/•; 



( /'IKI)UI<> ( 



In tt, .X 



o/'on 



ha. 



25 1 



and so close together that it is almost impossible for one to- 

 walk over them (see fig. S). On the southeastern side of [lha 



Raza the Bame etching of these calcareous sandstones has boon 

 caused by the surf blowing over the island (see fig. 6). 





.Kolian sandstone capping eruptive rocks on llha Rapta. 



The southern side of llha Raza has been deeply undercut by 

 the water. The rock here is more compact than at most of its. 

 exposures. Great caverns are at first worn along the line at 

 which the force of the surf is greatest at high tide. These ex- 

 cavations are carried forward upon a flat floor which extends all 

 along the southern side of llha Raza (fig. 6). When the under- 

 mining has reached a certain point the roof or edge of the 

 islands breaks and falls upon this floor whence it is soon re- 

 moved by the surf. It is to be remarked that all the debris 

 coming from the destruction of the sandstone at this place is 

 carried away by the undertow leaving the stone floor clean, 

 as shown in the accompany illustration. I am inclined to 

 think, however, that the removal of these rocks is due to a 

 considerable extent to chemical action. That the salt water 

 does dissolve the rock is clearly shown by the etching produced 

 by the spray wherever it is blown over these sandstones. 

 Undermining by the ocean is seen wherever these seolian sand- 

 stones occur — on llha Rapta, llha do Meio and llha Raza, on 

 the Chapeo and elsewhere, but the stone floor left by the 

 removal of the rock is most extensive on llha Ra r /.a. 



A shelf similar to that above described usually answers the 

 purpose of a landing place on llha Rapta. There is no beach 

 on that island, and a boat or jangada can be brought beneath 

 one of these projections at low water, enabling one to disem- 

 bark by climbing up on the ledge. 



