41 



In opposing the objections of some celebrated 

 naturalists, I am far from maintaining" the an- 

 cient contiguity of all the Smaller West India 

 islands. I am rather inclined to consider them 

 as islands heaved up by fire, and ranged in that 

 regular line, of which we find the most striking 

 examples in so many volcanic hills in Auvergne, 

 in Mexico, and in Peru. The geognostical con- 

 stitution of the Archipelago appears, from the 

 little we know respecting it, to be very similar 

 to that of the Azores and Canary islands. Primi- 

 tive formations are nowhere seen above ground*; 



central chain ; 4th, The idea, that in all places, where great 

 rivers take rise, we may admit great table-lands, or very 

 high mountains. 



* According to Messrs. Moreau de Jonnes and Cortes 

 (Journal de Physique, torn, lxx, p. 129). Dupuget and Le- 

 blond imagined they had recognised granite in the mountain 

 Pelee of Martinico, and in other parts of the Archipelago 

 {Voyage aux Antilles, torn, i, p. 87, 274, and 410). Gneiss 

 has been mentioned as forming a part of the solfatara, at St. 

 Kit's. We cannot be too much on our guard against these 

 indications of rocks in works, the authors of which are less 

 familiarized with the name than with the object. How 

 great was my surprise, when, during my stay at Santa-Fe- 

 de- Bogota, Mr. Mutis showed me in the Journal de Physique 

 -for 1786, p. 321, a paper of Mr. Leblond, where this tra- 

 veller, in other respects accurate, describes the table-land of 

 Bogota, where he resided during some years, as granitic. 

 We find there nothing but secondary formations, sand-stones 

 and gypsum ; not even detached fragments of granite. 



