Comment on the paper "Dark matter searches going bananas: the contribution of Potassium (and Chlorine) to the 3.5 keV line" by T. Jeltema and S. Profumo
We revisit the X-ray spectrum of the central 14' of the Andromeda galaxy, discussed in our previous work [1402.4119]. Recently in [1408.1699] it was claimed that if one limits the analysis of the data to the interval 3-4 keV, the significance of the detection of the line at 3.53 keV drops below 2 sigma. In this note we show that such a restriction is not justified, as the continuum is well-modeled as a power law up to 8 keV, and parameters of the background model are well constrained over this larger interval of energies. This allows for a detection of the line at 3.53 keV with a statistical significance greater than ~3 sigma and for the identification of several known atomic lines in the energy range 3-4 keV. Limiting the analysis to the 3-4 keV interval results in increased uncertainty, thus decreasing the significance of the detection. We also argue that, with the M31 data included, a consistent interpretation of the 3.53 keV line as an atomic line of K XVIII in all studied objects is problematic.