The following is a paper by H. Aspden published in Speculations in Science and Technology, v. 5, pp. 421-431 (1982).
Abstract: Light speed anisotropy experiments give unreliable results if account is not taken of the effects of anisotropy upon mirror reflections in lasers used in the tests. The deviation of ray paths when reflected by mirrors moving relative to the optical reference frame cannot be ignored. In particular, anomalous results are to be expected if curved mirrors are used and full account is not taken of all reflections at mirror surfaces. This is discussed by detailed reference to a recent important experiment reported by Brillet and Hall. Though finding a null anisotropy in light speed in the cosmic reference frame, they nevertheless noted a significant and persistent spurious signal indicative of anisotropy in the laboratory frame. This signal may be a measure of the laboratory's West-East motion owing to Earth rotation relative to a light reference frame coincidental with the Earth's inertial frame. This conclusion has a speculative aspect because it may impose some qualification upon the universal validity of the Principle of Relativity. Although the effect has been apparent in other laser experiments, specific experiments aimed at measuring the Earth's West-East motion, as opposed to angular speed of rotation, have yet to be performed. It is submitted that such tests are of fundamental importance.