College and Research Libraries can be no question of the need for such a volume, and the cooperative approach pro- vided by the ALA Statistics Coordinating Project should encourage general acceptance and use. Hopefully, as noted by Frank Schick in the foreword, "it may well prove the basis for an international standard for library statistics .... " The volume is organized with an opening chapter on General Concepts, followed by chapters on Statistics of College and Univer- sity Libraries, Statistics of State Library Agencies, Statistics of School Libraries, Sta- ~ tistics of Special Libraries, and Statistics of Library Education. Each is written by a specialist in the field. The chapter on Statistics of College and University Libraries was prepared by Mari- etta Chicorel, whose interest in this field is further represented by an article in CRL for January, 1966 (Marietta Chicorel, "Sta- tistics and Standards for College and Uni- versity Libraries," C RL, XXVII [January 1966], 19-22). It is suggested that a reading of this article will provide a background for understanding some of the recommendations made. While there may be some disagree- ment over decisions reached, for instance in the matter of using the physical volume rather than the bibliographical unit as the basis for count, we are at least provided with a clear statement on this and other items normally asked for in statistical re- ports. There also seems to be reasonable consistency in the definitions and principles among the chapters on Public Libraries, College and University Libraries, and Spe- cial Libraries. A glossary of terms is provided and is generally based on the ALA Glossary of Library Terms. In order that the volume be representa- tive of a broad spectrum, arrangements were made for a series of four regional confer- ences involving more than one hundred and sixty librarians. My only quarrel with the accuracy of the volume came for obvious reasons on page 148 where I found Mildred C. Langner, medical librarian of the Uni- versity of Miami, identified in a similar ca- pacity with the University of Mississippi! Joel Williams, director of the Statistics Coordinating Project, his staff and the Ad- visory Committee are to be congratulated Book Reviews I 285 upon the successful preparation of a very useful volume.-Archie McNeal, University of Miami. Computer Filing of Index, Bibliographic, and Catalog Entries. By Theodore C. Hines and Jessica L. Harris. Newark: Bro-Dart Foundation, 1966. ix + 126 pp. ( 66-23484) . The purpose of this book is to develop a code for computer or hand filing of li- brary, catalog, bibliographic, or index en- tries in a divided arrangement wherein authors, titles, and subjects are to be in separate alphabets. The authors assume "that filing should be a purely mechanical routine of handling entries whose written form actually determines their relative posi- tions." In other words, the cataloger or in- dexer prepares the entry for filing, and the computer does a simple, mechanical sort- ing into alphabetical order. However, the computer has far greater capabilities for filing than the mere ability to sort alphabetically catalog or index en- tries manually prepared for such a sorting. Use of a computer as a mere sorting ma- chine wastes much of its power, for the computer can relieve the cataloger or in- dexer of most of the work involved in set- ting up the entry for filing. For instance, the authors recommend that catalogers and in- dexers omit initial articles in the nominative case from title entries, but it is perfectly feasible to have the computer ignore these articles in mechanically preparing entries for filing; the article appears in the printed product, but was ignored in the alphabeting procedure. It is customary in sorting records with a computer to have the computer edit the category in the record under which the record is to be filed. The computer is in- sh·ucted to edit appropriate characters and set them up in a special sort field. The sort program then operates on this field. In setting up sort-field characters, the computer can alter original data in any way desired, providing that each character is always changed with the same algorithm. The algorithm may be quite complex and relate a given character with other elements so that in one circumstance it may be edited in one way and in another circumstance in 286 I College & Research Libraries • July, 1967 l quite a different manner. For instance, a A Plan for Indexing the Periodical Liter- ' diaeresis over an "o" in an English language ature of Nursing. By Vern Pings. With title would be dropped out of the sort con- an introductory chapter by Ellwynne M. trol, but when the computer detected a Vreeland. New York: The American diaeresis and on checking the language in- Nurses' Foundation Inc. 1966. xii, 202p. dication field found that the title was in ( 66-29223) . German, it would then place an "e" after the character over which the diaeresis oc- curred. With relatively few exceptions-per- haps one half of one per cent of entries-a computer can arrange bibliographic entries according to present library filing systems without human intervention. In part, the proposed code does not suc- ceed because it is not viewed as a segment in a comprehensive library system. It is only with the design of a machine-readable cataloging record to serve throughout li- braries that work should be undertaken on a machineable filing code. Even then characteristics of the machine must not be allowed to impose themselves on the code; rather, the objectives of the code should be firmly established and then the machine invoked to meet those objectives. Of course, the biggest obstacle to constructing an ef- fective new filing code-either machineable or manual-is that there are no adequate data to define the objectives of a filing ar- rangement. Much research needs to be done to attain understanding of how users use catalogs and indexes before thought should be given to construction, much less ac- ceptance, of a new filing code. The sum total of the book, however, is to propose a new filing code differing from existing codes. The differences are not great, and no evidence is presented to justify changes made from the present code. It seems unwise to invoke such a change without a clear demonstration of its bene- fit. Indeed, the proposed change involves greater human intervention in filing than would the computerization of present filing practices. Any increase in human interven- tion, such as manually preparing an entry for filing, diminishes the advantage of the machine. The code proposed in this book has been needlessly subjected to unneces- sary machine restriction, and can be con- sidered only as a departure from the old position; it is not a start in the right di- rection.-Frederick G. Kilgour, Yale Uni- versity. The proclaimed need for serious research in the library field together with the criti- cism now leveled at the superficial studies which have been dignified by that title are symptoms of our growing professional so- phistication. A research report such as this done in 1964 by Vern Pings for the Amer- r _ ican Nurses' Foundation proves that the J "working" librarian is capable of analytic examination of his field. The papers which make up the bulk of this report begin with a study of the growth of nursing as a pro- fession in terms of its formal communica- tion needs, continuing to a detailed study of the characteristics of the articles on nursing and their present bibliographical control, culminating in a formal series of plans for an index to this subject field. Despite the specificity of the subject in- vestigated, Dr. Pings' volume is of general interest to reference librarians; library edu- cators can use it as a teaching model, and it will serve other librarians contemplating similar studies as a planning guide. Each paper is formally organized, be- ginning with statements of hypotheses, de- scription of study methods , findings , con- clusions, and summary; extensive tables and appendices reinforce or demonstrate the points made. Chapters 3 and 4 which con- tain the analysis of the MEDLARS (Medi- cal Literature Analysis and Reb·ieval Sys- tem) coverage of nursing give a clear ac- count of MeSH (Medical Subject Head- ings) and the problems and inconsistencies which can frustrate the unwary user of Index Medicus. Conventionally, each chap- ter is accompanied by a good bibliography; the extensive one on nursing libraries in Chapter 9 of over three hundred items cov- ering the years 1903-1963 is especially im- pressive. The plans proposed by Dr. Pings for an index with broad geographic coverage and special subject headings but based on the already available MEDLARS foundation were closely followed in the new Interna- , I I