2G8 
VALUE OF THE BRITISH ASSOCIATION UNIT OF RESISTANCE. 
amount but in opposite directions. If SM be the error in M produced by an error Set 
Set 
in a the mean radius, it follows that in our case — is about 2 —, while in Lorenz’ 
M a 
method, M being the induction between the disc and the coils, for the arrangement 
adopted —2^ approximately—in the one case SM is of the same sign as Set, in 
the other the signs are opposite. The mean of the results then obtained from the two 
methods will be free from an error in a. But the two results are '98665 and '9867. 
They are identical to the fourth figure. Thus we infer that the error in a is probably 
very small. 
