PHYSICAL GEOGRAPHY. 293 



of direction extends in like manner to the chain of the Andes, 

 which is divided into two parallel branches, affecting not only 

 the littoral portions,* but even the eastern Cordilleras. In 

 the latter, civilization had its earliest seat in the South Amer- 

 ican plateaux, where the small Alpine lake of Titicaca bathes 

 the feet of the colossal mountains of Sorata and Illimani. 

 Further to the south, from Valdivia and Chiloe (40° to 42"^ 

 south latitude), through the Archipelago cle los Cho?ios to 

 Terra del Fuego, we find repeated that singular configuration 

 oi fords (a blending of narrow and deeply-indented bays), 

 which in the Northern hemisphere characterizes the western 

 shores of Norway and Scotland 



These are the most general considerations suggested by the 

 study of the upper sm-face of our planet with reference to the 

 form of continents, and their expansion in a horizontal direc- 

 tion. We have collected facts and brought forward some 

 analogies of configuration in distant parts of the earth, but we 

 do not venture to regard them as fixed laws of form. When 

 the traveler on the declivity of an active volcano, as, for in- 

 stance, of Vesuvius, examines the frequent partial elevations 

 by which portions of the soil are often permanently upheaved 

 several feet above their former level, either immediately pre- 

 ceding or during the continuance of an eruption, thus forming 

 roof-like or flattened summits, he is taught how accidental 

 conditions in the expression of the force of subterranean va- 

 pors, and in the resistance to be overcome, may modify the 

 Ibrm and direction of the elevated portions. In this manner, 

 feeble perturbations in the equilibrium of the internal elastic 

 forces of our planet may have inclined them more to its north- 

 ern than to its southern direction, and caused the continent 

 in the eastern part of the globe to present a broad mass, whose 

 major axis is almost parallel with the equator, while in the 

 western and more oceanic part the southern extremity is ex- 

 tremely narrow. 



Very little can be empirically determined regarding the 

 causal connection of the phenomena of the formation of con- 

 tinents, or of the analogies and contrasts presented by their 



* Humboldt, iu Poggeudorf's Annalen der Physik, bd. xl., s. 171. 

 On the remarkable fiord formation at the southeast end of America, see 

 Darwin's Journal {Narrative of the Voyages of the Adventure and Bea- 

 gle, vol. iii.), 1839, p. '^QQ. The parallelism of the two mountain chains 

 is maintained from S*-* south to 5^ north latitude. The change in the 

 direLtiou of the coast at Arica appears to be in consequence of the al- 

 tered course of the fissure, above which the Cordillera of the Andes 

 has been upheaved. 



