5 Colour-index curves

Examination of the colour index curves reveals the expected bluening in colour during the secondary eclipse (phase 0$.\!\!^{\scriptscriptstyle\rm p}$5) in the (b-y) and (v-y) curves, a result of decreasing the contribution of the red secondary relative to the blue primary. Note that the primary eclipse also exhibits a bluening in colour. During primary eclipse the secondary star makes a relatively larger contribution to the combined light. However, the nebular contamination also increases relative to the primary. Since the blue nebula contributes more light than the faint, red secondary, the light curve becomes bluer in the primary eclipse.

Another common feature in the colour-index curves of all 4 runs (data from 1994 in Fig. 3; from 1992 and 1993 in Fig. 2 of Paper III) is the small asymmetry in the bluening of the primary minimum, with the first half of the minimum being slightly redder than the second half. The effect was considered too small relative to the observational errors in Paper III, but it is real and present in all colours, even in y and becoming stronger at shorter wavelengths. Since this asymmetry is present in our observations from 1989 to 1994, it is stable and related to the orbital phase, probably connected to regions of unequal brightness (e.g. spots due to slight but continued accretion) on or near the surface of the eclipsing components.

Finally, the scatter in the 1994 colour curves appears to be significantly reduced during secondary eclipse (see the b-y and v-y colour curves in Fig. 3). A possible explanation is that there is a reduced contribution from surface activity on the secondary star when it is eclipsed. However, as noted above, such activity is unable account for the bulk of the observed non-eclipse variability.