organization of the guide, based upon the structure of the government at the time of publication, was able to encompass all accessioned records of all agencies, past and present. Although it was acknowledged that the guide would be out of date even before it would be published, it was felt that necessary revision and updating could be accomplished with a minimum of difficulty while preserving the overall organization. To further aid in general orientation, the introduction to the guide is followed by a graphic presentation of the inclusive dates of record groups - the "time line" for each record group is given on a chart covering the period 1790-1970. The appendixes include a list of record groups by record group number, with the page number to the guide proper, and in the lengthy index the titles of record groups that constituted agency names are printed in bold capital letters.

In some 880 pages, the guide describes nearly 1 million cubic feet of records, including about 1.5 million maps and charts; some 201,000 accessioned rolls of microfilm of agency records; nearly seven million photographs, including 2.4 million aerial photographs; and approximately 66,500 sound recordings. These holdings as of mid-1970 did not include a sufficient number of machine-readable (electronic) records to justify a separate tabulation. The initial plan was to revise, update and republish the guide at regular-from 4 to 8 years-intervals, but a succession of budget cuts prevented implementation of these plans.

While the guide was still being compiled, automated techniques were being introduced to facilitate the establishment of administrative or physical and intellectual control over NARS' holdings. The first experiments involved the use of selective permuted indexing (SPINDEX) at the file unit or folder level and for the preparation of indexes to finding aids, and for the preparation of indexes to the contents of the

records of the Continental Congresses. The next step was the development of a system (NARS-A1) involving a computer-assisted procedure for compiling all series-level inventories into a master file. This system was intended to improve administrative control over the holdings and to replace the many variations that had developed in describing series with a machine-readable file featuring a standardized format and a hierarchical addressing system that would preserve the organizational content of the records. Because of costs, no effort was made to design a system for information retrieval by subject. Such a system would have required new descriptions of all records after subject terms had been agreed upon, and the decision was made to develop instead a computer assisted program for text editing, a process now widely referred to as word processing. Since the system was to be computer assisted rather than computer centered, it employed batch processing rather than an on-line mode of operation. Other features of the system included fixed record lengths and fixed fields for descriptive information.

From the outset the system was handicapped with problems of obtaining computer services and with having to use systems analysts who had to design a system reflecting the capabilities of available equipment rather than agency needs. The overall cost of the system was relatively high when finally implemented. To have attempted to create a fully automated system with on-line retrieval by subject index terms would have been prohibitively expensive, and although this option is now more attractive because of the great advances that have been made in computer hardware and software during the past decade, the major obstacles still remain the time, the manpower, and the cost of not only converting existing series descriptions into machinereadable form, but of redescribing the records in terms of specific subject content.

Recent applications research projects undertaken by NARA involving digital image conversion, optical character recognition, and the use of function terminology in records description give promise of reducing, if not totally eliminating, these obstacles. The development of plans for a life cycle tracking system for accessioned records, and experiments with the MARC AMC format, will also influence the form and content of future finding aids.

Apart from the increasing influence of information-handling technology, a number of other developments will necessarily help determine NARA's future records description program. Since initial publication of the 1974 Guide, the total volume of the research resources in the custody or under the jurisdiction of NARA has increased from less than 1 million cubic feet to more than 1.5 million cubic feet, and this does not include the major holdings of the still-growing Presidential libraries system. The volume and variety of electronic records continues to grow rapidly, and accessioning by NARA will have to be accelerated if those of archival value are to be preserved and made accessable for use. It is thus no longer realistic to consider describing this mass of material in a single volume and in any degree of detail useful in assisting scholarly research.

A parallel development has been the changing character of research interests in the field of history. Traditional areas such as diplomatic, military, and institutional or administrative history continue to attract researchers, but the introduction of new methodologies based upon information

handling technologies have revolutionized the fields of social, economic, and political history and are continuing to influence research methodologies in other fields. Blacks, women, ethnic groups, peace, science, and technology - all have attracted students and scholars and are creating not only a demand for different types of records than those traditionally accessioned as archival, but are leading to a reexamination of the content of existing holdings in the light of these new interests. The growth of public history and the establishment of historical offices by federal agencies, many of them engaged in writing general administrative history and histories of particular functions, programs, and projects, will also influence the finding aid program of NARA. Finally, disciplines other than history are increasingly recognizing that government archives-whether high-level correspondence or economic or technological data bases-have widely varied and important research value.

Because scholarly research adds to the sum total of human knowledge and because archives are increasingly becoming a primary source for documentary research in all disciplines, NARA will continue its efforts to meet the needs of scholarship involving retrospective documentary research, whether by historians or by researchers in other disciplines. This singlevolume guide to the National Archives of the United States may be regarded as marking the end of one era and the transition to another in the achievement of that objective.

FRANK B. EVANS