of Edinburgh, Session 1870 - 71 . 
495 
ternal currents. If, in the circuit of the electric coil, a few inches 
of fine platinum wire he included, and the circuit of the magnetic 
coil half completed, so that one end of the connecting wire has only 
to touch the other binding screw to close it, and the handle be put 
in sufficient motion, the platinum wire becomes white hot, and this 
sinks to a dull red when contact in the magnetic circuit is made. 
The same takes place when the coils are reversed. Such an action 
as this suggests the supposition that what appears in the second 
coil is but electricity stolen from the first, and that the arrangement 
effects only a convenient distribution, and not an increase of the 
electricity available. I cannot, with the observations I have yet 
made, say that such is not true in all cases, but in one case, at least, 
the only one I have examined, such a supposition cannot be enter- 
tained, and that is when both coils work together in the same 
circuit. When both coils, as just mentioned, are ready to give 
external currents under the magnetism induced by one Bunsen cell, 
it is quite possible, by accustoming the ear to the note produced by 
the springs rubbing on the revolving collars, to get the arm to work 
at a uniform speed. If the cell be steady, you can, within a frac- 
tion of a degree, produce the same angle in the galvanometer in 
the same circumstances. I have made repeated observations in 
this way as to what current the electric coil would give when act- 
ing alone, as to what the magnetic coil would give, and as to what 
both together would effect. The circuits in these cases consisted of 
the coils themselves and the wires leading to a tangent galvano- 
meter some 12 feet off, and the working of the machine and the 
observing of angles were done by different persons. The resistances 
in both circuits were sensibly the same. The resistance of the 
electric coil was 32 inches of a G-erman silver wire in my posses- 
sion, that of the magnetic coil 34, and that of the galvanometer 
wire 5 inches. To these must be added the resistance introduced 
by the imperfect contact of the break-springs, which, at a high 
speed, and especially in the case of the machine exhibited where 
the armature is not quite truly centred, must be considerable. The 
difference between the two coils would thus almost disappear on the 
total resistances of their respective circuits. This being the case, 
the work value of the electricity appearing in each will be as the 
squares of the tangents of the angles observed. Now, in all the 
