108 



SIXTH REPORT — 183G. 



Here in the formula, « = ^ ^, T' = 645^-89 ; - T'T 



1 [r — T) 



_. 0^-6 ; T — t' = .38°-7. Whence a = -000048. 



In this experiment the time of vibration, as may be seen, 

 varied considerably in the cold temperatures at the commence- 

 ment and at the close, and gave reason to believe that the 

 change due to temperature might be overpowered by the 

 diurnal variation of the force. The experiment was therefore 

 repeated on the 15th November, as follows : 



Here T = 661^ ; T - T = 0^-76 ; t - t = 37°-6. Whence « = 

 '000061. This experiment appears more satisfactory than the 

 preceding one ; but as the results are so nearly the same, I have 

 taken the arithmetical mean -000055 for the value of a, which 

 being multiplied by M, the modulus of the common system of 

 logarithms, = -000024 the coefficient of t — x' in the correction 

 for temperature. 



In Table IV. the two last columns contain the value of the 

 intensity computed from the angles of deflection, and from the 



