The crystal-violet/tin-oxide solar cell which has been reported briefly in a recent patent was tested for u.v. stability under a variety of chemical conditions, and, in every case, was found to be unstable. Degradation effects produced by chemical reactions are strongly indicated by absorption and action spectra as the pH of the cell was varied. The p-n junction model is rejected for this type of crystal-violet/tin-oxide heterojunction cell. The potential of organic solar cells based on a dye-oxide semiconductor junction for cheap, lightweight and large-area devices is discussed, and some of the possible materials are reviewed.