arris on a New Electrometer, &c. &c.

288. ii. _Magnetism._--Voltaic electricity has most extraordinary and
exalted magnetic powers. If common electricity be identical with it, it
ought to have the same powers. In rendering needles or bars magnetic, it is
found to agree with voltaic electricity, and the _direction_ of the
magnetism, in both cases, is the same; but in deflecting the magnetic
needle, common electricity has been found deficient, so that sometimes its
power has been denied altogether, and at other times distinctions have been
hypothetically assumed for the purpose of avoiding the difficulty[A].

[A] Demonferrand's Manuel d'Electricite dynamique, p. 121.

289. M. Colladon, of Geneva, considered that the difference might be due to
the use of insufficient quantities of common electricity in all the
experiments before made on this head; and in a memoir read to the Academie
des Sciences in 1826[A], describes experiments, in which, by the use of a
battery, points, and a delicate galvanometer, he succeeded in obtaining
deflections, and thus establishing identity in that respect. MM. Arago,
Ampere, and Savary, are mentioned in the paper as having witnessed a
successful repetition of the experiments. But as no other one has come
forward in confirmation, MM. Arago, Ampere, and Savary, not having
themselves published (that I am aware of) their admission of the results,
and as some have not been able to obtain them, M. Colladon's conclusions
have been occasionally doubted or denied; and an important point with me
was to establish their accuracy, or remove them entirely from the body of
received experimental research. I am happy to say that my results fully
confirm those by M. Colladon, and I should have had no occasion to describe
them, but that they are essential as proofs of the accuracy of the final
and general conclusions I am enabled to draw respecting the magnetic and
chemical action of electricity (360. 366. 367. 377. &c.).

[A] Annales de Chimie, xxxiii. p. 62.

290. The plate electrical machine I have used is fifty inches in diameter;
it has two sets of rubbers; its prime conductor consists of two brass
cylinders connected by a third, the whole length being twelve feet, and the
surface in contact with air about 1422 square inches. When in good
excitation, one revolution of the plate will give ten or twelve sparks from
the conductors, each an inch in length. Sparks or flashes from ten to
fourteen inches