226 AVOYAGETO 



^>778- experienced here, as the habitations of the natives are quite 



February. 



V— >, — -> clofc ; and they fait both fifli and pork, which keep well, 

 contrary to what has ufually been oblerved to be the cafe, 

 when this operation is attempted in hot countries. Neither 

 did we find any dews of confequence, which may, in fome 

 meafure, be accounted for, by the lower part of the country 

 being deftitute of trees. 



The rock that forms the fides of the valley, and which 

 feems to be the fame with that fcen by us at diiTerent 

 parts of the coaft, is a greyifh black, ponderous flonc ; but 

 honey-combed, with fome very minute fliining particles, 

 and fome fpots of a rufty colour interfpcrfed. The lafl gives 

 it often a reddifli caft, when at a diftancc. It is of an im- 

 menfe depth, but fcems divided into Jlrata, though nothing 

 is interpofed. For the large pieces always broke off to a 

 determinate thicknefs, without appearing to have adhered 

 to thofe below them. Other flones are probably much 

 more various, than in the Southern iflands. For, during 

 our (liort ftay, befides the lapis lydius, which fcems com- 

 mon all over the South Sea, we found a fpccies of cream- 

 coloured whetftone, lomctimcs variegated with blacker or 

 whiter veins, as marble; or in pieces, as hrecciis ; and com- 

 mon writing flatc, as well as a coarfcr fort ; but we law none 

 of them in their natural Hate ; and the natives brought fome 

 pieces of a coarfc whiiifh pumice-flone. We got alfo a 

 brown fort of bdnjiotites^ which, from being ftrongly at- 

 tracted by the magnet, difcovcred the quantity of metal that 

 it contained, and fcems to belong to the fecond fpecies of 

 Cronftedt, though Linnxus has placed it amongft his hura^la- 

 h'd'ia. But its variety could not be difcovcred j for what we 

 faw of it, as well as the Hates and whetftoncs, was cut arti- 

 ficially. 3 



Befides 



