College and Research Libraries to one year. Still the author concludes that "the overall reviewing · pattern in all disci- plines has seemed to remain relatively con- stant over the past couple of decades" (p.123). We can possibly infer that the minor inconsistencies in data gathering did not materially affect the results. More useful and impressive than a rigid pattern of data gathering would have been a technique developed for easy, direct com- parison of the analogous findings from each study with the other and with a cpmposite profile. In spite of the numerous tables within the text, plus appendixes at the close, readers are forced to search out and compare for themselves (no easy task in view of the constantly changing presenta- tions of both text and tables). The final chapter of general conclusions does give a composite time-lag table, al- though none of the other summaries of findings is similarly documented. This vol- ume adds evidence to the impressions of many observers that, while the burden of book reviewing is carried by a comparative- ly small percentage of sci-tech journals, reliable book reviews for most significant general biomedical and general science books do appear in a year or less, frequent- ly in more than one source. The situation in special disciplines, we are told, is less re- liable, with astronomy and engineering titles being the most poorly reviewed.- Jeanne Osborn, School of Lihrary Science, The University of Iowa. Computer-Readable Bibliographic Data Bases: A Directory and Data Sourcebook. Compiled and edited by Martha E. Wil- liams and Sandra H. Rouse, Information Retrieval Research Laboratory, Coordi- nated Science Laboratory, University of Illinois. Washington, D.C.: American So- ciety for Information Science, 1976- Loose-leaf. $68.00 per year. LC 76- 46249. ISBN 0-87715-114-8. The rapid integration and growth of on- line bibliographic services in all types of libraries during the past four years has de- manded a single reference source of the available machine-readable data bases. Li- brarians and end users alike have required an up-to-date compilation of the publicly available data bases outlining their subject scope, format, acronym, full name, produc- Recent Publications I 261 er, and accessibility. This ambitious and urgently needed task has been accom- plished by Martha E. Williams and Sandra H. Rouse in their massive work, Computer- Readable Bibliographic Data Bases: A Di- rectory and Data Sourcebook. Compared with the 1973 ASIS publication, Survey of Commercially Available Computer-Read- able Bibliographic Data Bases, which pro- vided information on eighty-one data bases, the new directory expands coverage to 301 data bases and includes more pertinent de- tails about the files. Anticipating the dynamic nature of data bases which frequently change their name, format, ownership, or availability on vari- ous systems, the directory appears in a loose-leaf binder with update pages being promised at six-month intervals. Ranging from ABIPC (the Abstract Bulletin of the Institute of Paper Chemistry) to Zoological Record, the data base outlines follow a con- sistent format, the definition and method- ology of which are well described in the "Introduction." Approximately 58 percent of the data bases are produced by U.S. or- ganizations and about 41 percent by for- eign organizations. Although some of the minimal data base descriptions are not fully detailed, the majority of the file entries in- clude a great deal of information: acronym and complete name, issuance, correspon- dence with printed source, producer, dis- tributor and/ or generator, subject matter and scope, indexing, tape specifications, data base services, and user aids if offered by the producer. Of particular assistance is the name, address, and telephone num- ber of the person to con tact for further information about the data base. Included in the data base services category are the centers which process the file in either the on-line or batch mode; deliberately exclud- ed are centers which provide in-house ser- vice only or brokers which provide search services by remotely accessing other com- puter facilities. The directory is augmented by four indexes-broad subject categories, data base name, producer, and processor. In order to retain its usefulness, rigorous updating will be necessary. Uaving been published in October 1976 before the Jan- uary 1977 availability of the Bibliographic Retrieval Services, Inc., BRS system and before the actual loading of several pre- 262 I College & Research Libraries • May 1977 viously announced files of System Develop- ment Corporation's ORBIT and Lockheed Information System's DIALOG systems, the directory already needs several update modifications reflecting vendor additions. Overall, this work is recommended for any library which provides an active search service capability. To ensure its viability, it is urged that after the original one-year purchase, updated and additional pages be provided by ASIS at a nominal fee to the original subscribers.-Patricia E. Vaughan, Coordinator, NASIC, New England Board of Higher Education, Wellesley, Massa- chusetts. Pitkin, Gary M. Serials Automation in the United States: A Bibliographic History. Nletuchen, N.J.: Scarecrow, 1976. 148p. $6.00. LC 76-18116. ISBN 0-8108-0955- 9. According to the introduction, "the pur- pose of this book is to provide, in one place, access to all information published on the automation of serials control functions in the United States and cited in Library Lit- erature." The book is an annotated, biblio- graphic history, covering the period from 1949 to 197 4. Each entry, numbered and in chronological order, is annotated and contains complete bibliographic informa- tion. In addition, each entry briefly notes the major automation application, such as ordering, claiming, binding, etc., and the type of library, including the categories of academic, government, industry, medical, military, public, and secondary schools. The annotations are well-written and in- formative, varying in length from one to four paragraphs, with ample quotes from the original item. The appendixes contain an index of the articles by serials control function (binding, claiming, holdings in- formation, etc.), and there is also a com- plete author index. This book will help anyone considering serials automation by providing concise in- formation on the way in which other li- braries faced the challenges of serials automation and by locating specific articles relevant to the library's particular automa- tion needs. Its usefulness is diminished by its 197 4 closing date for entries and the limitation to the United States. Admittedly, most libraries contemplating the automation of their serials control func- tions will have Library Literature at their disposal. However, this book, reasonably priced at $6.00, provides a reliable, anno- tated literature survey and is highly recom- mended for any library or organization involved with serials automation.-]ean Hawks, Director of Public Services, North- ern Arizona University Libraries, Flagstaff. Sheehy, Eugene P. Guide to Reference Books. Compiled by Eugene P. Sheehy, with the assistance of Rita G. Keckeissen and Eileen Mcilvaine. 9th ed. Chicago: American Library Assn., 1976. 1015p. $30.00. LC 76-11751. ISBN 0-8389- 0205-7. It is difficult to consider doing a review of Kroeger/Mudge/Winchell/Sheehy with- out a feeling of reverence. Memories of the earlier editions from library school days on through years in the library profession bring proud recognition that the Guide has been a work of consistent excellence by li- brarians for the entire learning world. As one reads through the pages of this latest edition, there is the pleasure of recalling massive sets of volumes on shelves and of remembering authors whose careful work is always within reach and students whose puzzling questions have been answered. As noted in the preface, very little has been dropped from the previous edition and much has been added. This fact was confirmed by checking sections throughout the cumulated index of Supplement 3 of the eighth edition with the index of the ninth. The omission of reference works on individual authors appears to be the major deletion. One can sympathize with com- promises because of space requirements, but the omissions seem not to be done as consistently as would be desirable. Inclusions are said to be classical writers and "a few indisputably major authors" (Pre£. p.x). This decision results in rw list- ing for American authors (there were twelve in the eighth), and only Goethe is left for German writers-Hegel, Heine, Kafka, Mann, Schiller having been dropped. Corneille, Moliere, and Racine remain of fourteen French authors in the eighth. There is a satisfactory explanatory note as to the treatment of this type of reference work in the English literature section; a