358 REMARKS AND OPINIONS. 



do not dash and break, as against the u])per ledge 

 of a wall of rock. * It is the same stone, in which, 

 on the coast of Guadaloupe, petrefied human 

 skeletons are found inclosed. We have seen the 

 celebrated specimen in the British Museum, and 

 had an opportunity of accurately comparing the 

 stone in the Berlin mineralogical museum, t 



These coral reefs, low island-groups, and islands, 

 are extremely frequent in the Great Ocean, between 

 the tropics, and especially between the limits 

 assigned above to the two island provinces. Some- 

 times we found them singly ; sometimes in rows, 

 which seem to indicate a ridge in the bottom ol* 

 the sea ; sometimes in the vicinity of the high 



* We observed this with particular accuracy in Woahoo, be- 

 tween Hana-rura and Pearl River, where we passed in one oF 

 the boats of the natives, along the reefs, and several times 

 backwards and forwards through the surf. On the outside of it, 

 some boats were engaged in fishing in three or four fathoms' 

 water. 



f We met with this reef-stone, in the year 1817, atOwhyee, 

 at the foot of the lava, which flowed in 1801 from the moun- 

 tain of Wororai, and where there is no proper reef. Here it 

 contains fragments of lava ; otherwise it is identical with that 

 collected on the low islands. The stone of Guadaloupe is 

 precisely one and the same with the fine-grained varieties 

 of it. We found this reef-stone, and in some places reefs, at 

 Manilla and Guahon. With respect to the varieties, composed 

 of larger fragments, some local differences may arise, from the 

 difference of the species of madrepores, of which they are 

 chiefly composed. We are ef opinion that the species which 

 live on the spot, furnish the elements for the stone which is 

 formed. 



