at low temperatures. The soft annealed iron if cooled slowly to — 185° 0. recovers its original permeability when heated up again to ordinary temperatures. The unannealed iron, however, after cooling to the same low temperature, retains some of its increased permeability when heated up again to 15° C. The unannealed iron cannot be taken over the temperature range again and again with the same definite permeability values at each recurrent temperature, as in the case of the soft annealed iron. The unannealed iron is more or less permanently changed in magnetic character every time it is heated or cooled.
With this transformer, a long series of observations were taken to determine the hysteresis loss corresponding to different inductions when taken at the ordinary temperatures, and the temperature of liquid air. The hysteresis cycles were taken with the ballistic galvanometer, over wide ranges of maximum induction, the transformer being alternately at the ordinary temperature and in liquid air, but no constant magnetic condition could be obtained. In one set of observations, at a given maximum induction the hysteresis loss was increased when the transformer was raised in temperature, and for another series of observations at the same induction it was diminished. It is therefore impossible to make any definite statement with regard to the magnetic hysteresis loss in this unannealed iron ring coil at the two temperatures. The mere fact of immersing the unannealed iron in the liquid air changes its magnetic qualities to such a degree that it is no longer the same material, magnetically considered, after, as before its immersion. One curious fact, however, was noticed very soon with regard to unannealed iron, and that is, that if the unannealed iron ring coil has a small magnetising current passed through its primary coil, the secondary coil being connected to the galvanometer, the sudden immersion of this ring coil into liquid air invariably causes a deflection of the ballistic galvanometer, even when the primary magnetising current remains perfectly constant in value, thus showing a sudden and very large increase in the permeability of the unanneaied iron. Whilst the iron is in the liquid air it retains this increased permeability. If brought suddenly oixt its permeability again diminishes, but not with equal rapidity. This is partly accounted for by the fact that the iron is cooled with immense rapidity when it goes into the liquid air, but it heats up again much more slowly when it is brought out. The definite fact, however, remains, which has been repeatedly observed, that the cooling of this unannealed iron to a low temperature always increases its permeability, as far as we know, no matter whatever may be the magnetising force employed. One difficulty experienced in dealing with unannealed iron is the fact that in taking it up to the high magnetising forces, and by the process required to remove residual
