To test an electric cable for conductivity, it would be only necessary to expose the metallic conductor _G_, and put it in the water of the tank. If the conductivity were good, then the whole of the current would pass through the cable and the needle of the galvanometer would be violently deflected. If the continuity were broken, no deflection would be observed. _Defects observed in the Conductivity of the Cable._--To ascertain the position of a defect in the insulation of a cable, as indicated by the tests above described, it would be only necessary to keep a continuous current flowing through the cable, and gradually take it out of the tank. If the fault existed at a single point, the deflection of the needle would be suddenly reduced at the moment of that point of the cable being lifted out of the water, and therefore its position would be determined with considerable accuracy. Should several defects exist as each was lifted out, a sudden reduction of the deflection would occur. _Discharge Test._--The conductor of an electrical cable may be broken without destroying the insulation, and on applying the foregoing tests, good insulation would be indicated, but no conductivity, and no information would be given as to the position of the fault. Under such circumstances the following test must be applied:-- Put one pole of a very powerful battery to earth, and charge one end of the defective cable, then immediately discharge it through a reflecting galvanometer, and note the extreme limit of the swing of the needle, then, charge the other end of the cable in a similar manner, and discharge it through the same galvanometer, noting as before the swing of the needle. This should be done three or four times, and the average of the deflections taken. Then the position of the fault would be indicated by the proportion between the average deflections in each case, and the cable might safely be cut at that point. Should the precise position of the fault not be discovered in thus cutting the cable, each section should be tested again for conductivity, and that in which a fault was still found to exist should be again tested by the discharge as before. _Test of Electrical Resistance of Cable._--This is effected by balancing it against the Wheatstone balance, in a similar manner to that explained for a fuze. The electrical resistance of the conductor of a cable affords a very correct indication of the quality of the metal of which it is composed. For a very delicate test the reflecting galvanometer should be used. _Electrical Test of Insulated Joints._--Insulated joints and connections, whether of a permanent or temporary nature, should be tested electrically, in a precisely similar manner to that explained for electric cables. They should be soaked for forty-eight hours, and then tested for insulation, conductivity, and electrical resistance.