In 1971, D. A. Shirley conducted high-resolution x-ray photoemission spectroscopy on the valence bands of gold, using monochromatized Al Kα radiation and a single-crystal specimen. After background and scattering corrections, the results were compared with broadened theoretical density-of-states functions. The study concluded that relativistic band structure calculations are necessary to fit the spectrum. Both the KKR calculation of Connolly and Johnson and the RAPW calculation by Christensen and Seraphin gave density of states results that closely followed the experimental curve after broadening. Full Slater exchange was required, as fractional exchange gave too wide d band densities of states. Eastman's 40.8 eV ultraviolet photoemission spectrum was similar to the x-ray spectrum, suggesting little dependence on photon energy above 40 eV. Both conclusions imply an absence of strong matrix-element modulation in the photoemission spectrum. The experiment involved a gold single crystal of 99.9% purity, polished, etched, and annealed. Spectra were taken at room temperature and in a sample chamber pumped to about 10^-7 Torr on a Hewlett-Packard ESCA spectrometer. The spectrum showed a definite improvement over earlier spectra. The active sample was predominantly gold, as demonstrated by monitoring the Au N VI, VII doublet. Oxygen and carbon lines were barely detectable, indicating a clean sample. Spectra from specimens in the three crystal orientations were very similar. A (110) spectrum, which had the best statistical accuracy, was analyzed for comparison with theory. The data reduction procedure consisted of three steps: smoothing, background subtraction, and correction for inelastic scattering. The final corrected spectrum, IS'', was compared with theoretical density-of-states functions. Non-relativistic calculations gave density-of-states histograms differing so much from IS'' that plotting them was not useful. The two-peak form of IS'' agreed well with relativistic calculations, indicating that relativistic effects must be considered in discussing the band structure of gold. Six theoretical results were available for the relativistic band structure of gold. Sommers and Amar used the KKR approach but did not give ρ(E), so a direct comparison with IS'' was not possible. Their results gave a valence band width considerably larger than the experimental width. Christensen and Seraphin attributed this to the use of 2/3 Slater exchange by Sommers and Amar. Connolly and Johnson used full Slater exchange and obtained a smaller valence band width that agreed well with experiment. Three RAPW calculations were reported, with Ramchandani using full, 5/6, and 2/3 Slater exchange. His full-exchange results were in reasonable agreement with IS'', while the other two were not. Kupratakuln used 19/24 exchange and concluded that the best results were obtained with thisIn 1971, D. A. Shirley conducted high-resolution x-ray photoemission spectroscopy on the valence bands of gold, using monochromatized Al Kα radiation and a single-crystal specimen. After background and scattering corrections, the results were compared with broadened theoretical density-of-states functions. The study concluded that relativistic band structure calculations are necessary to fit the spectrum. Both the KKR calculation of Connolly and Johnson and the RAPW calculation by Christensen and Seraphin gave density of states results that closely followed the experimental curve after broadening. Full Slater exchange was required, as fractional exchange gave too wide d band densities of states. Eastman's 40.8 eV ultraviolet photoemission spectrum was similar to the x-ray spectrum, suggesting little dependence on photon energy above 40 eV. Both conclusions imply an absence of strong matrix-element modulation in the photoemission spectrum. The experiment involved a gold single crystal of 99.9% purity, polished, etched, and annealed. Spectra were taken at room temperature and in a sample chamber pumped to about 10^-7 Torr on a Hewlett-Packard ESCA spectrometer. The spectrum showed a definite improvement over earlier spectra. The active sample was predominantly gold, as demonstrated by monitoring the Au N VI, VII doublet. Oxygen and carbon lines were barely detectable, indicating a clean sample. Spectra from specimens in the three crystal orientations were very similar. A (110) spectrum, which had the best statistical accuracy, was analyzed for comparison with theory. The data reduction procedure consisted of three steps: smoothing, background subtraction, and correction for inelastic scattering. The final corrected spectrum, IS'', was compared with theoretical density-of-states functions. Non-relativistic calculations gave density-of-states histograms differing so much from IS'' that plotting them was not useful. The two-peak form of IS'' agreed well with relativistic calculations, indicating that relativistic effects must be considered in discussing the band structure of gold. Six theoretical results were available for the relativistic band structure of gold. Sommers and Amar used the KKR approach but did not give ρ(E), so a direct comparison with IS'' was not possible. Their results gave a valence band width considerably larger than the experimental width. Christensen and Seraphin attributed this to the use of 2/3 Slater exchange by Sommers and Amar. Connolly and Johnson used full Slater exchange and obtained a smaller valence band width that agreed well with experiment. Three RAPW calculations were reported, with Ramchandani using full, 5/6, and 2/3 Slater exchange. His full-exchange results were in reasonable agreement with IS'', while the other two were not. Kupratakuln used 19/24 exchange and concluded that the best results were obtained with this