274 Notes upon the Diluvial and Alluvial Deposits of the Taffe F'alley. 



These gravel beds seem due to two modifications of the action of running 

 water. First, that by which the gravel was formed and deposited, which 

 evidently left it in knolls, tnmps, and tongues, according to the direction 

 of the current ; and secondly, that by which the knolls and tongues when 

 deposited were scarped down, or half carried away, and by which the sur- 

 face of such plains has been wrought into a variety of channels. 



It seems probable, from the sinuosities affected by these escarpments 

 and channels, that they were cut by a stream much larger than the now 

 existing one, but left like tliat to find its own course down the valley and 

 across the plain ; and not by the far more powerful action of the currents, 

 by which the gravel was originally deposited. 



2. The proportion between the sand and pebbles, of which the gravel 

 beds are composed, is extremely variable; nor would it be easy, without 

 an extensive examination of the phaenomena, both of diluvial gravel beds, 

 and those of actually running waters, to determine at all accurately upon 

 what that proposition depends. 



In many cases, no doubt, the softness of the rock will account for a dis- 

 proportionate quantity of sand ; but where sand was most abundant, it 

 did not appear that the iew scattered pebbles differed materially from those 

 of other localities, nor that there remained any of them in a crushed or 

 semi-pounded state. 



In the beds of now existing rivers, the proportion of sand seems to be 

 greatest where the rapidity is least; and indeed in the gravel beds of such 

 rapid streams as the Rhonddiia, there appears at first sight to be no sand 

 at all, though on turning up a few stones, a proportion will be discovered 

 interspersed below. 



It should be observed that beds of sand and silt sometimes occur inter- 

 stratified with the diluvial gravel, and that their general direction seems to 

 be parallel to the surface of the gravel on which they repose, as might have 

 been expected. Attention to this circumstance may perhaps point out the 

 distinction between silt of the age of the gravel, and that deposited upon it 

 at a subsequent epoch. 



3. The bulk of the pebbles is also extremely variable ; and, in order to 

 make any satisfactory observations upon this point, it would be necessary 

 to examine a valley formed by rocks of considerable geological variety ; for 

 the size of a pebble must be affected by the nature of its parent rock, as 

 well as by the distance which it had travelled, or the degree of violence to 

 which it had been exposed. The upper pebbles appear on the whole to 

 be the largest, though this is doubtful; they are not usually so round as 

 the smaller ones, though to this there are numerous exceptions. 



The figure and shape of a pebble depends upon causes in a great measure 

 similar : schistose and fissile rocks will of course break up, as a general 

 rule, into flat pebbles,- the quartz conglomerate of the old red and mill- 

 stone grit into very round ones. A compact granular rock, such as 



