234 Correspondence. 



removable; the rest will yield to patient investigation, if only we 

 do not assume that there is nothing to be investigated. 



In the meantime I confidently rely on two conclusions, which in 

 our islands are specially applicable to Palaeozoic districts, but apply 

 mutatis mutandis to rocks of all ages. These are — 



1st. The sea has removed vast masses of roch, and left undulating 

 surfaces, the highest points of which ultimately become the summits 

 of mountains. 



2nd. When those undulating surfaces are raised high into the air 

 they are attacked by the atmospheric agencies, and hills, valleys, and 

 plains gradually carved out of the rock-mass below their particular 

 features depending on original varieties in the nature of that mass, 

 and variations in the action of the atmospheric agencies. The latter 

 depend largely on variations of temperatm-e, by which water is made 

 to assume the different forms of vapour, water, snow, and ice. 



It must be recollected that the forms of our Paleozoic grounds are 

 of very ancient date, anterior to the period of the New Eed Sand- 

 stone, and that the great denudation of the Older Pa,l£eozoic Eocks 

 took place even before the deposition of the Old Eed Sandstone. 

 The time, then, during which the atmospheric agencies have been 

 modelling the minor features is inconceivably great. The recent 

 temporary depression beneath the waters of the glacial sea did little 

 or nothing in the way of denudation, the principal effect then, being 

 the transport of blocks, or the washing about of materials, already 

 loose on the surface. 



Much instruction as to the amount of atmospheric action may be 

 gained by comparing volcanic cones with each other. I observed in 

 Java that small volcanic cones of recent origin had their sides quite 

 smooth and even, while others of older date, as was shown by the 

 young trees growing on them, began to show gullies widening and 

 deepening on all sides. The flanks of the great volcanic mountains 

 were a mere series of deep glens, separated by sharp knife-edged 

 crests, radiating like the spokes of a half-shut umbrella, as described 

 by Dr. Junghuhn. (See Lyell's Elements, 6th ed., p. 620.) 



Still older volcanos, as those in the South Pacific, described by 

 Dana, have merely narrow vertical walls, radiating from the central 

 mass, between flat-bottomed valleys, which gradually contracting 

 towards the interior where at the head of each may be seen a little 

 rill of water leaping from crag to crag, still going on with the work 

 it has performed, and to which it seems at first so utterly inadequate. 



It has sometimes occurred to me to ask how long grass has 

 existed ? and especially those grasses which make our matted turf ? 

 Conclusions as to the rate of atmospheric erosion drawn from our 

 turf-covered downs would be apt to lead us astray if applied to hills 

 not so covered. In many parts of Australia, for instance, where you 

 come to ride over a hill that looks quite green in the distance you 

 find you can see the ground between the roots of the grass, very 

 much as if you were riding through young wheat. The rain, when 

 it does come down in a torrent, must exert much more effect on such 

 ground than where there is matted turf. 



