Examples are given of the kinds of machine-readable data bases that should be developed in order to extend attempts at universal bibliographical control into neglected areas, the results of which can be used by researchers in the humanities, specifically those using books printed before 1801. The principles of bibliographical description, organization, and retrieval necessary for the design of such data bases and their control systems are explored and clarified. Theoretical considerations are based upon conceptual development from Dr. Ranganathan's writings and upon the author's experience with the Hand Printed Books Project. (Author)