STE'VE'NSON, COAL BASIN OF DECAZEVILLE, FRANCE 275 



acres, south from Valzergues, is covered "with a mass of granite blocks 

 with rounded angles, 13 to 16 feet in diameter, and bound together by a 

 granitic sand rich in feldspar. This rests on mica schist and in general 

 character is closely related to a granitic mass at a little distance south- 

 east. Along the western border and extending to the granite area inside 

 of the limiting fault, the earliest deposit is a "granite conglomerate" 

 which can be followed from near ]\Iontbazens to the granite north from 

 Eiou Vieux. The fragments are of large size even near Montbazens, but 

 they increase northwardly until beyond Eiou Vieux, as one approaches 

 the granite, very many of them are "10 metres cubes" and the intervals 

 are filled with granitic sand, holding blocks of granulite. A small patch 

 of similar conglomerate exists farther north on the granite itself. 



When one considers the colossal dimensions of the blocks at these 

 localities, the conception of transport by the insignificant streams enter- 

 ing and traversing the basin becomes at least improbable. The phenom- 

 ena point rather to atmospheric action ; the rounding of the angles in the 

 great blocks west from Montbazens is a commonplace feature of granite- 

 weathering beginning at the joint planes. The writer long ago observed 

 many instances in Colorado, one of which he placed on record.^ But it is 

 unnecessary to go far from Decazeville to find a deposit like the "granite 

 conglomerate." Along the railroad from Bort to Aurillac and thence to 

 Capdenac, one sees at many places a thick deposit of granitic sand hold- 

 ing great rounded and angular blocks of granite. At Viescamp-sous- 

 Jalles, near Aurillac, this is well exposed in a long cut, and it covers 

 the hillsides for a considerable distance southward toward Capdenac. 

 Where this readily disintegrating granite prevails, the valley widens, but 

 beyond in the schist it narrows. The granite area within the Decazeville 

 basin must be largely of the readily disintegrating type, for its topogra- 

 phy is hardly more abrupt than that of the Coal Measures. 



The presence of these conglomerates, the absence of deposits earlier 

 than Bourran in the western portion of the basin and the resemblance to 

 dejection cones shown by the earliest deposits in the southeast corner, all 

 indicate that for a long period much of the basin was dry land, that the 

 water encroached very slowly and that the entire area was not submerged 

 or water-soaked prior to deposition of the Bourran system. The isolated 

 conglomerate south from Valzergues cannot be regarded as affecting the 

 question of the original extent of the basin; and it appears to be alto- 

 gether probable that Bergeron, in his original paper, defined the limits of 

 the depression as nearly as possible. The width can have suffered very 



"U. S. Geogr. Expl. W. of 100th Mer., vol. iii, p. 348, 1876. 



