■ : n'^.^IL'T! OF THE GLOBE. 207 



76°.45', and 78°, from a series of 650 observations made by 

 himself, in nine successive years *. 



In the latitude of 76° 45' he found the mean temperature to 

 be 18° iVsdths.. My formula makes, it 18° iVodths, deviating on- 

 ly iVsdths of a degree. 



In the latitude of 78°, Mr Scoresby found the mean tempe- 

 rature to be 16 Voodths. My formula makes it I6°iVndths, devia- 

 ting only iBsdths of a degree. Mayer's formula makes the 

 mean temperature of these parallels above 34°, erring no less 

 than 33° upon both observations, whereas the error of my for- 

 mula is only > th of a degree. 



It appears, then, from the evidence of direct observation, 

 that the temperature of the North Pole must be considerably 

 lower than 16° iVsdths, and must therefore be more correctly 

 indicated by the new formula than by that of Mayer. Mr 

 Scoresby has attempted, by a very ingenious analogical pro- 

 cess, to deduce the temperature of the Pole from that of 

 76° 45'. He considers the difference between the actual tem- 

 perature of that parallel, viz. 18°. 86, and the temperature gi- 

 ven by Mayer's formula, or 33°. 8', as an anomaly produced by 

 the frigorific influence of the ice ; and having found what this 

 anomaly should be at the Pole, he subtracts it from Mayer's 

 polar temperature, in order to obtain the real polar tempera- 

 ture, 'which he thus finds to be 10°. This result, however, is 

 obviously too great, upon Mr Scoresby's own principle ; for 

 since Mayer's formula errs greatly in excess in those parallels 

 where there is no accumulated ice to produce an anomaly, it 

 must give at least an equal error in excess for the parallel of 

 76°.45'. Now, this error in the latitude of 63° and 65° in Lap- 

 land 



• This interesting investigation is now published, in Mr Scoresby's excellent 

 Account of the Arctic Regicms, vol. i. p. &56. 



