Soon after Curtis and Cole reported the first measurements of the impedance of the squid axon membrane in 1938 they made the important and dramatic observations that while the absolute magnitude of the impedance at 20 kilocycles fell during a propagated action potential to a few per cent of the resting value the measured capacitance fell by no more than a few per cent. The similarity to their earlier results with Nitella indicated that the impedance changes with constant capacitance were likely to be a general property of excitable cells.
In contrast to this reasonable behavior at higher frequencies Cole and Hodgkin noticed in 1939 that below 300 cycles the reactance of the membrane could become positive, a result typical of an inductor rather than a capacitor. Cole and Baker investigated this result in detail and later Cole pointed out that large inductive reactances could be measured in carbon filament incandescent bulbs and in thermistors and gave the proper interpretation that in each case it was the result of a non-linearity which occurred with a delay; the temperature dependence of the resistance of the filament and the thermistor and the voltage dependent ion permeabilities in the axon membrane. We now know that these voltage dependent, delayed changes in the ion permeabilities underlie all of the most interesting electrical phenomena in peripheral nerve fibers. In measurements of impedance care must be taken to distinguish between results due to these permeability changes and those due to the properties of the dielectric of the membrane.
The early measurements of the impedance gave no indication of non-linear effects at frequencies at least above a kilocycle, no appreciable changes in capacitances during applied current pulses