XIX.] GEOLOGICAL FORMATION OF THE COUNTRY. 711 



in the southern lands we now visited, I must, in the first place, 

 point out that these rocks next the surface of the earth in the 

 south have a much greater resemblance to strata of sand, gravel, 

 and clay than to our granite or gneiss rocks, the type of what 

 is lasting, hard, and unchangeable. The high coast hills, which 

 surround the Inland Sea of Japan, resemble, when seen from 

 the sea, ridges of sand (psar) with sides partly clothed with 

 wood, partly sandy slopes of a light yellow colour, covered by 

 no vegetation. On a closer examination, however, we find that 

 the supposed sandy ridges consist of weathered granitic rocks, 

 in which all possible intermediate stages may be seen between 

 the solid rock and the loose sand. The sand is not stratified, 

 and contains large, loose, rounded blocks in situ, completely 

 resembling the erratic blocks in Sweden, although with a more 

 rugged surface. The boundary between the un weathered 

 granite and that which has been converted into sand is often 

 so sharp that a stroke of the hammer separates the crust of 

 granitic sand from the granite blocks. They have an almost 

 fresh surface, and a couple of millimetres within the boundary 

 the rock is quite unaltered. , No formation of clay takes place, 

 and the alteration to which the rocks are subjected therefore 

 consists in a crumbling or formation of sand, and not, or at 

 least only to a very small extent, in a chemical change. Even 

 at Hong Kong the principal rock consisted of granite. Here 

 too the surface of the granite rock was quite altered to a very 

 considerable depth, not however to sand, but to a fine, often 

 reddish, clay, thus in quite a different way from that on the 

 coast of the Inland Sea of Japan. Here too one could at many 

 places follow completely the change of the hard granite mass 

 to a clay which still lay in situ, but without its being possible 

 to draw so sharp a boundary between the primitive rock and 

 the newl3^-formed loose earthy layers as at the first-named 

 place. We had opportunities of observing a similar crumbling 

 down of the hard granite at every road-section between Galle, 

 C(jlombo, and Ratnapoora, with the difference that the granite 

 and gneiss here crumbled down to a coarse sand, which was 

 again bound together by newly-formed hydrated peroxide of 

 iron to a peculiar porous sandstone, called by the natives 

 rahook. This sandstone fonns the layer lying next the rock in 

 nearly all the hills on that part of the island which we visited. 

 It evidently belongs to an earlier geological period than the 

 Quaternary, for it is older than the recent formation of valleys 

 and rivers. The cahooJc often contains large, rounded, un- 

 weathered granite blocks, quite resembling the rolled-stone 

 blocks in Sweden. In this way there arise at places where the 

 cabook stratum has again been broken up and washed away by 

 currents of water, formations which are so bewilderingly like 



