More on the narrowing of impact broadened radio recombination lines at high principal quantum number

Recently Alexander and Gulyaev have suggested that the apparent decrease in impact broadening of radio recombination lines seen at high principal quantum number n may be a product of the data reduction process, possibly resulting from the presence of noise on the telescope spectra that is not present on the calculated comparison spectra. This is an interesting proposal. However, there are serious problems with their analysis that need to be pointed out. Perhaps the most important of these is the fact that for principal quantum numbers below n=200, where the widths are not in question, their processed generated profile widths do not fit the widths of the processed lines obtained at the telescope. After processing, the halfwidths of the generated and telescope profiles must agree below n=200 if we are to believe that the processed generated linewidths above n=200 are meaningful. Theirs do not. Furthermore, we find that after applying the linewidth reduction factors found by Alexander and Gulyaev for their noise added profiles to our generated profiles to simulate their noise adding effect, the processed widths we obtain still do not come close to explaining the narrowing seen in the telescope lines for n values in the range 200<n<250. It is concluded that what is needed to solve this mystery is a completely new approach using a different observing technique instead of simply a further manipulation of the frequency-switched data.