524 
Proceedings of the Royal Society 
merits were slightly inclined to each other, instead of being exactly 
parallel. If the axes of the instruments made with each other 
an angle of tan -1 1/6 \i.e ., tan -1 l*6 = 9° 28'], scarcely any differ- 
ence was perceptible in the readings from those taken with the axes 
parallel; a marked difference is visible if the angle be tan -1 1/3 
\i.e., tan -1 -3 = 18° 26']. 
Experiments repeated with a small galvanometer on the Helm- 
holtz model instead of the sensitive galvanometer, proved that no 
great error had been introduced by any accidental change of 
relative position between the coils and needle of the reflecting 
galvanometer, such as might have been produced by a slight tilt 
to the side. 
The dimensions, &c., of the coils of the Helmholtz galvanometer 
are as follows : — 
Each coil consists of 16 turns of wire, and is of rectangular 
section. The wire is wound on in 4 layers, each of 4 turns. The 
mean radius of both coils is 24*65 ± ‘02 cm., the distance between 
their centres is 24*51 cms. The following are the measurements 
for each layer : — 
Bed of coils, 
. 
24*18 cm. 
Outer side of 1st layer, . 
24*41 cm. 
5> 
2nd „ . 
24*66 cm. 
JJ 
3rd „ 
24*89 cm. 
>} 
4th „ 
25*12 cm. 
Explanation of the Tables. 
The tables are arranged as follows : — In the first column is a 
reference number to the experiment ; in the second column the 
position of the centre of the needle is recorded; y being the 
direction along the Helmholtz axis, and x the direction perpen- 
dicular to y in the horizontal plane; the numbers being the 
distances in cms. from the origin. The third column contains the 
values of the resistance £ of the reflecting galvanometer circuit ; 
the fourth the logarithms of these resistances ; the fifth the result- 
ing value of the intensity g of that point of the field, obtained 
by the formula g = 
£ 
where £' is the value of £ 
for the centre of 
