250 
MESSRS. R, T. GLAZEBROOK AND J. M. DODDS ON THE 
Thus x and x are exceedingly small, and if we neglect squares and higher powers 
of SF, SG, x , and x', we obtain 
X P-Q 
(1 + a) —~—=x—x 
Now a ='036 mean B.A. unit. 
If then we know Q, the value of P can be found from the observations of x and x'. 
Two series of observations were taken, one by R. T. G. the other by J. M. D., each 
observation in each series being the mean of 2 or 3. 
The extreme difference between the two series was in no case more than 1 in 3000. 
The box remained in the room for some time before taking the observations, and 
the temperature was supposed to be that of the room as indicated by a thermometer 
laid on the box. A small correction was made for the resistance of the copper rods 
which connected the box to the bridge and the plugs in the box. This was deter¬ 
mined by one observer (R. T. G.) only, so that any error in it will affect both mea¬ 
surements equally. We shall show, however, shortly that it cannot affect the value 
of the B.A. unit as determined from our measurements. Having thus obtained the 
values of the resistances in the box in terms of the B.A. standard units, the value of 
R 0 in these units was determined by the ordinary method. We found thus the mean 
of several closely concordant measurements 
R 0 = 160-821 B.A. units 
the temperature being 14°*6 C. 
The resistance S of our secondary circuit and galvanometer was determined in terms 
of the coils in the same box, and we found 
S=3072-38 B.A. units 
at a temperature of 13°'2 C. 
Owing to the difficulty of determining the value of the resistance of the plugs in 
the box and the copper connecting pieces, either of these results may possibly have an 
error of about 1 in 1500. Now the value of S enters into the value of R 0 in absolute 
units and affects it in the same way. 
To determine the value of the B.A. unit we require the ratio 
Its value is 
Whence we get 
But the error we have been considering affects in exactly the same manner the 
numerator and denominator of this ratio. If in consequence of it one is too great so 
earth quadrant 
second 
R 0 in B.A. units 
158-626 
160-821 
ohms. 
1 B.A. unit =-98635 ohm 
