The sample is made in the form of a ring, A, most conveniently of rectangular cross section. The ratio of the radial thickness to the external diameter should not be greater than The dimen sions of the ring are measured. On it is wound, uniformly spaced, the search coil for B, consisting of a few turns of well insulated fine wire. The magnetizing coil is uniformly wound on top of this. (In the diagram only parts of the windings are indicated, the B coil being at the upper side of A). The magnetizing current derived from the battery B, is regulated by the resistance R, and measured by the galvanometer or ammeter K is a rocking key, which joins ac and bd when thrown over to the right and ae and bf when to the left. If the switch S is closed, K acts simply as a current reverser. If S is open, throwing the key from right to left reverses the current and also diminishes it by an amount depending on the adjustable resistance R2. The B coil is connected in series with a ballistic galvanometer G2, and with a secondary coil of a few turns wound over a standardizing solenoid E. By means of the three-way switch C, the primary current may be passed through the primary of E, to calibrate the galvanometer, or through the magnetizing winding on the ring. Knowing the dimensions of the ring, the field due to any current may be calculated from the num ber of turns on the magnetizing coil, and the change in induction corresponding to any throw of the galvanometer from the number of turns on the B coil.

To determine the simple B, H curve (fig. I I) the specimen is first demagnetized, by repeated reversals of the current which is gradually reduced to zero. The switch S is closed. The current is set at a suitable value, and the reversing switch K operated some twenty times, with F closed. F is then opened and a reading of the throw taken corresponding to a reversal of the current. This pro

cess is repeated to test whether the throw is constant (if not, the demagnetization process must be repeated). Half the throw for reversal gives the value of B on oab (fig. I I) corresponding to the field due to the current. Other points on oab are similarly determined.

When the hysteresis curve is to be obtained the current is adjusted by R to give the limiting values of the magnetizing force which it is intended to apply. After several reversals, the point a (fig. 13) is determined as before. With the key K to the left, S is suddenly opened, having been adjusted to a suitable value; the current is thus reduced, and from the galvanometer throw, the reduction of the induction and hence a point on the ac branch of the curve is found. By switching K from right to left with S open, points along cd can be found. The ascending curve dea is an inverted copy of acd.

The work done on a cubic centimetre of iron through a hys teresis cycle is equal to being proportional to the area of the hysteresis loop. This work appears in the form of heat. For technical applications a knowl edge of this heat loss may be very necessary—in transformers, for example, the hysteresis loss should be small. A number of arrange ments have, therefore, been de vised for measuring the hysteresis loss more expeditiously than is possible by the ballistic method.

In Ewing's hysteresis tester, the sample, arranged as a bundle of strips, is rotated about a horizontal axis between the poles of an upright C-shaped magnet, the magnet being supported so that it can turn about an axis in line with that about which the specimen rotates. The deflection of the magnet gives a measure of the hysteresis loss. Extensive use is made of wattmeter methods by which the total heat loss due both to hysteresis and eddy currents may be determined.

Using ring-shaped specimens, so that the uncertain correction for end effects is eliminated, the ballistic method is capable of great accuracy up to about H= 200 gauss, and is usually employed where reliable results are desired as to the magnetic properties of materials where the permeability is a maximum, and where the remanence and hysteresis losses require to be accurately known.