PLATEAU GEAVELS. 219 



the former follow the original valley., the latter have been carried 

 along the new course of the river. It may be noted that when 

 the terraces of Valley Gravel were formed, the bed of the Yar must 

 have been about 20 feet higher than now, that is at about the level 

 of the watershed. 



In some parts of the broad tract of Lower Greensand which 

 runs eastwards to Sandown, the remains of an old gravel-covered 

 plain are very striking. They occur at a fairly constant level, 

 but there are scattered patches also at a variable height on the 

 sides of the hills. South-west of Arreton, for example, several 

 patches of gravel, associated with brick-earth, occur in an ir- 

 regular manner on the flanks of St. George's Down. They are 

 clearly intermediate in age between the Plateau Gravel on the liill- 

 top, and the Valley Gravel of Horriugford, and, as might have 

 been anticipated, contain a larger proportion of Lower Greensand 

 material than does the older gravel. The best sections are to be 

 found in three road cuttings west-south-west of Arreton. 



Near Newchurch good examples of gravel-covered plateaus may 

 be observed. One extends through the village and along the top 

 of the steep bank overhanging the alluvial flat, showing in its course 

 a tendency to slope down towards the north, that is towards the 

 valley of the Yar. Another, cut by denudation into a sinuous out- 

 line, is well exposed at Skinner's Hill, on the road from Newchurch 

 to Borthwood, and is worked in many places for gravel. These 

 patches, more stony than those near Arreton, are associated also 

 with brick -earth in an irregular manner, which makes it impossible 

 to draw a hard and fast boundary for this deposit. 



The hill near Sandford is capped with a conspicuous outlier of 

 these gravels at a height of 200 feet above the sea ; and similar 

 but very thin patches occur near Apse and Apse Heath. At 

 Alverstonethe gravel caps the top of the steep bank which bounds 

 the modern alluvial flat, as at Newchurch. 



Two more patches belonging to this same series of outliers 

 occur on the top of the cliff between Shanklin and Sandown. In 

 the more southern o£ the two, at Little Stairs Point, may be seen 

 at different points on the cliff", sand and loam with flints, 9 feet 

 thick ; flint gravel, 12 feet thick ; and loam and brick-earth 6 feet, 

 with flint gravel 1 foot thick underneath. 



Lastly a few small patches occur on the north side of the Yar 

 between Alverstone and Yarbridge. Their mode of occurrence is 

 precisely similar, except that the ridges on which they lie slope to 

 the south, and more rapidly than those on the south of the Yar 

 slope to the north. 



It will be gathered from this disposition of the deposits that the 

 lowest part of the ancient valley in which this sheet of gravel was 

 laid down occupied about the same position as the bottom of the 

 existing valley, and that then, as now, the ground rose rapidly to 

 the north towards the Central Downs. Judging from their mode 

 of occurrence, we may infer that the gravels of Blake Down, 

 Newchurch, Alverstone, and the Sandown Cliffs were approximately 

 contemporaneous. 



