The $\alpha^4$ contribution to the lepton $g-2$ from a gauge-invariant set of 18 Feynman diagrams containing a light-by-light scattering subdiagram internally has been reevaluated by a method independent of the previous approach. Comparison of two methods revealed a program error in the first version. Correcting this error, the contributions of these 18 diagrams become -0.990 72 (10)$(\alpha/\pi)^4$ and -4.432 43 (58)$(\alpha/\pi)^4$ for the electron and muon $g-2$, respectively. The correction is not large enough to affect the comparison between theory and experiment for the muon $g-2$, but it does alter the inferred value for the fine structure constant $\alpha^{ -1}$ by 6 ppb.
Categories: hep-ph