IT is well known that the resistance (R) of a wire of bismuth, as measured with a constant current, increases under the influence of a magnetic field, and that this increase depends on the strength of the field and its direction with reference to the current in the wire. If the current traversing the bismuth is oscillatory, the resistance has a value O outside the magnetic field, or in a field in which the lines of force are parallel to the wire which is less than R. If, however, the wire is perpendicular to the lines of force of a field greater than 6000 C.G.S. units, the resistance O is greater than R; the difference O – R increases from this point pretty rapidly as the strength of the field increases. These changes are not due to alterations in the self-inductor, since they are independent of the form of the bismuth spiral. This curious phenomenon has lately been examined by M. I. Sadovsky (Journal de la Socitte Physico-Chemique de Russe, xxvi. 1894, and Journal de Physique, April 1895), who sums up the results of his experiments as follows: (1) The difference in the resistance of Dismuth observed with constant or alternating currents is measurable outside a magnetic field with 300 alternations per second, and can be detected in magnetic fields with only three or four alternations per second; (2) this difference depends on the number of oscillations per second, and without the magnetic field increases with the increase in the frequency of the alternations; (3) the resistance which bismuth, in a strong magnetic field, offers to an increasing current is greater, and that to a decreasing current less than the resistance for steady currents. The difference between the resistances to an increasing and decreasing current increases with the rate of change in the strength of the current (dC\dt), and this difference is more marked with strong currents than with weak. Thus M. Sadovskyhas discovered the remarkable fact that for variable electric currents the resistance of bismuth changes with any change in1/C or dC/dt where C is the C dt current. The author mentions that the effects observed cannot be due to self-induction, or they would occur when the bismuth is not in a magnetic field. In a note on the above paper in the Journal de Physique, M. Sagnac considers what would happen if the same series of experiments were repeated with an iron wire. A straight cylindrical iron wire becomes, when traversed by a current C, circularly magnetised; the energy due to this magnetisation being, according to Kirchhoff, ππlC2, where K is the susceptibility and l the length of the wire. This energy may possibly increase the coefficient of self-induction by ππl. From Klemenĉiĉ's data the order of the change in the apparent resistance can be calculated. For weak magnetic fields in which K has a large value, the difference between the value of the apparent resistance for steady currents and for increasing currents may amount to several hundredths of the value of the resistance for steady currents.