of Edinburgh, Session 1885-86. 
523 
5. Experiments on the Field of a Helmholtz Tangent Gal- 
vanometer. By E. Ellis, C.E. ; E. D. Clarkson ; and 
H. Eainy, M.A. (Plate XVIII.) 
The object of the following experiments, which were suggested 
by Professor Chrystal, was to determine the relative values of the 
intensity of the horizontal component of magnetic force, due to a 
current flowing in the coils of the Helmholtz galvanometer, for dif- 
ferent parts of its field. 
The experiments were conducted as follows : — A sheet of ground 
glass was carefully ruled into squares, whose sides were 1 cm. long, 
and it was placed horizontally within the coils of the Helmholtz 
galvanometer. On it stood a sensitive reflecting galvanometer, the 
horizontal plane of the centre of the mirror containing, as nearly as 
possible, the axis of the coils of the Helmholtz galvanometer. The 
position of the sensitive galvanometer with reference to the centre 
of the Helmholtz field could be readily obtained by means of the 
graduated glass plate. 
For the first experiment the sensitive galvanometer was placed in 
the centre of the field, when a current was split and the two parts 
sent in opposite directions through the coils of the sensitive gal- 
vanometer and of the Helmholtz respectively ; resistances being 
added to the former circuit until no deflection of the needle took 
place. The resistance of the whole circuit of the sensitive gal- 
vanometer was then determined ; after which the experiment was 
repeated to guard against any accidental error. The sensitive gal- 
vanometer was then removed to the next point of the field which 
was to be tested, and the same process repeated ; when the in- 
tensity at that point (that at the centre being taken as unity) 
was obtained by dividing the former by the latter resistance. 
The error in placing the sensitive galvanometer was within 
± 1 mm. at any part of the field ; the error in the graduation of the 
glass plate was less than 1 per cent. A very slight alteration of 
the resistance inserted in the circuit — less than ’0001 of the total — 
produced a very considerable deflection. 
No perceptible error was introduced if the axes of the two instru- 
