It is to be regretted that so sound a reasoner and so careful an experimenter had not the great advantage of the assistance of such suitable instruments for this class of research as the mirror-galvanometer and the telephone. But, although he could not practically demonstrate the effects which by him could be so clearly seen, it redounds to his credit that, as the improvement in instruments for this kind of research has advanced, the results he sought for have been found in the direction in which he predicted.

A and B will now be placed a definite distance apart, and comparatively slow reversals from ten Leclanché cells sent through spiral A; you will observe the amount of the induced current in B, as shown on the scale of the galvanometer in circuit with that spiral. Now midway between the two spirals will be placed a plate of iron, as shown in Plate 2, and at once you observe the deflection of the galvanometer is reduced by less than one half, showing clearly that the presence of the iron plate is in some way influencing the previous effects. The iron will now be removed, but the spirals left in the same position as before, and by increasing the speed of the reversals you see a higher deflection is given on the galvanometer. Now, on again interposing the iron plate the deflection falls to a little less than one-half, as before. I wish this fact to be carefully noted.

The experiment will be repeated with a plate of copper of precisely the same dimensions as the iron plate, and you observe that, although the conditions are exactly alike in both cases, the interposition of the copper plate has apparently no effect at the present speed of the reversals, although the interposition of the iron plate under the same conditions reduced the deflection about fifty per cent. We will now remove the copper plate, as we did the iron one, and increase the speed of the reversals to the same as in the experiment with the iron, and you observe the deflection on the galvanometer is about the same as it was on that occasion. Now, by replacing the copper plate to its former position you will note how rapidly the deflection falls. We will now repeat the experiment with a plate of lead; you will see that, like the copper, it is unaffected at the low speed, but there the resemblance ceases; for at the high speed it has but very slight effect. Thus these metals, iron, copper, and lead, appear to differ as widely in their electrical as they do in their mechanical properties.

Of course it would be impossible to obtain accurate measurements on an occasion like the present, but careful and reliable measurements have been made, the results of which are shown on the sheet before you, marked 3.

It will be seen by reference to these results that the percentage of inductive energy intercepted does not increase for different speeds of the reverser in the same rate with different metals, the increase with iron being very slight, while with tin it is comparatively enormous. It was observed that time was an important element to be taken into account while testing the above metals, that is to say, the lines of force took an appreciable time to polarize the particles of the metal placed in their path, but having accomplished this, they passed more freely through it.

Now let us go more minutely into the subject by the aid of Plate IV., Figs. 1 and 2. In Fig. 1 let A and B represent two flat spirals, spiral A being connected to a battery with a key in circuit and spiral B connected to a galvanometer; then, on closing the battery circuit, an instantaneous current is induced in spiral B. If a non-magnetic metal plate half an inch thick be placed midway between the spirals, and the experiment repeated, it will be found that the induced current received by B is the same in amount as in the first case. This does not prove, as would at first appear, that the metal plate fails to intercept the inductive radiant energy; and it can scarcely be so, for if the plate is replaced by a coil of wire, it is found that induced currents are set up therein, and therefore inductive radiant energy must have been intercepted. This apparent contradiction may be explained as follows:

In Fig. 2 let D represent a source of heat (a vessel of boiling water for instance) and E a sensitive thermometer receiving and measuring the radiant heat. Now, if for instance a plate of vulcanite is interposed, it cuts off and absorbs a part of the radiant heat emitted by D, and thus a fall is produced in the thermometer reading. But the vulcanite, soon becoming heated by the radiant heat cut off and absorbed by itself, radiates that heat and causes the thermometer reading to return to about its original amount. The false impression is thus produced that the original radiated heat was unaffected by the vulcanite plate; instead of which, as a matter of fact, the vulcanite plate had cut off the radiant heat, becoming heated itself by so doing, and was consequently then the radiating body affecting the thermometer.

The effect is similar in the case of induction between the two spirals. Spiral A induces and spiral B receives the induced effect. The metal plate being then interposed, cuts off and absorbs either all or part of the inductive radiant energy emitted by A. The inductive radiant energy thus cut off, however, is not lost, but is converted into electrical energy in the metal plate, thereby causing it to become, as in the case of the vulcanite in the heat experiment, a source of radiation which compensates as far as spiral B is concerned for the original inductive radiant energy cut off. The only material difference noticeable in the two experiments is that in the case of heat the time that elapses between the momentary fall in the thermometer reading (due to the interception by the vulcanite plate of the radiant beat) and the subsequent rise (due to the interposing plate, itself radiating that heat) is long enough to render the effect clearly manifest; whereas in the case of induction the time that elapses is so exceedingly short that, unless special precautions are taken, the radiant energy emitted by the metal plate is liable to be mistaken for the primary energy emitted by the inducing spiral.