The logical structure of the data bank is described from the user's point of view, without reference to machine implementation. The length of files, records, segments, and fields are variable and unlimited; duplication factors are unlimited; sequence of files, records within file, and segments within record are random. All data are subject to unrestricted revision and expansion. Access is by directory, in the form of multiple inverted indices. While the data files are identical in structure, they differ in content. Five "types" of record are identified: (1) unique object description, (2) biography, (3) publications, (4) events and (5) places. A separate file, associated with its own set of inverted indices, is planned for each type. The description of the data bank structure and contents is followed by three appendices: Appendix 1: acronyms in use, June 1970. Appendix 2: definitions of annotation classes in use, June 1970. Appendix 3: model for a combined thesaurus and inverted index. (Author/NH)