

178 



Mr J. D. FORBES on the Horary Oscillations 



Lat. 



Oscill. 



Selecting nine of the possible combinations by threes, we ob- 

 tain the following values for n : 



Lat. 



Lat. 



Lat. 



9 2 



21 22 

 46 4 



9 2 

 38 3 

 53 14 



9 2 

 33 51 



73 48 



}" 

 H 



} 



-~ 2 /2 



Mean value of n, = 2.60, or about -. 



15. In order to modify the co-efficient and constant in the 

 arbitrary formula, so as best to agree with observation, we shall 

 select a series of observations from M. BOUVARD'S Table, which, 

 from their intrinsic merit, or from their representing accurately 

 a group of minor observations, shall seem best fitted to give ac- 

 curate results. We shall first take an approximate value of 

 a and from the observations at Cumana and Toulouse, which 

 represent pretty nearly the whole of the rest. From them we 

 obtain the two following equations : 



2.55 = cos* 10 28' + = .959 + 6, 



Whence = 3.02 ; = 0.35. 



Substituting these values in the expressions for the selected 

 oscillations, we obtain the following approximations : 



