148 GEOLOGY OF THE WEST INDIES. [Nov 



five to six hundred yards thick. Part of this madrepore 

 rock is changing into silex, havinqj the pan that surround- 

 ed the animal already converted into chalcedony. A con^ 

 siderable quantity of gypsum is found near the same place, 

 in a crystalline state. 



Saba, This litde island seems to finish the volcanic 

 formation, and consists of one mountain, rather rougher 

 and more rugged than St, Eustatia, but apparently of 

 nearly the same kind of rocks. 



The foregoing description of the volcanic islands may 

 perhaps authorize the following general remarks. 



1st. That there is a great similarity in the substances 

 ejected, which are marked by a family feature running 

 through all the rocks, cinders, Sec. of the different islands; 

 and it is to be observed that the proportion of cinders, 

 pumice, and other light substances, is much greater than 

 of the solid lavas, which are but thinly scattered; also that 

 the cinders are always the lowest stratum on a level with the 

 sea; and the masses of solid lava, near that level, repose on 

 a bed of cinders, in every place where 1 had access to them. 



2d. The madrepore and coral rocks, mixed with shells, 

 partly similar to those found at present in the sea, are 

 found in many places alternating with the cinders, and 

 other volcanic rocks, presenting much the appearance of the 

 whole having been ejected from the bottom of the ocean. 



3d. The direction of the islands, running from north to 

 south, a little easterly, corresponds with the direction oT 

 the strata of those stratified islands, lying to the eastward: 

 such as Barbadoes, St. Bartholomew, Sec. which should 

 seem to support the supposition, that the seat of combus- 

 tion occupies a stratified substance, running parallel to the 

 general stratification of the surrounding rocks. 



