The insulation surrounding the conductor of a cable is supposed to be uniform in regard to quality of material, density, and thickness. The resistance which it offers to the passage of a current through it will then vary inversely with its length. In comparison the insulation resistance of 1 mile of cable is taken as the standard. This insulation has a large negative temperature coefficient; that is, an increase of temperature lowers its resistance. It is customary to reduce all insulation resistance to that at a standard temperature of 60° F., and for this purpose reduction factors applicable to the particular insulation compound should be furnished with the cable. (Note: It has been found that for most compounds, if the logarithms of the resistance are plotted as ordinates against the temperature in degrees F. as abscissæ, the resulting curve will be very nearly a straight line.) The ordinary methods of measuring resistance—that is, by means of a Wheatstone bridge, or by fall of potential, or by voltmeter—can not be used in measuring resistance as high as that of the insulation of a submarine cable. For this the direct deflection method is employed. In brief, this consists of the following steps: First. The deflection produced in a galvanometer by a current from a battery through a known resistance, usually 100,000 ohms, is determined, whence is calculated the resistance through which this same battery would produce a deflection of one point using the unity shunt. This is expressed in megohms and is called the galvanometer “constant” under the conditions. Second. The deflection produced by the current from the same battery through the insulation of the cable is determined, whence, from “First,” the corresponding number of megohms is calculated. Third. This multiplied by the length of the cable in miles and corrected for temperature gives the required insulation resistance per mile. This testing can be made most satisfactorily on dry days, but a close adherence to the instructions herein given relative to the preparation of the cable ends, the insulation of the cable lead and of the battery, and the drying out of the test room and instruments should enable satisfactory work to be done under adverse conditions of weather or climate. The following apparatus is required: Reflecting galvanometer, universal shunt, special testing key, 100,000-ohm resistance box, battery of dry cells giving approximately 100 volts, and stop watch. Figure 16 shows diagrammatically the arrangement of the apparatus for testing a reel of cable. As a rule the instruments should be so placed that one person may manipulate the key and the shunt while at the same time observing the galvanometer. The 100,000-ohm box, as a protection to the galvanometer in testing, is always kept in the circuit and its value should be subtracted from the resistance determined, except in the case of high insulation resistance when it will not be necessary to make the subtraction.