156 MR. A. NORMAN SHAW: A DETERMINATION OF THE ELECTROMOTIVE 
There are nineteen double turns so that the mean radius 
2972-60 + 2972-92 
2 x 19 x2tt 
= 24-9016 cm. 
The mean radius of both coils is therefore 2+8892 cm. 
(A rough verification of these measurements was made by measuring the channel 
before and after with a steel tape.)” 
The measurements were taken at an average temperature of about 17° C., hence 
correcting for the temperature change in the coil frames and tape we have the mean 
radius 
a = 24-8905 cm. at 20° C. 
The mean depth of the winding was approximately +72 cm. 
It will be noted that the determination of this radius was made by Mr. King 
some time before the present work. The constancy of the dimensions of the coils 
could, it was thought, be assumed without chance of error, as the coils had not been 
used at all in the interval and the insulation (silk ribbon in the case of the fixed coils) 
was not of such a nature that strains could be caused in the winding by temperature 
or humidity variations in the laboratory. A check on the validity of this assumption 
can, however, be found in § IV. (c), p. 167. It will be seen from the value of IJf 
that Mr. King’s value for the ratio of the dimensional constants for the two systems 
of coils differed from the present one by less than 3 parts in 100,000. The 
construction and the arrangement of insulation in the two systems was so different 
that this constancy in ratio can be taken as evidence of their individual constancy. 
The system of interchanges adopted in the series of observations recorded later 
made it possible to use the mean value of the radii without the introduction of any 
further correction on account of their slight difference. 
(b) The Determination of the Distance between the Large Coils. 
The distance between the planes of the large coils, 2d, was determined by measuring 
the distance between the external edges at five evenly distributed positions round 
the coils, and subtracting the mean thickness of the channel-frames at these points 
from the mean of these measurements. The reversing of the coils during a series of 
observations eliminated errors due to any asymmetry in the position of the copper 
tape in its channel. An extension micrometer gauge of special size and shape for 
adjustment on the dynamometer was constructed by attaching a sensitive screw 
gauge to one end of a thick brass tube and a suitably shaped end-piece and stud at 
the other end. The brass part was wrapped in asbestos and a thermometer inserted 
in the tube. It was found that the variations due to the temperature changes while 
handling could be followed with sufficient closeness to avoid any possibility of error. 
These calipers were calibrated" by means of rectangular blocks with sharp edges 
