Volume 48 number 2 (2020)
We use three complete light curves for MU Cancri, a faint W-type W UMa contact binary, to investigate possible mechanisms for changes in such binaries’ light curves. The standard Roche model, as implemented by the Wilson-Devinney code, does not fit the observations at their level of precision. Most of this discrepancy can be explained by placing one or two moderate starspots (rspot ~ 10–12°) on the more massive component. However, this does not resolve the discrepancy, since the solutions for the three epochs have different mass ratios, implying unmodelled changes in eclipse depths. This, in turn, implies that more spots are changing the depths in unpredictable ways. Thus we are confronted with limits on just how precisely light curve solutions can define the physical properties of a contact binary. We use spectra to classify the star (G3–G7) and to measure a spectroscopic mass ratio (q = 2.63), significantly closer to 1.0 than the photometric mass ratios (3.0–3.3), but this difference is unlikely to be caused by third light. And we also extend the period study of Alton and Stępień.