Dear colleagues,
Jack Martin sent me a lot of spectra which did fill some gaps of the year 2015 and 2016 (see the attached monitoring). As we can see, the variability of the peak height of the V component becomes more and more smaller the more the moment of the eclipse is approaching.
I think that in the current phase (shortly before the eclipse and in the line of sight) the large and expanded Halpha emission-forming area (where the scattering will be produced) more and more is diving into the semi-transparent atmosphere of the M supergiant, and therefore the divergence of the peak height variability becomes smaller and smaller.
The divergence should become almost zero (ideal) and the peak height should remain stable, if the moment of the eclipse is reached.
Ernst Pollmann