IX STREAMS OF STONES 207 



those found in the Silurian formations of Europe ; the hills are 

 formed of white granular quartz rock. The strata of the latter 

 are frequently arched with perfect symmetry, and the appear- 

 ance of some of the masses is in consequence most singular. 

 Pernety^ has devoted several pages to the description of a Hill 

 of Ruins, the successive strata of which he has justly compared 

 to the seats of an amphitheatre. The quartz rock must have 

 been quite pasty when it underwent such remarkable flexures 

 without being shattered into fragments. As the quartz insensibly 

 passes into the sandstone, it seems probable that the former 

 owes its origin to the sandstone having been heated to such a 

 degree that it became viscid, and upon cooling crystallised. 

 While in the soft state it must have been pushed up through 

 the overlying beds. 



In many parts of the island the bottoms of the valleys are 

 covered in an extraordinary manner by myriads of great loose 

 angular fragments of the quartz rock, forming " streams of 

 stones." These have been mentioned with surprise by every 

 voyager since the time of Pernety. The blocks are not water- 

 worn, their angles being only a little blunted ; they vary in 

 size from one or two feet in diameter to ten, or even more than 

 twenty times as much. They are not thrown together into 

 irregular piles, but are spread out into level sheets or great 

 streams. It is not possible to ascertain their thickness, but the 

 water of small streamlets can be heard trickling through the 

 stones many feet below the surface. The actual depth is 

 probably great, because the crevices between the lower fragments 

 must long ago have been filled up with sand. The width of 

 these sheets of stones varies from a few hundred feet to a mile ; 

 but the peaty soil daily encroaches on the borders, and even 

 forms islets wherever a few fragments happen to lie close 

 together. In a valley south of Berkeley Sound, which some of 

 our party called the " great valley of fragments," it was necessary 

 to cross an uninterrupted band half a mile wide, by jumping 

 from one pointed stone to another. So large were the fragments, 

 that being overtaken by a shower of rain, I readily found 

 shelter beneath one of them. 



Their little inclination is the most remarkable circumstance 

 in these "streams of stones." On the hill-sides I have seen 



1 Pernety, Voyage mix Isles Maloiiiues, p. 526. 



