The harmonic structure of the cochlear microphonic (CM) and that of a sound-elicited signal which we have considered as an (apparent) changing resistance (CR) were simultaneously determined in scala media of the first turn of the guinea pig cochlea. We analyzed our data in the context of the Davis variable resistance hair-cell model (1965), which predicts CM and CR to be proportional to each other. But, plotted as functions of the sound-pressure level, CM and CR were found to have qualitatively similar but quantitatively disproportionate spectra. The preparations with the highest endolymphatic potential showed the least correspondence between the spectra of the two measured quantities. The phase angles of the fundamental components in CM and CR were equal within approximately 10 degrees, but the phase of the even harmonics of the two independent measures commonly differed by approximately 180 degrees at lower SPLs. Although most data were collected using 160-Hz tonal stimulation, tones with frequencies up to 1280 Hz produced qualitatively similar results. The CM and the CR both varied slightly with the level of the alternating current used to probe the CR. Considered on a quantitative basis, consistent with the accuracy of our measurements, any model which reduces to a fixed source, a fixed resistance, and a single linear, time-varying resistance cannot mimic the most significant, commonly found aspects of our CM and CR data. An alternate model incorporating a nonlinear, time-invariant resistance is able to account for some of the data. The output of the model is correctly considered a (time) changing resistance, or apparent changing resistance; but the model demonstrates that similar experimental results are not necessarily evidence for a time-varying resistor as originally proposed by Davis.