EXPERIMENTAL AND THEORETICAL.

543

Instead of making the usual connections for showing the spark, a piece of iron wireis screwed into each metallic cylinder, pointed at the other end, and bent nearly atright angles, as seen in the figure; and, when the machine is in motion, these wiresare carried round in a vertical plane in front of, and parallel to, an annular iron disc,i i, which is firmly fixed in the base-board of the machine by the pillar, s. The pointsof the iron wires are directed obliquely towards the iron disc, against which theypress gently during the whole revolution. By this means the two cylinders, and con-sequently the ends of the coil wires, find a metallic connection which is frequentlyinterrupted whilst the points are scraping over the asperities on the surface of the irondisc, and two concentric circles of brilliant scintillations are thus produced. This is,perhaps, the most pleasing experiment yet shown in Magnetic Electricity.

By this arrangement it sometimes happens that the revolving points do not succeedone another properly, in the contacts with the iron disc, in order to produce completecircles of scintillating fire. When this is found to be the case, it may be necessary toremove the longer wire from the outer cylinder, and connect the latter with the iron discby a piece of copper wire, having a loop at one end for the reception of the cylinder,and bent into a hook at the other to hang on the inner edge of the annular iron plate.This wire, however, is a mere substitute for a spring of either steel or brass wire, oneend of which is screwed tight to the back part of the iron disc, and the other pressesagainst the outer revolving cylinder. This spring can he turned off the cylinder atpleasure by slackening the screw in the iron disc, and replaced by the iron point, asin the figure.

The iron disc, i i, may be either rough or smooth, bright or dull; and as much ofthe display depends upon the combustion of the ferruginous matter, steel points arepreferable to iron ones. If the points be filed or ground very fine, each wire maycarry two, at different distances from the centre of motion, which will usuallyproduce four concentric circles of fire.

If, instead of a steel point, one of platinum be used, leaves of laminated metals,covering the face of the annular disc, i i, will be deflagated. Gold, silver, and theDutch metals will, by this means, respectively display the same characteristic colouredlight as by the action of a Voltaic battery. The metallic leaves will be sufficientlyattached to the plate, i i, by first breathing on the latter, and then pressing the formeragainst it. In these experiments, however, it is necessary that the plate, i i, bebright: it may be of iron or any other metal.

The deflagration of mercury is beautifully shown by an apparatus of this kind. Letthe annular disc, i i, be of copper, and the whole face over which the points revolveslightly grooved with a file in close radiations from the inner to the outer edge. Letthis radiated face be amalgamated with a dilute solution of nitrate of mercury, thenwashed in clean water, and afterwards covered with clean mercury, so that the whole