The elimination of the residual image problem in the Mariner 10 vidicons allowed photometry of moderately high accuracy to be carried out during the 5 February 1974 flyby of Venus. Photometry of celestial objects allowed in-flight verification of camera linearity, shading and absolute photometric calibration. Due to the rapid rotation of the upper atmosphere the planet exhibits temporal brightness variations in the UV greater than 10% over a few hours. The observed terminator is 4° past the geometric terminator because of the detached haze layers at altitudes around 85 km. No indications of local cloud-top elevation variations greater than a few hundred meters were found in the Mariner 10 data. The cloud tops are probably diffuse. The distribution of brightness across the planet at 23° phase angle is described better by a cloud of isotropic scatters than by Mie spheres. In the UV the bright and dark regions both have low albedos in all scales, showing that the UV absorber is not just confined to the dark markings. Correlated contrasts can be seen on frames taken through the ultraviolet (UV), blue and orange filters. The outlines of the UV markings are diffuse even at the highest resolution; the contrast gradients are characterized by lengths of the order of 10–15 km. Regions which are brighter in the UV also have higher polarizations than darker areas. A cloud model which is consistent with the Mariner 10 observations has a third type of particle present in the clouds in addition to atmospheric gas molecules and sulfuric acid droplets. This particle is UV-absorbing, scatters nearly isotropically and is weakly polarizing; elemental sulfur has these properties. The UV absorbers are well-mixed vertically but incompletely mixed horizontally, thus causing the UV markings. The UV markings apparently represent stratospheric rather than tropospheric processes. There is little in the Mariner 10 pictures to suggest evaporation or condensation processes or strong horizontal wind shears.