of the Movable-coil Galvanometer. 81 
is evident, if it is remembered that the weak current of one 
thousandth of an ampere flows round the needle one thousand 
times, but the stronger current of one ampere only once. The 
solid ring and the convolutions are thus arranged that their 
cross sections have a common centre of gravity, thus both 
acting exactly in the same way upon the magnetic needle. If 
this simple plan is adopted, the calibration of the galvano- 
meter for difference of potential in volts, which is readily per- 
formed with a few cells of known E.M.F., at the same time 
gives the graduation of the instrument for strength of current 
in amperes. I have been using various modifications of the 
Daniell cell with solutions of copper and zinc sulphate 
of equal specific gravity. At present I am engaged in con- 
structing a standard cell for such purposes, which is always 
at disposal; and, as far as the preliminary experiments show, 
the E.M.F. of the new cell will closely approach one volt. 
Further on I shall communicate some measurements, by which 
I intend to show how accurately the calibration of the fine-wire 
coil in volts can serve for that of the solid ring in amperes. 
The Constant Shunt and Leading Wires. 
With the size of the ring usually employed, viz. 30 centim. 
diameter, and our horizontal component of the earth's magne- 
tism, currents of greater strength than about 50 amperes 
would require the ring to be at the multiplying powers 9 or 
10, i. e. near the horizontal position. If, as a rule, such cur- 
rents have to be measured, it is desirable to raise the con- 
stant of the galvanometer, say two or threefold, without a 
proportionate increase of the dimensions. This can be done 
by the use of a so-called " constant shunt," thereby allowing 
only half or one third of the current to flow round the needle. 
In our case the shunt is made of exactly the same metal as 
the solid ring itself; it has no soldering-places, and consists 
in fact of three or four little bridge pieces left standing 
instead of cutting the ring quite open where the terminals join. 
By comparison with an instrument of the same description 
having an open ring, the shunt can be adjusted to any power 
desirable. 
However, by the introduction of the " constant shunt " the 
accuracy of the measurements is somewhat impaired. Expe- 
riments in which the shunt-pieces were touched with a thin 
stick of low-melting material during the passage of very strong 
currents, proved that they did not become hot, on account of 
their extremely low absolute resistance and their contact with 
the large metallic mass of the ring conducting away the heat. 
Variations of temperature, to which both the ring and the 
