1913-] STEVENSON— FORMATION OF COAL BEDS. 85 



the bed, where the roof or the underclay replaces more or less of the 

 coal. Rolls in the roof usually consist of material differing in char- 

 acter and arrangement from the overlying shale, as though deposited 

 in channels of streamlets made after formation of the coal. The 

 underclay swellings may have been laid down in drowned channels 

 made anterior to formation of the coal and occupied after that 

 formation had been begun. Variation in direction of channels 

 during accumulation of the beds might account for distribution of 

 the fragments but the existence of such waterways within this area 

 is problematical and it is well to seek another explanation. 



Phillips's hypothesis that the fragments were transported by 

 trees, uprooted from banks of streams, has found favor with alloch- 

 thonists and autochthonists alike; but there are serious objections to 

 it. The weight of some fragments, upwards of lOO kilogrammes, is- 

 too great to admit of transportation by Stigmaria, while the presence 

 of blocks of mud would suggest that hollow trees had shared in 

 the work. In any event, there would always remain the remarkable 

 purity of the coal, so difificult to explain in view of the great amount 

 of inorganic material known to be transported by floating trees. 

 There seem to be insuperable difficulties in the way of a conception 

 that the presence of fragments is due to the agency of trees grow- 

 ing outside of the area in which coal was forming. Objection to the 

 hypothesis of transport by floating ice is equally serious. Beyond 

 doubt there were widespread changes in climatic conditions toward 

 the close of the Palaeozoic, but attempts to reconcile the tropical 

 character of the Nord-basin flora with a cold climate have not been 

 successful. The markings on the fragments do not resemble those 

 made by glacial action. 



The presence of fragments in the mur is proof that they were 

 brought in prior to formation of the coal, when streams were dis- 

 tributing the detritus which became the mur. Stigmaria: became 

 rooted in that and enlaced the fragments, which some day they were 

 to transfer to the coal. The fall of trees, overturned in the marsh 

 by age or wind, tore portions of the mur from below ; fragments, 

 there encased, came gradually to the surface of the coal ; at times a 

 stump fell into a stream and its load would be deposited in the 



