I 



l°+ 



ATMO- 

 SPHERE. 



\ 



REMARKS 



N 



THE 



by a copious dew. 



I am therefore inclined to believe, that a heavy 



L 



fall of dew, cannot any longer be confidered as a certain or infallible 

 fign of the neighbourhood of lands. 



Secondly y The Tropical ifles having been vifited by us at two dif- 

 ferent times, vi%. once from Auguft to Ocftober, in 1773, and again 

 from March to September, 1 774, we can hardly from thence form 

 an idea of the changes and the returns of the feafons, as, each time, 



V 



during our whole ftay, the fun was in the Northern hemifphei-e : 



h 



however, if we confider that we were at O-Taheitee at the latter 

 end of Auguft, and again in April and the beginning of May, and 

 that we found a very material difference in the external appearance 

 of vegetation, we may be allowed to conclude, that this was owing 

 to the difference of feafons ; for, in countries between the tro- 

 pics. Nature feems to be more uniform and conflant in the fetting 



of winds and rains, than without the trop 



The difPeren 



was too flriking not to be noticed -, and, if it were not caufed by 



the natural 



of the feafons, it mufl have been produced by 



F 



very extraordinary deviation from the uniform tenor of the changes 

 produced within the tropics 3 which may be poiTible, but hardly 



probable. 



The hills in O-Taheitee were covered in Augufl with dry and 



^ 



dead herbage, a great part of it had been burnt by the natives. 



and 



ave 



the country a barren and drearj look. The iHe had very 



\ 



I 



little 



