19 



a load of varied vegetation, — the effect of the whole is at once picturesque and imposing 

 in an extraordinary degree. I can only afford to notice a few of the most remarkable rocks, 

 although a faithful description of the whole is desirable, since the Chinese quarrymen are 

 proceeding so rapidly in their work of destruction that it is to be feared these grand and 

 singular natural phenomena will, in a few years, have been entirely obliterated. 



One of the most striking of the rocks is a connected pile of great bulk and extent which 

 stretches from the base of the hill across the beach into the sea. It is cleft in a few pla- 

 ces by narrow dark chasms, three of which, in particular, divide it into four principal por- 

 tions. The inner is a great cuboidal mass based in the land, and its sides, being very 

 slightly furrowed, rise perpendicularly like solid walls. At the NW. angle rises what may 

 be called a great columnar turret, partially severed from the mass by a deep narrow chasm, 

 and traversed from top to bottom by deep channels divided by narrow ridges, and wilh its sum- 

 mit ascending some yards above the level of the pile. A tree rises from the pinnacled sum- 

 mit, and the whole of the western side of the columnar mass is reticulated by its roots. 

 The principal of these run down the channels to the ground, and, as they descend, 

 give out numerous lateral branches which closely embrace the rock, folio wing its sinuositie3 

 and entering the cleft between it and the main mass. The roots are so numerous and so 

 intcrlaced that they conceal the greater part of the rock; and, towards the summit, where 

 they converge beneath the trunk, only some narrow portions of the ridges are visible. 

 These differ so little in shape and colour from the roots, and are so closely united with them, 

 that, from some points of view, the whole appears as the solid stole of the tree. From other 

 points of view some of the acicular summits of the turret are free from the net work of roots, 

 and others pierce through it. One of the long nearly horizontal branches of the tree which 

 stretches seaward above the pile is literally covered with air plants save at its extremity. 

 The second mass of the pile is also cubical, but distinguished by a bulky awkward looking 

 protuberance, which rises above it towards it edge and leans to one side, but for which it 

 is impossible to find any architectural prototype. Its rounded shapeless edges and partially 

 depressed sides give it a form more resembling a bit of ginger root than anythiog ever 

 shaped by art. The W. side of the mass above which it rises is traversed by a few fur- 

 rows nearly vertical, but dipping a little from N. to S. The upper portion, or about two 

 thirds, of the east side is rather more furrowed. The lower portion bulges out and is 

 smooth. The furrows incline to the southward, bending more decidedly in that direction 

 as they reach the bulging portion of the rock, on the upper surface of which they termi- 

 nate. The third portion of the mass is smaller than the preceeding ; and its upper surface 

 slopes seaward. Its sides are much channeled. The upper portion of the southern face 

 of the fourth or external mass slopes curvately from the cleft that separates it from the 

 preceeding mass till its lower portion approximates to vertical and dips beneath the 

 sea. Its surface is on all the three exposcd sides an uninterrupted series of sharp 

 ridges and included furrows which vary in depth and breadlh. The depth is from 

 two to live feet. Some are broad and deep semicylindrical concavities — the surfaces of 

 which are grooved or fluted all round. The grooves are continued without interruption, 

 and with all their regularity of form, beneath the levcl of the sea where they are quite 



