1817.] GEOLOGY OF THE WEST INDIES. 145 



larly thrown about in every direction. At the northern 

 extremity of this crack lies what is called the cave, whence 

 there issued, about 15 or 20 years ago, a flood of water 

 and rocks, which ran down the valley, at present called 

 the valley of Faujas, in the utmost disorder. I am inclin- 

 ed to think that water only came from the crack, and that 

 it ran over the mountain, sweeping in its course all the 

 small stones and cinders, leaving those that were too large 

 to be moved. Ihis irruption of water was cool, and 

 without any apparent connection with heat, though it was 

 most probably ejected by the force of some elastic fluid. 

 Montserrat. I passed close to the leeward side of 

 the island of Montserrat, but did not land. The soutli 

 side had an appearance of being partly composed of solid 

 rock, and the rest of the island might be supposed to be 

 constituted of cinders mixed with loose rocks, as it consists 

 of one mountain, the sides of which are furrowed by the 

 rain, gently, and not in precipices, as would have been the 

 case had there been many currents of solid lava, which 

 circumstance, with the flatness of the coast, and the gradual 

 ascent of the mountain, would seem to indicate a great 

 proportion of cinders. 



Ncvu consists of one mountain in 'the middle, a trun- 

 cated cone, I suppose about 2000 feet high; and one 

 small elevation to the south, called Saddle-Hill, and ano- 

 ther to tlie north, called Round- Hill; the rest of the island 

 is a gradual descent from these three hills to the sea. It 

 is composed of large masses of rocks, in beds of cinders, 

 gray, red, and black, of various degrees of solidity, from 

 the pumice to the compact lava; the black crystals I take 

 to be augite, or perhaps what Werner calls the basaltic 

 hornblend, of the Cape de Gate in Spain, many of the 

 Vql. I. K 



