Theories of the initiation and propagation of transverse sinuous disturbances on an air jet emerging from a flue slit into a transverse acoustic velocity field are reviewed and brought into a form suitable for comparison with experiment. Available experimental data is then reviewed and is found to be only moderately well accounted for by the existing theory. A new series of experimental determinations of wave velocity on a jet under conditions relevant to sound generation in musical instruments is then described, these measurements extending over a larger range of frequencies and blowing pressures than earlier measurements. For jets with Reynolds number below about 1000 the wave velocity is found to be constant along the length of the jet and to depend not upon the jet velocity V but rather upon the integral J of V**2 across the width of the jet. At low frequencies, omega , the wave velocity is approximately 0. 7 (J omega )**1**/**3, in agreement given by 50J (where S. I. units are implied). Wave amplification along the jet is also critical and the wave amplitude found to increase like exp( mu x) where mu is positive only below a critical frequency omega *, dependent on the jet parameters. For the jets studied, mu is of order 1 multiplied by 10**3m** minus **1 below omega *. A simple theory based on dimensional analysis is proposed to describe these experimental results.