70 



PROCEEDINGS OF THE GEOLOaiCAL SOCIETY. 



top surface of tlie bed of stones on which it lies, just as the last bed 

 is almost parallel with the mean angle of the surface of the rock on 

 which it is spread by the action of water (see E). This section is 

 to natural scale. 



Fig. 6*. — Section along cutting of Taff-Valley Railway. 



BRIDGE 



The exposure of rock at many different points of the horseshoe 

 bend near the piers of the viaduct, and in the bed of the river, adds 

 to the interest of the sections at Quaker's Yard. I shall now de- 

 scribe the character of this large rock-basin in which the gravel 

 under consideration is contained. The first place at which I ob- 

 served the rock was near the river, where a small quarry is worked 

 close to one of the piers of the viaduct. The dip is slightly south-east, 

 and the stone is divided by nearly vertical joints ; there is another 

 quarry of the same character on the opposite bank, between the 

 tramway-bridge and the south-east end of the railway-viaduct, but 

 close to the latter. In following the tramway round the horseshoe 

 bend the rock is frequently visible ; in the river-bed it is concealed 

 by a coating of gravel as far as the junction of the stream at the 

 north-east point of the bend. The gradient of this small stream is 

 steep, and the water enters the river with some velocity, running 

 nearly parallel with the Taff near the junction. The water of the 

 Taff, up to the junction, is smooth, and flows quietly over a gravel 

 bottom; but the lateral stream, coming in with added velocity, 

 causes the stream of the Taff Eiver almost to split into two different 

 paths, one taking the east, and the other, or larger portion, crossing 

 the bed of the river, and hollowing out a deep channel in the rock 

 close to the west bank. The river flows below the junction at H 

 entirely over rock, the gravel not being able to rest on account of 

 the velocity of the current. 



The action on the river-bed corresponds with what I have repre- 

 sented in other sections that I have made, which show that the 

 work of denudation, or cutting out of the solid rock, is principally 

 due to running water, either in the lateral streams themselves, or in 

 the main river immediately after the side-stream has emptied itself 

 into it. Of this I shall bring forward other examples at another 

 ^ Fig. 6 is also reversed. See Plate IV. fig. 1 at E. 



