454 STEVENSON— THE FORMATION OF COAL BEDS. [Nov. i, 



regarded as derived from the cannel bench of an underlying bed. 

 That cannel pebbles should retain their shape and size better than 

 those of glance coal is normal, because cannel is tough and glance 

 is brittle ; but there is no reason to suppose that sapropelic muds 

 became hard and resistant with great speed while the underlying 

 felted peaty materials became compacted slowly. It would seem 

 more probable that cannel pebbles are more numerous because 

 cannel survived the shocks of transport and the glance coal was 

 reduced to minute grains. The evidence that the pebbles have 

 undergone much change since entombment seems to be slight and of 

 hardly material importance. The writer, during a second visit to 

 Commentry in 1910, examined very closely more than 100 coal 

 pebbles scattered through the sandstone up to 60 feet above the 

 Grande Couche. In only two, possibly three, was there appearance 

 of contraction. Almost without exception, the pebbles were coated 

 by a thin film of clay, such as commonly covers, in whole or in 

 part, coal fragments on beds of actual streams in the Appalachian 

 basin. This mud-coat, by drying, might leave a space in which a 

 film of calcite could be deposited; but. aided by strong pocket- 

 glass, the writer could find no evidence of contraction in the 

 pebbles. There had certainly been no change after the sandstone 

 was compacted. Fracture planes are rarely seen but, in very many 

 pebbles, the typical cleavage is distinct. The pebbles of shale are 

 not clay balls ; they are fragments of laminated shale. In any event, 

 the form of the pebbles, shale and coal alike, is that due to stream 

 transport. Some indeed are flattened like coast shingle, but that 

 is due merely to the original form, a block with laminated structure. 

 Every feature of these pebbles appears abundantly on the beds of 

 streams flowing across the Pittsburgh coal area of southwestern 

 Pennsylvania. They leave no possible room for doubt that the 

 coal fragments, like those of sandstone and shale, were deposited 

 by streams flowing over outcrops of coal, shale and sandstone. 



In the larger areas, pebbles are not distributed indiscriminately 

 throughout the mass of sandstone or shale ; they are localized. 

 Petrascheck's descriptions as well as those of the Silesian area by 

 Gothan recall the conditions observed in western Pennsylvania but 



