We used just the HG system (Bowell et al. 1989) to characterize the phase darkening behaviour of the nucleus and its absolute magnitude, H_R(1,1,0). Our best-fit values for both parameters are: G = 0.11 ± 0.12 and H_R(1,1,0) = 15.31 ± 0.07 which are determined from chi-squared fitting, results of which are shown in the left panel along with the 1–3σ co-variance ellipses. This fit is very satisfactory with very little scatter, as illustrated on the right panel. We point out that our NTT data from 10th May, 2005 (α = 0.1°) has been excluded from the analysis above due to a suspected calibration error of unknown source leading to magnitudes 0.35 mag fainter than might be expected. This has no bearing on our shape modelling analysis as we use relative lightcurves. While not included in our fit, we overplot the HST data (Lamy et al. 2006) and the VLT data (Tubiana 2008) for comparison. See Sect. 3.2 for more discussion.