Page:Scientific Memoirs, Vol. 2 (1841).djvu/26

 14 Sciences of Berlin, stated to me that he had not obtained the least development of magnetism by means of a magneto-electric helix, which he had introduced into a piece of a gun-barrel. I was then of the opinion of this philosopher, that this negative result must be attributed to some accidental property of the iron; the more so, as a helix which he had coiled around the barrel had not produced any considerable effect. M. Dove does not appear to have published any additions to this isolated experiment. Altogether independent of this previous experiment, M. Parrot has made the discovery which he has published in his memoir, (Bulletin scientifique, No. 16) that a magneto-electric helix, placed in the interior of a hollow cylinder, does not communicate any magnetic force to the soft iron. This coincidence of observations does not allow of the admission of any accidental state of the iron which has been exposed to the influence of the current. We might probably arrive at the explanation of this curious fact, by following up the electro-dynamic views of M. Ampère on the constitution of magnets. Regarding the magnets as an assemblage of elementary wires, parallel to the axis, and surrounded by electric currents, which are all in the same direction relatively to the axis, M. Ampère is obliged, in order to comply with known facts, to adopt a covered surface, which represents the total effect of these isolated currents, or of those solenoids which may be substituted in their stead. In conformity with experiment, it is unnecessary to take any notice of the inner currents, as they for the most part destroy each other. The direction of the magnetization is determined by that of the currents; so that if we suppose the magnet to be directed by the action of the earth, the currents proceed from east to west in the upper part, and from west to east in the lower part. This is merely a small omission which has escaped this illustrious philosopher, if he has not specially added that, in order to be able to replace the elementary effects by the magnetic covering, it is absolutely necessary to admit also the condition, that the enveloped currents have, relatively to the axis of the whole body, the same direction as the elementary currents have relatively to the axis of the wires. It is this condition, however, which constitutes the distinctive character of a magnetic surface. For it suffices to look at fig. 1. in Pl. I. to judge of the discordance of these directions relatively to the inner surface, and of their concordance relatively to the external surface. The experiments of Mr. Barlow, and the calculations of M. Poisson, to