'91-'.] STEVENSON— THE FORMATION OF COAL BEDS. 447 



local, for the upper division of the Pittsburgh has been removed and 

 its fragments distributed in the overlying sandstone. The condition 

 is in no wise unusual. In a later report, he says that the Home- 

 wood sandstone, the closing deposit of the Beaver, contains in its 

 lower portion disks and pots of coal of shape such as one might 

 expect where a soft mass has been entangled in sandstone and sub- 

 jected to heavy pressure. At another locality, the sandstone con- 

 tains coal fragments, some quite large, which he thinks represent a 

 coal bed removed from the area during deposition of the coarse 

 more or less pebbly rock. As that rock in this locality fills an old 

 valley, it is possible that the larger fragments represent boggy 

 material accumulated on undermined banks. In some places men- 

 tioned by this observer, the fragment-bearing sandstone rests on 

 the uneroded Pittsburgh coal. In all probability, a sand-laden 

 stream, sweeping across the area, removed part of the unconsoli- 

 dated coal and carried it away to be deposited on an overflow sur- 

 face along with living plants, as well as with the sand and hardened 

 coal brought from a distance. The same observer mentions a 

 locality, where the coal had been cut out and replaced with a con- 

 fused mass of coal, slate and sandstone. I. C. White"* has recorded 

 several instances similar to the last, in widely separated localities, 

 where the coal has been replaced with a conglomerate of worn and 

 rounded pieces of limestone, coal, slate, and sandstone. 



" Washouts " similarly filled have been reported by other ob- 

 servers in Pennsylvania and elsewhere. The conditions are pre- 

 cisely such as one can see in the streams now flowing across the 

 bituminous field and the conclusion is the same in both cases — that 

 the stream flows across the outcrops of rocks whose fragments are 

 in its bed or in the " bottom " deposit of its banks. The Banc des 

 Chavais in the Grande Couche of Commentry belongs to this type. 



Renault"*" studied very carefully the coal pebbles found in the 

 sandstones of the Commentry coal basin. Some resemble freshly 



'*I. C. White, Second Geo!. Surv. Penn., Report Q, 1878, pp. 114, 268; 

 Report Q2, 1879, pp. 270, 274-276; Report Q3, 1880, p. 176. 



^^ B. Renault, " Quatrieme note pour servir a I'histoire de la formation 

 de la houille," Comptes Rendus, Vol. 99, 1884, p. 201. 



