Construction of a wide-frequency range double heterodyne conductance bridge and its use in the investigation of polarisation errors in conductance measurements

(1) Developments in a.c. conductance techniques during the past ninety years have been reviewed, and a brief outline is given of the older theories regarding electrolytic polarisation. (2) A conductance bridge - incorporating the double heterodyne principle - has been constructed, capable of giving resistance readings to an accuracy of 0.01% over a range of frequencies covering the best part of 100 kc/s. It has also been found possible to calibrate the oscillator so that frequency settings can be guaranteed to an accuracy of , at least, 0.1% in the range: 2 kc/S to 50 kc/S. (3) The Wheatstone Bridge Network has been slightly modified to enable measurements at the high frequencies. (4) Resistances in the measuring arm of the bridge have been calibrated 'in situ' by the method of intercomparison. (5) A brief description is given of the modern theories regarding electrode processes and modern methods of eliminating electrode effects. (6) Two types of conductance cells, with bright Pt electrodes, have been used to carry out measurements on potassium chloride solutions: (a) Thomas- Gledhill Cell (b) Nichol-Fuoss Cell. The latter incorporates concentric, cylindrical electrodes with the lead to the outer electrode acting as an electrical shield for the lead to the inner electrode. This cell was constructed and used for the first time in this laboratory. (7) From resistance-frequency graphs plotted, it is shown that the Jones and Christian extrapolation procedure cannot be applied (with any degree of confidence) to obtain the true resistance, when measurements are effected over an extended range of frequencies. (8) The method of resistance-reactance diagrams is discussed and applied to various networks of resistances and capacitances. (9) By drawing resistance-reactance diagrams for the experimental readings obtained, equivalent circuits have been derived - for all the solutions investigated in the N-F cell, and for the approx. O.OlD solution in the T-G cell - which approximate to cell behaviour in the range: 500 c/s to 75 kc/s. The less concentrated solutions in the T-G cell show peculiar behaviour at the high frequencies. (10) Probable reasons are advanced for deviations from linearity on resistance-frequency graphs. (11) A new method is proposed for determining the true resistance of solutions measured in cells of the N-F type. Summary, p. 166-167