1847.] including Notices of Sumatra, the Malay Peninsula, fyc. 529 



an inch to many feet, below the surface, are frequently discovered by 

 sections for roads and pits for planting spice trees, &c. It is obvious 

 that the hypothesis which I am now considering will not explain such 

 cases. 



There is another phenomenon of frequent occurrence connected with 

 the position of fragmentary rocks which this hypothesis ought to 

 include if it be made the foundation of any general theory. In sections 

 across strata they are almost invariably seen to be more or less curved 

 as they approach the surface. Before reaching it however they some- 

 times gradually, but often abruptly, lose their compact form and become 

 masses of fragments. In some cases these are almost insensibly min- 

 gled with the superincumbent soil till all trace of the stratum disap- 

 pears. But it is not uncommon to see the curve pass into a line more 

 or less horizontal, and even bent downwards, and the fragments streaming 

 away as it were in a layer of which the direction seems to have no 

 relation to the parent stratum, but which generally possesses or 

 approaches to parellelism with the plane of the surface. It is true that 

 of some of these cases the hypothesis which we are at present pursuing 

 might seem to afford a solution. Thus suppose a thin layer of hard 

 sandstone to rest on a bed of soft sandy clay or unlaminated shale, both 

 inclined and having their outcrop on the slope of a hill, a certain depth 

 from the surface of the slope would be subject to the action of meteoric 

 forces which would cause the sandstone to break up into fragments 

 and the sandy clay to become loose and open. The sandstone rubble, 

 if heavy, might possibly tend to descend or settle in a perpendicular 

 line through the upper pulverulent to the lower and more compact soil, 

 and, at all events, as the soil below it was carried away, the rubble would 

 descend along the line of the slope, the heavier fragments remaining 

 at and near the point of outcrop, those of medium size streaming 

 further down the slope, and the smallest borne away with the fine 

 sand and clay to lower levels ; — the possibility of the existence of such 

 lines of rubble, their breadth down the slope from the line of outcrop, 

 and the quantity and size of the fragments, being always determined 

 by the texture of the recipient bed of clay or sand, and the declivity 

 of the hill. Where the slope of the hill consisted of a succession of 

 similar layers and beds, the lower layers of rubble would, in course of 

 time and in favourable positions, become covered with soil brought 



3 z 2 



