There have been many measurements of the optical properties of V . Although in the 1960s they were generally obtained on bulk samples mechanically polished, in later years films were preferentially used. The optical properties of the V films are strongly dependent on the preparation technique and the physical-chemical conditions (in particular, deposition rate and substrate temperature). To some degree this can explain scatter among the data. Moreover, V is particularly sensitive to oxide, but no author has quantitatively estimated the oxide-coating effect on the optical response. In any case, the best films give a response very similar to the bulk.The room-temperature values of n and k tabulated here (Table I and Fig. 1) were obtained from the following works and references therein: Sonntag et al. [7] from 300-120 eV, Wehenkel and Gauth6 [13] from 120-40 eV; Seignac and Robin [8] from 40-6 eV; Nestell and Christy [19] from 6-0.5eV; Weaver et al.[12] from 0.5-0.1eV. As a general rule we preferred data that were obtained "directly," for example, from R and T by inverting the Fresnel relations or from ellipsometry, which do not rely on extrapolations outside the measured spectrum. However, we also considered values derived from Kramers-Kronig analysis of reflectance [12] or electron-energy-loss spectra [13], based on a very wide energy range and accurate measurements: This ensured reliable results, which were in good agreement with the others in the overlapping regions. We have disregarded data that were insufficient and ambiguous, in terms of either the sample preparation and characterization or the measurement accuracy, especially when inexplicable or in clear disagreement with other researchers. This is particularly true in the IR region, where despite numerous ellipsometric measurements, we preferred the results of calorimetry [12], although limited to 0.1 eV.