176 MR. A. NORMAN SHAW: A DETERMINATION OF THE ELECTROMOTIVE 
By maintaining strong pressure it was possible to reduce these values to about 
14'4974 as obtained at the outer position. This difference is due to the small distance 
apart at which the cylinders remained under ordinary circumstances. These observa¬ 
tions show that this distance is approximately 0'0005 cm. Direct measurement on 
the distance between the edges, when no pressure was exerted to keep the cylinders 
together, gave from 0’0003 to 0'0006 cm. 
It was advisable to check the apparent possibility of the influence of these small 
variations by some observations on T\ and T 2 with the cylinders in the different 
positions. 
Table XV. 
Position of the cylinders. 
Tl 
Cylinders in initial external position 
11-6566 
Reversed in direction 
11-6563 
Rotated 90 degrees 
11-6564 
„ 180- „ 
11-6564 
„ 270 „ 
11-6565 
T 
Cylinders at the initial internal position 
±2* 
6-5621 
Rotated through 90 degrees 
6-5620 
„ „ 180 ,. 
6-5622 
n » 270 ,, 
6-5619 
Reversed in direction 
6-5620 
(d) The Time Measurements. 
The observations for T were obtained by recording the transits on an electric 
chronograph controlled from the McGill Observatory. It was necessary to take the 
time of a large number of swings in order to get the required accuracy. When the 
period T is determined in this way it is necessary to apply a correction to the 
calculated value in order to reduce it to the value T 0 , for infinitely small oscillations. 
In the experiment we really obtain the average period T 7 . Now if the amplitude “a” 
is very small, we have, by a well-known calculation, 
T = T »(l + h)’ - 
and hence 
1 ~ H 1+ -82N- 7 
where a T is the amplitude fie., the arc described in T for the r tb transit, and 2N is 
the number of transits considered. 
