The experimental and theoretical results are presented with their peak deflections coincident, which is compatible with the 01 mm experimental uncertainty of the beam position. In practice, one expects the theoretical curves to be slightly displaced, since they are derived from a path-difference calculation, rather than by ray tracing, and do not take account of the displacement of the beam while passing through the cell. Although this error term is related to the deflection angle, its effect on the theoretical curve is small and the simple theoretical calculation has proved quite useful for predicting the cell performance for different thickness/width ratios. When the cell is switched from parallel alignment to clinotropic alignment, or back, there is a region of the cell where the molecules will rotate in opposite directions and cause turbidity. Unfortunately, this includes the region giving maximum deflection, and so it is not possible to produce a continuous scan if the beam is focused near that position,