42 
the coil of 10 feet (namely, the grains of zinc electroliti- 
cally dissolved per hour). 
Finding it very desirable to ascertain more precisely than 
I had before done, that the indications of the galvanometer 
are consistent with each other, and in strict relation to the 
quantity of the current passing, I adopted the following 
methods. In the first place a coil of the same length and 
thickness as the 10 feet coil of the galvanometer was pro- 
vided, and as accurately as possible adjusted to the like 
resistance as the galvanometer coil, by means of a diffe- 
rential resistance measurer (Wheatstone's). I found that the 
mere measuring similar lengths of wire from the same 
bundle was not sufficient, inasmuch as the connexions, 
although made by good binding screws, varied the relative 
resistances. Such coil was attached so as to divide the 
circuit with the galvanometer coil. Another length of wire, 
of about five feet, was added to the circuit, and its resistance 
adjusted so that the resistance of the five feet coil, and of 
the two ten feet coils, side by side, were equivalent to the 
original resistance of the galvanometer. 
A pair of elements being put in circuit, the galvanometer 
in its ordinary state balanced 58 grains, the dividing coil 
being attached, and also the 5 feet coil, to equalise the 
resistance, as before-mentioned; exactly one-half of the 
current, of precisely the same quantity as before, passed 
through the galvanometer coil, which then balanced 29 
grains; and on repeating the experiment with a smaller 
pair of elements, the undivided circuit balanced 38 grains, 
and the divided circuit 19 grains; thus proving that within 
very considerable limits, the weight balanced by the galva- 
nometer may be relied on as indicating the quantity of 
current passing. 
It appeared also desirable to ascertain the relative indica- 
tions of the galvanometer as compared with the voltameter. 
