98 THE ROYAL SOCIETY OF CANADA 



interchanging R1 + R2 and V with regard to the rest of the circuit we 

 get, in the same way as we obtained equations (4), (5) and (o), the 

 three equations 



P R *+ R > + Kd 



Q v 



P+ 1 R x + R t 



-q~ = ~Y~ + Kd 



P V 



+ Kd" 



(21) 



Q Ri+Rs 

 which gives us ( since we may take -~ = j 



V 1+ 4000 (d-d') { } 



The deflections were always obtained by reversing the current 

 through the galvanometer. In one case the deflections, expressed in 

 scale divisions, were • 



d=+ 139, d'=-l29 and d" = +26, 



This gives — =z — =1.00011. The following is a list of values ob- 

 tained for this ratio, before, during and after our calibration. 



It will be seen that such variations from the mean value 1.00011 

 can be neglected in the calculation of the complete corrections. 



r> 



The determination of -=■ is made in a similar manner, but in this 

 R\ 



case two one-thousand ohm coils are used for the auxiliary arms and 

 it is necessary to make one of the galvanometer contacts by means of 

 a copper wire to the stud between coil 1 and coil 2, the glass cover 

 of the main slide being removed for the purpose of this test. Per- 

 forming the same operations as before we get the three analogous 

 equations reducing to 



*• - 1 + d ~ d " (23) 



~K~ 1+ 2000 (d-d') {26) 



