STEVENSON— INTERRELATIONS OF FOSSIL FUELS. 477 



oped' in shale, with roots at the lower end, which do not penetrate 

 the coal ; at least, if they do, they have been converted into coal and 

 become unrecognizable. 



During a number of years, Barrois, P. Bertrand and some other 

 geologists had studied the roofs of coal beds and they succeeded in 

 classifying them into (i) roofs of sandstone, the grains varying in 

 coarseness, containing much vegetable debris, but leaves have disap- 

 peared; (2) roofs of shale, carbonaceous, with plants, the leaves in 

 fine condition, showing that they had not been transported far and 

 that the deposition had been made in quiet, shallow water; (3) roofs 

 of bituminous shale, black, ampelitic, and with fish remains ; always 

 very thin ; the deposit was made slowly and the water was not free 

 from mud ; (4) roofs of bituminous shale, brown, contains lamelli- 

 branchs of deep- and brackish-water types ; these also were formed 

 slowly and the water was not deep or agitated violently; (5) roofs 

 of calcareous shale with marine shells ; the water was deeper and 

 liable to greater movements. 



If the trees had been floated in, they should occur in roofs of 

 deep water origin, they should not be in roofs, formed in water so 

 shallow that they could not be introduced in vertical position. But 

 they are present in shallow water roofs. At Lievin, they have been 

 obtained from 19 veines or passees (veinules) with typical shallow 

 water roofs and from 7 roofs of intermediate types. Distinctly deep 

 water roofs are not wanting, there being 28 of them, not one of 

 which contains erect trees. Barrois legards the evidence as sustain- 

 ing the assertion of in situ origin for the stems. 



The Central Plateau of France. — The coal basins of central 

 France, about 300 in number, are in large part of little more than 

 local importance ; but some of them are extremely important because 

 the seams attain great thickness and yield a high-grade fuel. All are 

 of limnic origin and the Coal Measures deposits belong mostly to the 

 Stephanien. The general features are much the same in all, so that 

 it is necessary to refer only to the basins with which all are familiar. 



The Coal basin of the Loire or of St. Etienne was studied by 

 Gruner, whose report was published in 1882 and by Grand'Eury,^^ 



81 L. Gruner, Bassin houiller de la Loire," Paris, 1882, pp. 168-173, 

 204-237, 483-486 ; C. Grand 'Eury, " Bassin de la Loire," C. R. Cong. Int. 

 Geo!., Paris, 1900, pp. 521-543, Livret-Guide des Excursions," Xlb., 1900. 



