An experimental verification of Rayleigh’s interpretation of the Sondhauss tube