Some implications of ultraviolet observations of quasars and active galaxiesThe problem of the order of magnitude discrepancy in the expected and observed ratios of the Lyman and Balmer lines in quasars and active galaxies is reviewed. Whereas early photoionization models for the emission line regions predicted F(L-alpha)/F(H-beta) not less than 40, the observations give values for this ratio in the range 3-8. Attempts at explaining the observations have involved dust, both external and internal to the emission line regions, and improved treatments of the collisional processes and radiative transfer effects in dense (Ne about 10 to the 10th/cu cm), optically thick clouds. None of the effects considered is, by itself, able to explain all the observations, and a combination of several of them is probably required.