148 MR. A. NORMAN SHAW: A DETERMINATION OF THE ELECTROMOTIVE 
Page 
(d) The time measurements.176 
(e) The masses of the cylinders.177 
(/) Further remarks on the accuracy of the method.178 
(, g ) The value of B during the deflection observations.179 
VII. The deflection observations. 181 
(a) Method of observation.181 
(b) The correction for torsional variations in the controlling couple.181 
(c) The electrical connections.183 
(i d ) The potentiometer correction.184 
(< e ) The deflection tables.185 
(/) The evaluation of corrections.189 
(g) The final value of S, the double scale deflection. .191 
( h ) The determinations of /?, the distance between the scales . . ..191 
(i) The calculation of the mean tan 9 .192 
VIII. The resistances.193 
IX. The “ ratio of the currents ” during the deflection observations. ... 194 
X. The standard cells.195 
XI. Insulation and magnetic tests.195 
XII. The electromotive force of the international mean Weston normal cell.196 
Appendix. Note on the calibration of the Kelvin-Varley slide for accurate electrical measure¬ 
ments.197 
Preface by Prof. H. L. Callendar. 
The measurement of the E.M.F. of the Weston cell affords the best means of comparing 
the performances of different methods and instruments for the absolute determination 
of the ampere. Great progress has been made in the last six years, but the most 
recent determinations by independent methods, giving equal promise of accuracy, still 
show discrepancies covering a range of 2 parts in 10,000, which must be debited for 
the most part to the difficulty of the absolute determination of current. Each 
method in itself appears to give an order of accuracy of repetition approaching, or 
even exceeding, 1 in 100,000. It is therefore of special interest and importance to 
compare the results of methods differing as widely as possible in experimental details 
in endeavouring to arrive at a value comparatively free from the constant errors 
which may beset any particular type of method. 
The measurements described by Mr. Shaw in the following paper were made by 
the method of the Weber bifilar electrodynamometer, as modified by Clerk Maxwell 
and Latimer Clark, which has not hitherto been employed for work of the highest 
accuracy, and which merits attention on account of its many fundamental points of 
difference from recent methods. The instrument originally supplied to McGill 
College for this purpose was a faithful copy of Clerk Maxwell’s instrument at 
Cambridge, of which the theory is given together with a figure and description in his 
‘ Electricity and Magnetism,’ vol, 2, p. 367. The chief sources of error in this instru¬ 
ment were (1) the uncertainty of insulation of the coils, which proved to be of the 
