1836.] 



Southern India. 



331 



volcanic hills still existing unaltered. That the whole of the granite 

 rocks, however, as well as the more recent basalts, have resulted from 

 successive streams of lava, scarcely can be doubted : at the same time 

 it is impossible that they could have been erupted as they now stand : 

 they appear to be subverted fragments, belonging to a former crust 

 which has suffered a general bouleversement. Admitting the vertical 

 position to be the result of such an event, the surface of the crust is 

 unexpectedly level, for all considerable inequalities are referable to 

 subsequent dislocation. 



Another phenomenon connected with the question of subversion is 

 the exfoliation of mountain masses, as well as the smaller fragments, 

 in concentric layers. This is interesting as being the cause to which 

 the remarkable convexity of the trap and granite hills is attributable : 

 it is by no means confined to this country, the same convexity of sur- 

 face being found everywhere, both in the primitive and transition rocks. 

 On the slightest examination it will be distinctly visible, that the layers 

 exfoliate, without any reference to the original structure or stratifica- 

 tion of the rock itself; in gneiss, and granites having vertical stratifica- 

 tion, or rather having the strata placed in a vertical position, we see the 

 layers scaling off horizontally in cavo-convex fragments, the fissures 

 cutting across the original lines of stratification ; and not only this* but, 

 in porphyritic granites, a single crystal of felspar is often intersected , 

 in several places, across its own lines of cleavage. This is a satisfactory- 

 proof that exfoliation is in no way influenced by the original structure 

 of the rock, but is, perhaps, in some way connected with the radiation of 

 heat from the surface : whatever may be the cause, the process itself is 

 sufficiently apparent, and in its effects important to theory, as it appears 

 to have altered very considerably the outline of the primary rocks. 

 We learn, also, by the numerous instances of half demolished peaks, 

 that the exfoliation does not proceed from the top downward, but that 

 the peak or fragment is at once truncated near the base in repeated 

 concentric lines, leaving a convex projecting surface. The thick plates 

 or caps may be often seen still adhering to the convex surface, in which 

 situation they usually split into divisions, each separate piece again 

 exfoliating in its own mass. Many of the blocks on hill-tops are boul- 

 dered, but, when these blocks are of the same nature as the rock on 

 which they rest, and have their lines of stratification also vertical, it 

 may be taken for granted that they have at some time formed part of the 

 subjacent rock, and are not transported boulders. The convexity of 

 the dome-shaped hills, with their exfoliating surface, is very apt to lead 

 to the conclusion that they are volcanic masses erupted in a fluid state 

 on the spot ; the layers have been taken for successive flows of lava, 

 while corresponding craters have been sought for in vain : these layers 

 have also sometimes been confounded with gneiss. 



