“…Usually several interference maxima from each source are recorded simultaneously (by chopping at different frequencies) as the path difference changes slowly. Air pressure scanning, either at atmospheric pressures with a piston [Rowley and Wallard, 1973;Blaney et al, 1977] or from vacuum with a controlled leak Hanes et al, 1973;Ito and Tanaka, 1978;$chweitzer et al, 1973], is the most usual way of achieving a smooth linear scan. Problems of nonlinearity arise from the linear rate of volume (rather than pressure) change with piston scanning and from temperature changes induced by the pressure changes.…”