearer approach to the 

 nature of the vegetable material forming the coal, it may give ap- 

 proximately a conception of what may be termed the original inor- 

 ganic material ; while study of the whole deposit may give a clue also 

 to foreign matters introduced during formation. Yet one finds him- 

 self confronted at once by a question as to the significance of part- 

 ings ; in one locality they may be mere films of f usain and impalpable 

 clay separating benches of the bed, whereas in another, one or more 

 of these partings may have swollen to a mass of shale or sandstone 

 or both, many yards thick. Some coal beds, like the Waynesburg, 

 have clay partings, 6 to 12 inches thick. Occasionally one of these 

 persists after the underlying or overlying bench has disappeared. 

 The question arises Should the sample be taken where the partings 

 are thin or where they are thick? .Should the sample be taken where 

 the bed is practically single and another where the bed is divided, the 

 latter to include the intervening sandstone, shale and perhaps 

 limestone? 



