638 STEVENSON— FORMATION OF COAL BEDS. [November 3, 



Stone, interlaced with Stigniaria roots, as if the sand had been pre- 

 vented from drifting away by a tree or stump." The Millstone Grit, 

 in three divisions and somewhat less than 6,000 feet thick, shows 

 erect Sigillaricc in the lower portion, but most of the stems occurring 

 in the sandstones are clearly driftwood. 



One must keep in remembrance that Dawson's observations were 

 made on the outcropping face of the beds along the coast; one may 

 only conjecture what might be seen if some of the old soils were 

 stripped as was the coal at Parkfield colliery. 



Robb^''* described the rooted stump of Sigillaria, seen by him in 

 the roof of a coal mine. The roots cross the slope, which is 11 

 feet wide ; the conditions are shown well in the plate accompanying 

 his report. He observed many Sigillaria stumps with their attached 

 roots reaching into the coal seams. He notes one case, where a coal 

 parting contains the roots of an erect tree, " which had apparently 

 forced the layers asunder, 6 or 8 inches, for several feet from the 

 extremities of the roots, beyond which the layers of coal unite 

 again." In another, " where a large upright stem appears rooted 

 in a coal seam, the latter seems to have been actually bent down 

 by the superincumbent weight and, at a little distance, to have re- 

 sumed its normal attitude." 



Gosselet studied^*''"' the occurrence of erect stems at various levels 

 in a vertical section and discussed their bearing upon the formation 

 of coal beds. At Lens, there are three coal horizons within a ver- 

 tical distance of 42 meters — Alfred, Leonard and Louise, with thick- 

 ness respectively of 1.4, 1.6 and 0.6 meter. In the one meter thick 

 shale under the Alfred, overlying an irregular passce, he saw two 

 trunks, one resting on the passce, with the underlying coal slightly 

 depressed, while, above, it terminates abruptly in a faux-mur, 5 

 centimeters thick, in direct contact with the coal. No roots could 

 be traced. Five trunks were seen in the Leonard ; one evidently 

 had been floated in and a second was too indefinite for determina- 



'** C. Robb, Geol. Surv. Canada, Rep. I'rogress, 11^74-5, pp. ig6, 20.3, 204, 

 23s, 237. 



'"J. Gosselet, "Note sur les troiics d'arlires verticaux dans le terrain 

 houiller dc Lens," Ann. Soc. Gcol. du Xord., Vol. XXI IL, 1895, PP- I74. 1/5. 

 177-179, 181. 



