The work functions of silicon single crystals have been obtained by measuring the contact potential differences between the crystals and a gold reference whose work function was measured by the Fowler method. The measurements were carried out in high vacuum (p<10−9mm Hg). For nonfloating‐zone silicon, the work function was higher for samples which were radiation quenched from 990°C than it was for samples which were annealed at 500°C. This effect was not observed for floating‐zone material and was presumably a function of occluded oxygen. The work functions for the different faces after ion‐bombardment cleaning were in the sequence (100)>(110)>(111). The values after quenching were: (100), 4.92 ev; (110), 4.89 ev; (111), 4.77 ev. The values after annealing were: (100), 4.82 ev; (110), 4.70 ev; (111), 4.67 ev. The work function of the (111) face of a 2000‐ohmcm,p‐type, floating‐zone sample was 4.73 ev after quenching or annealing. Changes in both surface and volume properties were observed when silicon was heated above 1000°C in contact with quartz. Oxygen exposures of 5×10−5mm Hg min resulted in work function increases which were larger on the annealed surfaces than on the quenched ones. This was observed for both floating‐zone and oxygen‐doped material and is therefore a property of the silicon itself. Photoelectric evidence indicated a diffusion of oxygen into the silicon (1) when silicon was heated in oxygen, (2) when silicon was heated at 990°C in high vacuum after O2adsorption, and (3) when silicon with residual surface films from chemical etching was heated at 990°C prior to ion bombardment. Hydrogen exposures of 7.5×10−4mm Hg min resulted in work‐function decreases only in the presence of a heated filament. Nitrogen exposures of 4×10−6mm Hg min resulted in no work‐function changes. Heating silicon in high vacuum at 990°C for about 100 hours developed thermal‐etch patterns which were rectangular on the (100) face, hexagonal on the (110) face, and triangular on the (111) face. Heating at 1080°C for 24 hours caused further etching which developed the same triangular patterns on the (111) face but left the (100) and (110) faces with a sandblasted appearance.