with emery-paper. The wire so treated may, after a long time, exhibit partial recovery of its responsive power.
The facts described above seem to show that the enhancement or depression of response may, at least to a considerable extent, be due to the increase or diminution of molecular mobility conferred by the chemical reagents. With a given stimulus, the height of response and the form of the response curve will be determined by the element of molecular friction. In connection with this, it is instructive to obtain records of the vibrations of a torsional pendulum, the friction of which may be gradually increased by immersing the pendulum more or less in a viscous fluid or sand. The various types of response-curves in metals are found to be very similar to those thus obtained.
Of these I give an interesting example. With moderate friction the successive curves obtained with the pendulum are like those given in the left of fig. 13 (a). With increased friction the height of the curve is diminished, the maximum is reached later, and the recovery is prolonged (like the curve in the right of fig. 13 (a)). With still greater friction the recovery is arrested.
Fig. 13.—Photographic records showing the effect of "molecular arrest." The two curves to the left of each set show the normal response; curve to the right in (a) shows partial and in (b) complete arrest, produced by the reagent.
It would appear as if the reagents which abolish response in metals produce a similar molecular arrest. The following photographic records seem to lend support to this view. If the oxalic acid be applied in large quantities, the abolition of response is complete; but
