The difficulties experienced in connection with cable-ends when testing at high voltages in air are dealt with; and the reason for the preliminary sparking at the lead sheath and its tendency thereby to produce spark-over is explained. Various methods in normal use of overcoming this disadvantage are noted and are described in the figures. In particular, a suggestion is put forward for the use of semi-conducting material, as a temporary method of attaining greatly improved results for ordinary routine testing work. Comparative results are given both in tables and in curves, showing the advantages obtained by using semi-conducting carbon paper applied in various ways on or over the surface of the cable insulation.Theoretical lines-of-force diagrams are shown to explain these results and in what manner they are produced, and these are supported to some extent by practical measurements. Fusing-current tests are given for various samples of the material, and in view of these a possibility of the more permanent use of the principle is suggested in connection with the installation of insulators and bushings, etc.