There is a wonderful scene near the end of Frank Baum’s The Wizard of Oz wherein the curtain hiding the wizard is suddenly whipped back and the wizard is exposed as a mere mortal, pushing buttons to make things happen, mostly through illusion. I often think my library users would have the same shock as Dorothy and her friends if they took the time to find out how we go about providing them with information. They’d whip back the curtain and find me madly pushing keys on the terminal, shouting “Ignore that woman over there!” This talk is supposed to be on the use of computers in astronomy libraries or at least in one astronomy library — and it is, but it also about providing an information service when you have only a modicum of information and a great demand for service. Since the Institute library has been in existence for only five years and because of the enormously transitory nature of much of astronomy publishing, we have remarkably little available in-house, particularly of the historical material. That’s where the creativity of a wizard (or the computer of the librarian) comes in. We may not have a specific item or we may not have the item exactly as requested, but we know who does and we use any and every means possible to obtain the item quickly. We concentrate on providing the specific information needed even if the user is unaware that what was requested may be available in ways other than that suggested in the request. (My favorite example of this phenomenon is the one pointed out by D.A. Kemp of a paper that appeared in five different series plus twice in report form.)