Lov. 30, 1882] 
Hence, by Coulomb’s law of the proportionality of the 
force of torsion to the twist given, we have for the 
couple corresponding to a deflection @ the value 
6 
27 —a 
If then under the action of a current in the coil the deflec- 
tion of the needle is 9, the equation of equilibrium is 
Homi sin a. 
Cm2™1 cos 0= mH 1 (sin Oo+ ware sin a) 
ee 27 —a 
and therefore instead of (g) we have 
C=(1+- LEE 
27 — 20 IN 
tan 6. 
. (13) 
a sin 0 
6 sin 4) 
If a be an angle of say 1°, and @ be 45°, pill Poms very 
2 —a 
Sinaia) 0 I I 
nearly T and s— X or ——. Hence 
8 SUING enibi75 eel One AO OMe as 
OS (: ee) a 
The error therefore is somewhat less than 4 per cent. 
The accuracy of the measurements of currents, made 
according to the method of which I have just given the 
theory, of course altogether depends on the careful adjust- 
ment of the standard galvanometer, and the care and 
skill of the observer. The standard galvanometer should 
be of such a form that the values of its indications can be 
easily calculated from the dimensions and number of 
turns of wire in the coil. Such a galvanometer can be 
made by any one who can turn or can get turneda wooden, 
or, preferably, brass ring with a rectangular groove round 
its outer edge to receive the wire. It is indeed to be pre- 
ferred that the experimenter should at least perform the 
winding of the coil and the adjustments of the needle, &c., 
himself, to make sure that errors in counting the number 
of turns or in determining the length of the wire, or in 
placing the needle at the centre of the coil, are not made. 
The breadth and depth of this groove ought to be small 
in comparison with its radius, and each should not be 
greater than ;/5 of the mean radius of the coil. which 
should be at least 15 cms. The size of the wire with 
which the coil is to be wound must be conditioned 
of course by the purposes to which the instrument is 
to be applied, but it should be good well insulated 
copper wire of high conductivity, and not so thin as 
to run any risk of being injured by the strongest cur- 
rents likely to be sent through the instrument. For 
the exact graduation of current as well as potential gal- 
vanometers directly by means of the standard instrument, 
it is convenient sometimes to have two coils—one of 
comparatively high, the other of low resistance. The 
latter may very conveniently be a simple hoop of say 15 
cms. radius, made of copper strip 1 cm. broad and £ mm. 
thick. To form electrodes to which wires can be attached 
the ends of the strip are brought out side by side in the 
plane of the ring with a piece of thin vulcanite or paper 
between for insulator. Insulated wires are soldered to the 
ends of the circle thus arranged, and are twisted together 
for a sufficient distance to prevent any direct efiect on the 
needle from being produced by a current flowing in them. 
In constructing a coil the operator should first subject the 
wire to a considerable stretching force, and then carefully 
measure its electrical resistance and its length. He should 
then wind it on a moderately large bobbin, and again 
measure its resistance. If the second measurement differs 
materially from the first the wire is faulty and should be 
carefully examined. If no evident fault can be found on 
the removal of which the discrepance disappears, the wire 
must be laid aside and another substituted. When the 
two measurements are found to agree the wire may then 
be wound on the coil. For this purpose the ring may 
either be turned slowly round in a lathe or on a spindle 
NATURE 
107 
so as to draw off the wire from the bobbin, also mounted 
so as to be free to turn round. The wire must be laid on 
evenly in layers in the groove, and the winding ended 
with the completion of a layer. Great care must be taken 
to count accurately the number of turns laid on, The 
resistance should now be again tested, and if it agrees 
nearly with the former measurements the coil may be 
relied on. The ring carrying the coil thus made should 
now be fixed to a convenient stand in such a manner that 
if necessary it can be easily removed. The stand should 
be fitted with levelling screws so that the plane of the coil 
may be made accurately vertical. A shallow horizontal 
box witha glass cover and mirror bottom should be carried 
by the stand at the level of its centre. Within this the 
needle and attached index are to be suspended. The 
needle should be a single small magnet about a cen- 
timetre long, hung by a single fibre of unspun silk 
about 10 cms. long from the top of a tube fixed to the 
cover of the shallow box, so that the centre of the needle 
when the coil is vertical is exactly at the centre of the 
eoo0000000 
es000000 
oo 
oso 
oo 
Q°0 
oo 
oo 
90 
oo 
Fic. 3. 
coil. To allow of the exact adjustment of the height of 
the needle, the fibre should be attached to the lower end 
of a small screw spindle, made so as to be raised or 
lowered, without being turned round, by a nut working 
round it above the cap of the tube. The needle should 
carry a thin glass index, about 6 inches long, made by 
drawing out a bit of thin glass tube at the blowpipe. In 
order that the zero position of the index may not be 
under the coil, the index should be fixed horizontally 
with its length at right angles to the needle, so as to 
project to an equal distance in both sides of it. To 
test that this adjustment is accurately made, draw a 
couple of lines accurately at right angles to one another 
on a sheet of paper. Then suspend a long thin straight 
magnet over the paper, and bring one of the lines into 
accurate parallelism with it. Remove then the magnet, 
and put in its place the little needle and attached index. 
If the index is parallel to the other line the adjustment 
has been carefully made. The needle may then be sus- 
pended in position and the box within which it hangs 
closed to prevent disturbance from currents of air. 
