Comment on "Comparison of Potential Models with the PP Scattering Data Below 350 Mev"

We point out two flaws in the recent test of nucleon-nucleon (NN) potentials conducted by Stoks and de Swart. First, in some cases, the neutron-proton ($np$) version of an NN potential was compared to the proton-proton ($pp$) data, which is improper and yields (large) $\chi^2$ that are essentially meaningless. Second, for a proper test of the quantitative nature of a NN potential, it is insufficient to compare to $pp$ data only, since this leaves the T=0 potential untested. Thus, it can happen that the $pp$ version of a potential predicts the $pp$ data accurately, while the $np$ version of that same potential is poor in $np$ (where also the T=0 potential is involved). An example for this is the Nijmegen potential, which predicts the $pp$ data well with a $\chi^2$/datum of 2.0, but yields a $\chi^2$/datum of 6.5 in $np$.