An ammeter inserted in the circuit will show at a glance what current is passing at any particular speed, and voltmeter readings are taken at the terminals of the machine, when the same is standing still as well as when the armature is running, because the E.M.F. indicated when the armature is at rest alone determines the commercial efficiency of the motor, whereas the E M.F. developed during motion varies with the speed until it nearly reaches the E.M.F. in the leads; at that point the theoretical efficiency will be highest. Calculations are greatly facilitated, and the value of tests can be ascertained quickly, if the constant of the brake is ascertained; then it will be simply necessary to multiply the number of revolutions and the weight at the end of the lever by such a constant, and the product gives the horse power, because, with a given Prony brake, the only variable quantities are the weight and the speed. All the observations, electrical and mechanical, are made simultaneously. The electrical horse power put into the motor is found by the well known formula C x E / 746; this simple multiplication and division becomes very tedious and even laborious if many tests have to be made in quick succession, and to obviate this trouble, and prevent errors, I have constructed a horse power diagram, the principle of which is shown in the diagram (Fig. 1). Graphic representations are of the greatest value in all comparative tests. Mr. Gisbert Kapp has recently published a useful curve in the _Electrician_, by means of which one can easily compare the power and efficiency at a glance (Fig. 2). The speeds are plotted as abscissae, and the electrical work absorbed in watts divided by 746 as ordinates; then with a series-wound motor we obtain the curve, EE. The shape of this curve depends on the type of the motor. Variation of speed is obtained by loading the brake with different weights. We begin with an excess of weight which holds the motor fast, and then a maximum current will flow through it without producing any external work. When we remove the brake altogether, the motor will run with a maximum speed, and again produce no external work, but in this case very little current will pass; this maximum speed is om on the diagram. Between these two extremes external work will be done, and there is a speed at which this is a maximum. To find these speeds we load the brake to different weights, and plot the resulting speeds and horse powers as abscissae and ordinates producing the curve, BB. Another curve, e = B/E made with an arbitrary scale, gives the commercial efficiency; the speed for a maximum external horse power is o a, and the speed for the highest efficiency is represented by o b. In practice it is not necessary to test a motor to the whole limits of this diagram; it will be sufficient to commence with a speed at which the efficiency becomes appreciable, and to leave off with that speed which renders the desired power.