When an iron wire is magnetised longitudinally, it lengthens in the direction of magnetisation, according to an old discovery of Joule's. More recently Wiedemann showed that when wire is at the same time magnetised circularly it tends to twist. Thus, if an iron wire be fixed at one end, and stretched vertically by means of a mass attached to the free end, the free end will twist round when the wire is both traversed by one current and magnetised by another which traverses a helix surrounding it. If the wire is magnetised so as to have the north pole down, a down current will make the free end twist in the direction of the hands of a watch as looked at from above. Eeversal of either current reverses the direction of twist; reversal of both produces no alteration. Maxwell and Chrystal have pointed out that Wiedemann's phenomenon can be explained by means of Joule's.