Abstract
It is shown that rescattering corrections to γp→π0p through the sequence γp→π+n→π0p, at threshold, and to γn→π0n through the sequence γn→πp→π0n, are large. The cusps at π+n and πp thresholds, kL=151.5 and 148.5 MeV, respectively, are inverted. It is also argued that an explanation of the discrepancy between experiment and theory through a Σ-like term is, most likely, erroneous.