Magneto-Electric Induction. 451 



III. Vary diameter of disk. 



IV. Cut metallic disk. 

 V. Vary thickness. 



VI. Vary nature of metal of disk. 

 VII. Examine liquids in view of the determination of their 

 conductivities. 



I. The Relation beticeen Velocity and Deflexion. — The appa- 

 ratus used for establishing this relationship is, in its later 

 form, shown in PI. XIII. fig. 1 one third the true size, the 

 difference being that the torsion-thread was held by a clip 

 attached directly to a glass shade instead of to the sliding 

 arrangement shown. 



The motive power was a Fromont's electro-magnetic engine. 

 The elastic band from the engine passed round one or other 

 of the wheels on the vertical spindle, p, carrying a pair of 

 magnets, n .<?, each 3 T ^- inches long x 1J x J, placed with their 

 similar poles together. A screw on the spindle geared with 

 a worm-wheel, iv, with ninety-seven teeth. The time of a 

 revolution of this could be measured either by watching a 

 mark on the wheel pass a fixed pointer, or by its making a 

 contact every revolution so as to sound a bell. The torsion- 

 thread used was the hair spring of a watch, to which was at- 

 tached the copper disk by means of a screw clip. The speed 

 was altered by placing the driving-band on one or other of 

 the pulleys on the vertical shaft, or on the Froment engine, or 

 by altering the strength of the battery driving it, or by means 

 of a friction-break. 



The disk was graduated into degrees on its upper face, and 

 the reading made by keeping a vertical edge, the eye, and 

 the torsion-thread in one plane, and noting the division or 

 part of one intersected by that plane, so as to avoid parallax. 

 When the deflexion of the disk exceeds 20° or 30° the wire 

 takes a " set," especially if it is kept long at its maximum 

 twist. It appears completely to recover itself in time ; but 

 such recovery is at last so slow that it is preferable to rede- 

 termine the zero immediately after each experiment, and to 

 allow for this, assuming that at the moment of observation (that 

 is, when the disk is at its greatest excursion) the set in the wire 

 is the same as it is immediately after when the zero is rede- 

 termined. When the deflexion is considerable it is impossible 

 to keep it absolutely constant ; accordingly the mean was 

 taken of several maxima and minima, and the rate was regis- 

 tered after each set of observations. The means of the means 

 of deflexion and rate were finally taken. When the rate is 

 high and the deflexion considerable, the disk begins to swing, 

 the axis describing a cone, until reading becomes impossible, 



