581 
of Edinburgh , Session 1868 - 69 . 
cated in the sketch, is permanently established between the two 
central holes and the holes connected with the electrometer. 
The rocker consists of four wires, supported on an insulating bar 
of vulcanite, the two outermost having three points, the middle 
one longer than the others, and the two inner being similar, but 
wanting one of the extremities. When the four middle stems dip 
vertically into the four central mercury cups, the other stems do 
not reach the mercury in any of the other six cups. If the instru- 
ment be inclined to the right the four prongs enter the holes to 
the right — thus simultaneously connecting the electrodes with the 
decomposing battery, and the electrometer with the charging 
battery. When the instrument inclines to the left, the electrodes 
are shunted from the decomposing battery on to the electrometer, — 
the latter having just before, by the same action, been cut off from 
the charging batter}', and thus left charged. 
The modus operandi is simply this Leave the rocker leaning to 
the right by its own gravity, decomposition and polarisation going 
on ; adjust the wires B 2 to different points in a wet string (or a 
narrow canal of water) closing the circuit of the charging battery ; 
work the rocker quickly to the left, and allow it instantly to fall 
back again, — a process which need not occupy more than a small 
fraction of a second ; yet which must not be performed too quickly, 
on account of the inertia (small as it is) of the needle and mirror 
of the electrometer. If the deflection of the electrometer be suddenly 
increased or diminished by this action, slide one of the wires If, 
along the wet string, a little farther from or nearer to the other, 
and rock again, — continuing this process till a charge is found which 
leaves the electrometer at rest when the rocking to and fro is per- 
formed. Beverse a commutator attached to the wires E, and repeat 
the operation. The difference of the scale readings in these two 
cases gives a number proportional to the electromotive force of the 
polarised plates — (I say difference , because the scales commonly 
used with Sir W. Thomson’s instruments are, to avoid confusion, 
graduated from one end to the other, as they ought to be, instead 
of being graduated opposite ways from the middle). To enable 
this measure to be reduced to absolute units, a normal Daniell’s 
cell was applied at intervals, during each day’s work, directly to 
the electrodes of the electrometer, then reversed ; and the differ- 
