ON THE ABSOLUTE UNIT OF ELECTRICAL KESISTANCE 163 



I. Torsion of fibre. 



The needle of the tangent galvanometer was sustained on a point 

 and so required no correction. The correction for the torsion in the 

 other galvanometer is the same for /? and d and hence only affects T. 

 Therefore, if t is the coefficient of torsion, 



b= - It. 



c. Rate of chronometer. 



Let p be the number of seconds gained in a day above the normal 

 time 



P 

 ~ 86400* 



d. Reduction to normal metre. The portion of this reduction which 

 depends on temperature must be treated under the variable corrections. 

 Let m be the excess of the metre used above the normal metre, ex- 

 pressed in metres; then 



d = + m. 



e. Correction of T for the arc of vibration. This arc was always the 

 same, starting at c^ and being reduced by damping to about c n , 



where c^ and c a are the total arcs of oscillation. 



/. Correction for length of needles. For the tangent galvanometer, 

 the correction is variable. For the circle it is 



/= + 



where I is half the distance between the poles of the needle and A the 

 radius of circle. For the other galvanometer it is included in the 

 formula for G. 



A. Reduction to normal metre. As the dimension of R is a velocity 

 and the induction coils were wound on brass, the correction is 



where f is the coefficient of expansion of brass or copper, t' the actual 

 and t" the normal temperature. 



B. Correction of standard resistance for temperature. Let a be the 

 variation of the resistance for 1 C., ?" be the actual and T the normal 

 temperature 17- C. ; then 



