Detections of variations in the frequency of the Chandlerian compounent of polar motion have been reported by several investigators, starting at least as early as Chandler's (1892) report. The concept has never gained general acceptance because no credible cause has been identified. Investigations reported in this paper suggest that the Chandlerian component may be frequency modulated as a linear function of the polar motion magnitude. The variation in frequency is estimated to be ~0.03 tp 0.06 cycles per year increase per 0.1 second of arc decrease in the polar motion magnitude. Nonequilibrium response of the oceans to polar motion, which takes the form of a pole tide that varies with time in enhancement and phase lag, is suggested as a possible cause of the frequency modulation.