120) BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY. 
have been denuded. These beds appear to have been the same ~ 
Cretaceous and Tertiary strata which occur in patches along the coast 
in dislocated positions. The borders of the Coastal Cordillera are 
determined by faults so that the Cordillera is itself a long relatively 
uplifted fault-block or “horst.’”’ Relative to it, the western side of 
the Longitudinal Valley on the east has sunk as has also the Pacific 
Ocean border on the west. It is along the fault-zone on the west, 
partly if not altogether, that the epicenters of great Chilean earth- 
quakes mainly lie as was the case with that of Coquimbo in 1906, 
which destroyed Valparaiso and neighboring towns. 
It follows from this structure of the west coast of South America 
throughout the greater part of Chile that the earthquakes which take 
place along the faulted edges of the Coastal Cordillera arise from the 
movements of this horst and are not directly connected with the 
folded chain of the Andes on the east. Accepting as I think we must 
the essential accuracy of Darwin’s and Fitzroy’s observations upon the 
elevation of the coast about Concepcion at the time of the earthquake 
of 1835, it does not follow, however, as Darwin thought, that the 
Andes were simultaneously elevated. 
In the Valparaiso earthquake of 1906 Professor Steffen has found 
evidence of a local elevation amounting to half a meter on the western 
border of the Cordillera at that place; but it is a question whether the 
elevations which take place at the time of these earthquakes are 
permanent. At Concepcion it is the opinion that a subsidence follows 
the uplift of the coast. Nevertheless the horst stands as an indubita- 
ble block of evidence of uplift in relation to the ocean bottom and the 
Longitudinal Valley. 
The date of beginning of the faults which bound the horst of 
the Coastal Cordillera is Post-Eocene and apparently late Tertiary. 
Whether movement has taken place in modern times along the fault 
on the east I am not able to say. The Pleistocene gravels at San 
Rosendo abut against the horst without signs of disturbance; and the 
rivers, such as the Bio Bio, Calle Calle, and Maule, have cut channels 
across it on their way to the sea from a superposed position. Neither 
the Rio Bio Bio nor the Calle Calle, through whose valleys I passed in 
traversing the width of the horst, display rapids such as might be due 
to recent uplift at a greater rate than that of the cutting power of 
these streams to maintain a free grade to the sea. 
It seems most probable that the eastern fault bounding the Longi- 
tudinal Valley is of early Pleistocene or late Pliocene date, and that 
the present valley is due to erosion of the lavas and sediments part 
passu with the cutting of the gorges. 
