THE PACIFIC OCEAN. 187 



Jatiuarjr. 



forts of rock-fifh ; one with numerous foots of blue, and ,^77^ 



* I an on r 



the other with whitifliftreaks fcattered about. 



The foil of this ifland, in fome places, is light and black, 

 evidently compofed of decayed vegetables, the dung of birds, 

 and fand. There are other places again, where nothing 

 but marine produdtions, fuch as broken coral ftones, and 

 fhells, are to be feen. Thefe are depofited in long, nar- 

 row ridges, lying in a parallel direcftion with the fea-coaft, 

 not unlike a ploughed field -, and muft have been thrown 

 up by the waves, though, at this time, they do not reach 

 within a mile of fome of thefe places. This feems to 

 furnifli an inconteflable proof, that the ifland has been pro- 

 duced by acceflions from the fea, and is in a (late of in- 

 creafe ; for not only the broken pieces of coral, but many 

 of the fliells, are too heavy and large to have been brought 

 by any birds, from the beach, to the places where they now 

 lie. Not a drop of frefli water was any where found, though 

 frequently dug for. We met with feveral ponds of fait 

 water, which had no vifible communication with the fea, 

 and muft, therefore, in all probability, be filled by the water 

 filtrating through the fand, in high tides. One of the loft 

 men found fome fait on the South Eaft part of the ifland. 

 But, though this was an article of which we were in want, 

 a man who could lofe himfelf, as he did, and not know 

 whether he was travelling Eaft, Weft, North, or South, was 

 not to be depended upon as a fit guide to conduct us to the 

 place. 



There were not the fmalleft traces of any human being 

 having ever been here before us ; and, indeed, Ihould any 

 one be fo unfortunate as to be accidentally driven upon the 

 ifland, or left there, it is hard to fay, that he could be able 



B b 2 to 



