Voltage clamp experiments measured transients from 10 mV steps applied about different membrane potentials, VT. Their analysis employed the Fourier transform relationship between dielectric spectra of permittivity as a function of frequency, and the step transient admittance, as well as established methods. 2. The membrane capacitance measured between -85 mV and about -50 mV rose monotonically and was associated with simple decaying transients in both 'on' and 'off' steps. At more depolarized potentials the capacitance increased sharply and was associated with a charging current of complex form, before falling again beyond the transition potential. 3. Step-transient responses for dielectric analysis were sampled in the above voltage range. Dielectric spectra of the non-linear transients were obtained by subtracting Fourier transforms of transients at VC = -85 mV from test transforms at VT. 4. The imaginary transform coefficients represent a spectrum of dielectric loss against frequency. These showed two non-linear components. The q beta component formed a broad peak, when charge movements were simple monotonic decays. A sharp low-frequency q gamma peak became superimposed at particular voltages when charge-movement kinetics became complex. 5. In contrast, 'off' transients were simple monotonic relaxations. Their transforms showed only one dielectric loss peak, whose frequency was relatively voltage-independent when q gamma occurred in 'on' transforms. 6. Both altering holding potential from VH = -85 to -50 mV and adding 1 mM-tetracaine to the bathing solution reduced the dependence of capacitance on voltage. The non-linear polarization currents became simple monotonic relaxations at both the beginning and end of the voltage step. 7. It is concluded that charge movements are composed of at least two components: q beta, and the tetracaine and voltage-inactivated q gamma. Any causal relationship between q beta and q gamma and the membrane processes they might underlie would be expected to be complex.
