CONTINUOUS ELECTRIC CALORIMETRY. 
81 
(C.) Current. 
(10.) The Electrodynamometer. 
For the absolute measurement of the E.M.F. of the Clark cells employed in this 
investigation in terms of the standard ohm, I proposed to employ an electro¬ 
dynamometer of the pattern constructed by Latimer Clark for the Electrical 
Committee of the British Association. This instrument is described and figured in 
Maxwell’s “ Electricity and Magnetism,” vol. 2, p. 339. The moment of the 
couple due to the current is measured by means of a bifilar suspension, the constant 
of which is determined by observations of the time of oscillation. It has been 
usual of recent years to prefer the current-weighing method, in which the force due 
to the current is directly balanced by weights, to methods depending on the 
observation of an angular deflection; but it appeared to me that the British 
Association apparatus, with certain modifications of detail, would escape most of the 
errors generally urged against the deflection method, and that it possessed important 
advantages in other respects. 
The electro-dynamometer in the possession of McGill College, as originally 
constructed by Messrs. Nalder, was intended to be an exact duplicate of the 
apparatus in the Cavendish Laboratory at Cambridge. The coils of the Cambridge 
apparatus were wound under the supervision of Maxwell, whose measurements of 
the large coils were assumed by Lord Rayleigh in his determination of the 
electrochemical equivalent of silver. As a preliminary step, the apparatus was set 
up and tested in October, 1894, with the assistance of Mr. R. 0. King, then a 
fourth-year student. The defects of the electrodynamometer, in its original form, 
were found to be so serious that it was necessary to re-wind the coils, and make 
other extensive alterations. This involved considerable delay, while the necessary 
materials were being obtained and preparations made; and the apparatus was 
not set up in its final form until the return of Mr. King from Harvard in 
September, 1897. 
The principal defects of the original apparatus were found to be as follows :— 
(1) The framework of the large coils, and the pulley arrangement for equalising the 
tensions of the bifilar suspension were not sufficiently rigid. It was not possible 
to obtain the desired order of accuracy in the observation of the times of oscillation, 
or in the deflections. (2) The most serious defect w T as the uncertainty of the 
insulation between adjacent layers of the windings of the coils. This could not be 
directly tested, but was inferred from a comparison of the mean radii of the coils 
determined by electrical methods with the values calculated from the measurements. 
The difference could only be attributed to defective insulation, as the number of 
turns and the measurements were verified by unwinding the coils. The error 
amounted to nearly 1 part in 500 for the large coils (324 turns each), and about 
YOL CXCIX.— A. 
M 
