ENDOGENOUS OR ERUPTED ROCKS. 239 



syenite. ( 247 ) Where granite is found in large insulated 

 masses,, having a slightly vaulted ellipsoidal form, whether 

 it be in the Hartz district, in Mysore, or in Lower 

 Peru, it is surmounted by a kind of crust divided 

 into blocks. These " seas of rocks," as they are 

 sometimes called, are probably occasioned by a con- 

 traction of the distended surface of the granite when 

 first upheaved. ( 248 ) In Northern Asia, on the ro- 

 mantic shores of Lake Kolivan on the north-western 

 slope of the Altai, ( 249 ) as well as at las Trin- 

 cheras ( 25 ) on the declivity of the maritime chain of 

 Caraccas, I have seen divisions in the granite, which were 

 probably caused by similar contractions, but which ap- 

 peared to penetrate deep below the surface. Farther to 

 the south of Lake Kolivan, towards the boundary of the 

 Chinese province of Ili (between Bucktarminsk and the 

 river Narym), the appearance of the erupted rocks, in 

 which there is no trace of gneiss, is more remarkable 

 than I had ever before seen in any part of the globe. 

 The granite, which always scales at the surface, and is 

 characterised by tabular divisions, rises, on the steppe, in 

 small hemispherical hillocks of six or eight feet in height^ 

 and sometimes, like basalt, in small mounds with narrow 

 streams on opposite sides of their base. ( 251 ) At the cata- 

 racts of the Orinoco, as in the Fichtelgebirge in Bavaria, 

 and in Gallicia, as on the Pappagallo between the high 

 lands of Mexico and the Pacific, I have seen large flat- 

 tened globes of granite, which could be separated into 

 concentric layers like certain basalts. In the valley of 

 the Irtysch, between Buchtarminsk and Ustkamenogorsk, 



VOL. I. S 



