A research investigation of the ionization phenomena which occur in paper-insulated, high-voltage power cables is being made at the Harvard Engineering School under the auspices of the Impregnated Paper-Insulated Cable Research Committee, The paper presents some of the preliminary results which have been obtained, certain tentative conclusions that are suggested by the data and a description of the method developed for making the measurements. It is a preliminary report and in subsequent papers it is intended to record the progress of the investigation as it proceeds. The paper consists essentially of four parts as follows: (a) A number of curves of power, power factor and capacitance taken on samples of cable at two frequencies and over a wide range of temperature. These curves show well-known characteristics, but in order to exaggerate these effects, cable models were made up to simulate the general conditions in a cable — one model consisting of glass and air only and another of glass, air and paper. Tests of these cable models gave very interesting results, particularly with reference to power factor. They also show rather clearly the baffling action of paper and the effects on power, power factor and capacitance when this baffling action is eliminated. In practically all the curves, the power and power factor begin to increase rapidly at a lower voltage gradient than the capacitance. (b) Discussion of the results obtained and certain tentative conclusions which may be drawn from the results so far obtained and reported herein. 1. Ionization in the dielectric of an air condenser increases its capacitance slightly at first and then rapidly as the electron is separated from the atom. In our measurements we accordingly found that the power and power factor increase rapidly at a lower voltage gradient than the capacitance. 2. Ionization which occurs within air spaces in a dielectric may be called “restricted ionization” in that the current is limited because of the remainder of the dielectric in series. 3. With “restricted ionization,” the voltage across each air space reaches a constant finite value with indefinite increase in the over-all voltage. The resistance of the ionized space, therefore, must be inversely as the current. 4. Consequently when this condition is reached, the ionization loss may be proportional to the charging current. 5. The increase of power factor with increase in voltage gradient with subsequent decrease of power factor is due to the fact that the capacitance of the solid dielectric, which is substantially constant, is in series with ionized air spaces whose resistances are inversely as the current. A simple mathematical analysis of this type of electric circuit shows that power-factor curves should have the form obtained in the measurements. Hence power factor alone is not a criterion of the degree of the completeness of saturation by compound. 6. Ionization by its bombarding action may destroy the baffling action of paper. 7. Ionization may produce potential gradients tangential to the surface of the layers of paper which, in conjunction with the bombarding action, may be the cause of the so-called “tree designs.” (c) Discussion of methods for measurements of this sort and a description of a new type of bridge which was devised for the measurement of dielectric losses at these extremely low power factors with the necessary high degree of accuracy. A large air condenser used as a standard and a vibration galvanometer of unusually high sensitivity and wide range of tuning were designed for use as a detector. The bridge is quickly and accurately balanced by varying a resistance and a mutual inductance. The angle of defect of the standard air condenser, although extremely small, is nevertheless very important in measurements of this character. This angle was measured by a substitution method and corrections made accordingly. (d) Four appendixes discussing respectively (1) method of measuring defect angle of the standard air condenser, (2) mathematical analysis of a condenser of composite dielectric consisting of air and solid homogeneous material, (3) measurement of the high-voltage ratio, and (4) measurements of the capacitance, and other electrical constants of an ionized air space in series with a solid dielectric.