The frequency of phase oscillation in a synchrotron is so low as to make possible resonances with harmonics of the power‐supply frequency. The interaction may take place through ripple in the magnetic field or in the frequency or amplitude of the accelerating voltage, the first two interacting very strongly and the third moderately. Thus, a small ripple in the magnet power supply output, though it produces only an extremely small ripple in the magnetic field because of the magnet's inductance, may tend to cause large phase oscillations. The detuning effect of the fall in phase oscillation on frequency with increasing amplitude limits the resulting amplitude, however, and may even reduce the amplitudes of initially large phase oscillations. The ripple in the accelerating field is unlikely to be sufficient at a resonant frequency to cause trouble. In addition to ripple, there will be noise which, particularly through modulation of the frequency of the accelerating field, may result in time in considerable phase oscillations because the noise contains a wide range of frequencies, one of which is always in resonance throughout the acceleration cycle.