In 1985 the VEGA 1 and VEGA 2 spacecraft dropped two descent probes into the nightside of Venus. On board was the French-Russian ISAV ultraviolet spectroscopy experiment, consisting of a UV light source absorbed by atmospheric constituents circulating freely into a tube attached outside the pressurized modules. ISAV generated a wealth of absorption spectra in the 220- to 400-nm range with an unprecedented vertical resolution (60--170 m) from 62 km of altitude down to the ground. On the basis of known instrument properties and a careful examination of the light curves recorded in 13 wavelength intervals in the UV, we show that most of the recorded differential absorption (at each wavelength with respect to 394 nm) can be explained by a combination of gaseous SO2 absorption and absorption by aerosols deposited on the mirrors during the crossing of Venus' lower cloud. The spectral signature of this absorber, termed X, was obtained, thanks to an unexpected shock on VEGA 1 which removed this absorber from the mirrors at 18 km of altitude. The UV spectral signature of X resembles that of croconic acid, C5O5H2, whose absorbing power as a contaminant of H2SO4 droplets at 2.5% dilution is compatible with the observations. However, the nonidentity of the spectral signature, together with stability arguments, makes this identification less plausible. Whatever its nature, the relevance of this new absorber X is discussed in connection with the albedo of Venus and the IR variable leak windows. If this absorber X detected by ISAV in the lower cloud were also present in the upper cloud, it would be a good candidate to explain the UV part (λ<400 nm) of the Venus albedo. Three layers of absorbing material, called b, c, and d, are identified in the data of both ISAV 1 and 2 in the altitude range 49--43 km. The higher layer b is inside the lower cloud identified by the nephelometer of Pioneer Venus, while the two other layers are well below the lower cloud boundary as measured by Pioneer Venus. The SO2 profile (from 60 km down to 10 km) is characterized for ISAV 1 by a double peak of the mixing ratio (150 ppmv at 51.5 km, 125 ppmv at 42.5 km) separated by a deep trough at 30 ppmv at 45.6 km. For ISAV 2 there is a single peak at 43 km. Both SO2 profiles are quite compatible with recent ground-based measurements, showing 130¿40 ppmv in the altitude range of 35--45 km <B¿zard et al., 1993>. Below the clouds the measured SO2 mixing ratio decreases steadily on both probes, down to 25¿2 ppmv at 10 km for ISAV 1, which is lower than previously reported values from gas chromatograph measurements (shown to be incompatible with ISAV measurements). The variation of SO2 mixing ratio with altitude implies a strong vertical transport which is given as a function of altitude, showing the source and sink regions of SO2 from 10 to 60 km of altitude. These data should impose severe constraints on future chemical models of the atmosphere of Venus, occurring after volcanic episodes or impact cratering events. The total SO2 column density (0--60 km) was measured to be 4¿1022 molecules cm-2 or 4.2 g cm-2, a factor of 5 below previous estimates. With an average reaction rate of 4.6¿1010 molecules cm-2 s-1 with calcite, CaCO3, estimated by Fegley and Prinn <1989>, it would take only 27,000 years to get rid of SO2 and associated H2SO4 droplets if calcite were present all over the surface of Venus. Therefore SO2 and its associated ubiquitous cloud cover may only be transient phenomena in the life of Venus. ¿ American Geophysical Union 1996