46 JOURNAL. 



July, to the ftate of the thermometer, which I was furprifed to 



find differ fo little about noon and midnight ; its greateft 

 height was 58"^, at eleven in the forenoon; at midnight 

 it was 51°. 



On the 1 6th, at noon, the weather was remarkably 

 fine and clear. The thermometer in the fhade being at 

 49°, when expofed to the fun rofe in a few minutes to 

 Sg''!, and remained fo for fome time, till a fmall breeze 

 Springing up, made it fall 10° almoft inftantly. The 

 weather at this time was rather hot ; fo that I imagine, if 

 a thermometer was to be graduated according to the feel- 

 ings of people in thefe latitudes, the point of temperature 

 would be about the 44th degree of Fahrenheit's fcale. 

 From this ifland I took a furvey, to afcertain the fituation 

 of all the points and openings, and the height of the moll 

 remarkable mountains : the longeft bafe the ifland would 

 afford was only 618 feet, which I determined by a crofs 

 bafe, as well as aSual meafurement, and found the refults 

 not to differ above three feet. To try how far the accu- 

 racy of this furvey might be depended upon, I took in a 

 boat, with a fmall Hadley's fextant, the angles between 

 feven objedls, which interfefted exadly when laid down 

 upon the plan. I had a farther proof of its accuracy 

 fome days after, by taking the bearings of Vogel Sang 

 and Hacluyt's Head Land in one^ which correfponded 

 ■exadly with their pofition on my chart. 



On 



