Wide-Angle Electron-Pair Production

An experiment testing quantum electrodynamics at high energies and small distances is described. The photoproduction from carbon of electron-positron pairs was measured at laboratory angles of 4.60°, 6.23°, and 7.46°. Symmetrical electron-positron pairs in the energy range from 1 to 5 BeV were detected with a magnet-counter system which consisted of two mirror-image arms. Extensive internal checks of the apparatus were made and the results were reproducible. The theoretical values for the electron-pair yield were calculated by integrating the differential pair-production cross section over the acceptance of the apparatus using a Monte Carlo technique. The ratio $R=\left(\mathrm{experimental}\mathrm{yield}\right)/\left(\mathrm{theoretical}\mathrm{yield}\right)$ was not 1.0. $R$ was approximately given by $R=0.62\left\{\right(1.00\mathrm{}\pm 0.05)+\frac{k^2}{{(4.31\mathrm{}\pm 0.17)}^2}\}$ where $k$ is the energy in BeV of the photon which produced the pair, and by $R=0.67\left\{\right(1.00\mathrm{}\pm 0.04)-\frac{{Q_f}^2}{{(313\mathrm{}\pm 13)}^2}\},$ where ${Q_f}^2$ is the four-momentum of the virtual fermion in ${\mathrm{}\left(\mathrm{MeV}\right)\mathrm{}}^2$. The apparatus studies and a comparison of the measured single-electron yields with the theoretical yields suggest that an error exists in the absolute normalization of the results. There are no indications that the observed variation of the electron-pair yields with momentum or the large excess of wide-angle electron pairs at high energies is due to any systematic error. The experimental results do not agree with the predictions of quantum electrodynamics; they indicate a breakdown of the theory or the presence of other processes.