The atmospheric-electric data obtained at College, Alaska, during the International Polar Year have been analyzed in order (1) to separate the universal and local features in their diurnal variation and (2) to compare them with the theory of the electrode-effect. After eliminating the universal diurnal-variation component from the potential-gradient a local component remains. In winter this local component is largely the result of variations in the conductivity near the Earth's surface but in summer there is no correlation between these elements. The calculated ratio of the columnar resistance of the air over College to that over the oceans shows a diurnal variation. The statistical significance of this variation is questioned. If it is significant various hypotheses which could account for such a variation are considered.
The diurnal variation in positive conductivity retains its character throughout the year although its amplitude is about 50 per cent less in summer than in winter. Its variations are attributed chiefly to variations in q and N. In contrast the diurnal variation of negative conductivity changes markedly from summer to winter. The negative conductivity depends greatly on the field in winter whereas in summer the increased wind or factors closely associated with it effectively oppose the action of the field.
The theory of the electrode-effect is reviewed and amplified. Following the method of Scholz additional relations based on both the large- and small-ion theory are developed so that a variety of conditions are covered. Some of the results are presented graphically for ease in comparing with these and other results. The air at College was very quiet in winter and apparently met this restriction of the theory for the results obtained then are shown to be in satisfactory quantitative agreement with the theory.