39 



have belonged to the same chain, or have always 

 been isolated. Supposing- an irruption of the 

 ocean to take place either in the eastern part of 

 the island of Java*, or in the Cordilleras of 

 Guatimala and Nicaragua, where so many burn- 

 ing mountains form but one chain, that chain 

 would be divided into several islands, and per- 

 fectly resemble the Caribbean Archipelago. The 

 union of primitive formations and volcanic rocks 

 in the same range of mountain has nothing in it 

 strange : it is very distinctly to be seen in my 

 geognostical sections of the Cordillera of the 

 Andes. The trachytes and basaltes of Popayan 

 are separated from the system of the volcanoes 

 of Quito by the mica-slates of Almaguer ; the 

 volcanoes of Quito from the trachytes of Assuay 

 by the gneiss of Condorasto and Guasunto-j~. 

 There does not exist a real chain of mountains 

 running South -East and North -West from 

 Oyapoc to the mouths of the Oroonoko, and of 

 which the Smaller West India islands might be a 

 northern prolongation. The granites of Guiana, 



• Raffles, History of Jam, 1817, p. 23—28. The principal 

 line of the volcanoes of Java, on a distance of 160 leagues, 

 runs from West to East, through the mountains of Gagak, 

 Gede, Tankuban-Frahu, Ungarang, Merapi, Lawn, Wilis, 

 Arjuna, Dasar, and Tashem. 



+ Consult the Nivellement Barometrique, and Tableau des 

 Formations des Andes, in my Obs. Ast, 9 vol. i, p. 303, and 31 1 ? 

 (N, 125—220.) 



