IS 



change of temperature ou the quantity of refraction is taken 

 too great. This appears certain by the direct experiments of 

 T.Mayer, Dalton and Gai-Lussac on the expansion of air 

 at different temperatures. It also appeared evident to nie 

 by observations of low stars in different temj^eratures. The 

 consequence of which is, that even supposing the utmost 

 accuracy in the instruments and in the observations, the 

 zenith distances of stars will appear greater in winter than in 

 summer, and the more so the greater the zenith distance. 



The column P shews the quantity to be applied to the 

 fourth column to obtain the north polar distances according 

 to M. Piazzi, at Palermo. His north polar distances given 

 in the Conn des Temps, 1812, having been reduced to Jan. 

 1, 1813, and also reduced to what they would have been ac- 

 cording to Bradley's refractions. I do not know the exact 

 date of these observations, but I suppose them recent. I 

 believe also that M. Piazzi takes the mean refraction at 

 45' = 57",4 and makes the same allowance for changes in 

 the thermometer as Dr. Bradley. If so, the correction to 

 be applied to the north polar distances, as determined by 

 M. Piazzi, to give what would have resulted from the use of 

 Bradley's refractions = — 0",69— 0",5 (tan. N,P.D,—55\ 53% 

 This quantity has been applied accordingly. 



The column D is the difference between my results in 

 1809 and 1813. The quantities according to their signs are 

 to be applied to the results in column 4, to give what would 

 have resulted from the observations in 1809- In makiug 



VOL. XII. M 



