239 



Hornblende, sometimes pure and forming in- 

 termediate strata, at other times mixed with 

 granite, as in the basanites or basalt of the an- 

 cients, may, by itself, furnish all the iron con- 

 tained in the black and stony lavas. This quan- 

 tity amounts in the basalt of the modern mine- 

 ralogists only to 0*20, while in hornblende it 

 exceeds 0 30. 



Were these granites and these mica-slates of 

 Gomera anciently united to the chain of Atlas, 

 as the primitive mountains of Corsica appear to 

 be the central nucleus of Bochetta and the Apen- 

 nines ? This question can never be solved, till 

 mineralogists shall have visited the islands that 

 surround the Peak, and the mountains of Mo- 

 rocco covered with eternal snows. Whatever at 

 some future day may be the result of these in- 

 vestigations, we could not admit with Mr. Pe- 

 ron % u that in none of the Canary Islands do 

 we meet with true granites ; and that, the whole 

 of the Archipelago being exclusively volcanic, 

 the partisans of the Atlantis must suppose, what 

 is equally destitute of probability, either a con- 

 tinent perfectly volcanic, or that only the vol- 

 canic parts of that continent were spared in the 

 catastrophe, by which it was swallowed up." 



From the information of several well instruct- 

 ed persons, to whom I addressed myself, I 

 found, that there are calcareous formations in 



* Voyage de Decouvertes aux Terres Australes, t. i, p. 24. 



