AbstractFrequencies and amplitudes of oscillations for nineteen pure systems were measured photographically. The period of oscillation was longer than that predicted by Lamb. The discrepancy was not due to wall effects, viscosity, or velocity of fall but to amplitude of oscillation. Modification of previous expressions included an amplitude function which could be experimental or empirical. Oscillations began near the peak velocity, and a vortex trail was necessary for them to take place. Oblate‐prolate oscillations did not cause drop breakup, as all systems ceased to oscillate and wobbled randomly below maximum drop size. Oscillations do not decay with tim