TYLOR— -QITATERNAPvY GBAVELS. ' 69 



bited in the centre of the cutting. From this section it is evident 

 that the Taff river flowed at one time in a straight line from near 

 Quaker's-Yard Junction to Aberdare Junction, excavating the small 

 valley shown in fig. 5, before it formed the horseshoe bend shown 

 on the plan (PI. Y.), through which the river now reaches the same 

 point, G ; this valley at D is about 40 feet deep, and 200 yards across, 

 and is excavated out of the wedge-bedded pennants, which dip south- 

 east at a smaU angle, but have both horizontal and vertical joints 

 well marked, dividing the rock into cubical masses at angles varying 

 from 80° to 85°. 



The railway engineers have taken advantage of this natural valley, 

 and cut through it to a depth of 60 feet below the bottom of the 

 gravel. 



I wish to caU attention to the position of this well-rolled loose 

 gravel, containing many pieces of Old Eed Sandstone and Millstone- 

 grit, brought from above Merthyr and deposited at a very high level 

 in the small valley cut out of the sandstone rock at Quaker's Yard. 

 The rails are 393 feet above the datum-line at Cardiff, and the gravel 

 is about 460 feet above the datum-line, and 163 feet above the river 

 Taff, close by. If the horseshoe bend (Plate Y.) had been gorged 

 with ice at the gravel-period so as to prevent the passage of the 

 river round the valley B H C, Plate Y,, then the river might have 

 risen to a height which would have enabled it to pass along the line 

 BDEFC ; and the rolled gravel, which is a loose superficial deposit, 

 may have been left by the last great flood. 



Returning to Plate TV. fig. 1 — before reaching E the rails cross 

 a small lateral valley on an embankment 20 feet high, the sur- 

 face of the ground dipping at right angles to the rails at an angle 

 of 15°, and so reaching the river at an angle of 30°. The great 

 gravel- cutting E P is now entered, and shows a maximum thickness 

 of gravel of 70 feet. 



The great cutting, E F, nearly in the line of the old incline, com- 

 mences with a series of large blocks of sandstone from 6 to 9 feet 

 long, laid at an angle of 1° N., and forming the last or topmost 

 layer of the gravel. These great pieces of rock have evidently only 

 descended a few yards from the mass from which they were detached. 

 The gravel becomes 55 feet thick at a point (E) where the sand-rock 

 is exposed (see fig. 6, p. 70). The sand-rock in the cutting is visible 

 for nearly 60 yards to a height of 7 feet above the rails at the maxi- 

 mum. This gravel may be observed to be roughly parallel to the 

 top surface of rock, which slopes gently to the south. The greatest 

 thickness of gravel with large boulders occurs near F, reaching 

 70 feet. The slope of the surface is given at different points. 



The larger blocks are often arranged in lines or planes of deposi- 

 tion, forming a distinct measurable angle. In the larger gravel-cut- 

 ting, E F, they are not in any case more than 15° ; near the viaduct 

 they reach an angle of 35° near the rock. These planes follow, and 

 are nearly parallel with, the surface of the rock which crops out at E, 

 and are so arranged that the last bed of gravel forming the surface, 

 being also the last superficial deposit, is also almost parallel with the 



