468 PLEISTOCENE DEPOSITS. , CHAP. 



By observations of this kind the fresh-water origin of the gravel- 

 beds is sufficiently plain, as far down as the great vale expanding 

 below Abingdon. 



The gravel -beds were formed in broad waters, which were divided 

 into irregular expansions above Yarnton and Sandford, and above 

 the long- winding valley which extends from Moulsford to Reading. 

 Whether we regard these ' broads' 7 as formed at two or more levels 

 by natural impediments at Yarnton, Sandford, and Moulsford, or 

 treat the whole as one ramified loch gradually emptied by the 

 deepening of one gorge below Wallingford, we shall arrive at the 

 same general idea of the arrangement of the gravel being due to 

 the fluctuations of these waters. As they retired southward they 

 left the gravel at lower and lower levels, with no distinct steps, 

 but rather on continuous gentle slopes, broken here and there by 

 irregularly undulated mounds. 



The gravelly strata in the uppermost of these sheets of water 

 lie chiefly on the north side from Ampney to Yarnton, and on the 

 east side from Kidlington to Cowley. One may believe this to be 

 in some degree the effect of the prevailing winds from the south- 

 west, for these depend on conditions far older than the pleistocene 

 ages, and their effects have been traced 'in mesozoic and even in 

 palaeozoic periods by philosophical observers like De la Beche and 

 Sorby. Thus, perhaps helped by floating ice, we may understand 

 the drift of flints which runs so far in these gravels and beyond 

 them to the north of the Berkshire downs, and lies in deep accu- 

 mulations about Stadhampton, north of the Chilterns. 



In a general manner, we may assign to the true sea-drift from 

 the north the earliest in our history in the course of the Thames 

 drainage above Pangbourn, a height not exceeding 750 feet ; to 

 the flint-drift from the south, which may have been marine, and 

 must have been, at least in part, of later date, something like two- 

 thirds of that elevation; and to the gravel-beds, in which both 

 flints and quartzites have been mixed with oolitic fragments, some- 

 thing above one-third. 



In descending the Valley of the Thames below Reading we 

 remark the scattering of gravel (with northern pebbles) on the hill 

 summits and slopes, and the gathering of thicker beds in the lower 

 ground. At Hurley-Bottom, below Henley, a very productive spot 

 for mammalia (mammoth, rhinoceros, bear, horse, ox, deer), the 



