College and Research Libraries RICHARD DE GENNARO Harvard University's Widener Library Shelflist Conversion and Publication Program This paper briefly reviews the essential background and reports on the progress, evolution and future outlook of this unique program. It then describes in some detail the present and potential uses that can be made of the shelflist of a major research library after it has been con- verted to machine readable form and can be manipulated by com- puter. The uses fall in.to two broad categories: publications and read- er services, and library management and statistical uses. It is a non- technical article for librarians and library users. INTRODUCTION · IN 1964 WIDENER LmRARY, the central research collection of Harvard Univer- sity, developed a system for converting its manuscript sheaf shelflists to machine r eadable form and embarked on a proj- ect to computerize the 1.6, million entries in the list. To date, more than 600,000 records from some of the most active classes in the library have been con- verted and used in various ways, and the project continues as an accepted and im- portant part of the library's automation operations. The project has now come of age; its feasibility and usefulness are firmly established and it seems appro- priate at this time to review briefly the essential background of the program and to report on its progress, evolution, and future outlook. The previous papers on the Widener At the time this paper was written, Dr. D e Gennaro was Senior Associate University Librarian, Harvard University. He is now Director of Libraries, University of Penn- sylvania. 318 / shelflist project were largely concerned with the strategy and the techniques of converting this large and unique fil e into machine readable form. This paper will stress the present and potential uses that can be made of the shelflist of a major research library after that conversion (or a large part of it) has been completed . Two main categories of uses will be dis- cussed: ( 1) the production of publica- tions of various kinds and the provision of other reader services; and ( 2) library management uses including the genera- tion of statistical and other data for fur- ther automation, for managerial pur- poses, and for general research. REVIEW OF THE PROJECT The justification for embarking on the ambitious project to convert the estimat- ed 1.6 million handwritten entries in the old loose-leaf sheaf shelflists in Widener Library can be found in an article en- titled "A Computer Produced Shelflist" which appeared in C RL in 1965.1 The project was placed in the larger context of the Harvard Library's overall automa- tion program in another article, "Auto- .... , Con version and Publication Program I 319 mation in the Harvard College Library" which was published in 1968. 2 A tech- nical description of the operation in its early stages was written by Foster M. Palmer in 1966.3 No detailed technical descriptions of the computer systems have been published since that time, al- though specific information can be ob- tained from internal working documents. The preparation for publication of tech- nical descriptions of a rapidly evolving system of a local nature is time-consum- ing and difficult to justify. This article will merely sketch in enough of the project's background to make it compre- hensible without reference to the earlier papers. No technical material will be in- cluded. A library shelflist is a record of the books arranged in the order in which they appear on the shelves. It is main- tained primarily as a tool for assigning new and unique numbers to books that are added to the collection and as an in- ventory record of the books in a library. Since the book collections in most Amer- ican libraries are arranged in classified order, the shelflist is potentially useful to scholars, particularly if it can be made available in convenient form and if clas- sification schedules and author and title indexes are provided. For most libraries the maintenance of a shelflist is a routine process and involves merely filing a copy of each main-entry card into the card shelflist in call-number order. However, the Widener shelflist, for historical rea- sons, is largely handwritten in loose-leaf volumes, rather than on cards, and is therefore difficult to use and maintain. In 1964 it became evident that, through the use of computer technology, the li- brary could modernize the shelflist main- tenance procedure and at the same time make an expanded version of the shelf- list available as an additional approach to the library's holdings. Accordingly, an experimental system was designed to con- vert the shelflist to machine readable form and, after a successful pilot project, a full-scale conversion and publication program was begun in 1965. The initial system was somewhat primitive, with input and output limited to the standard uppercase character set that was then commonly available on computer print chains. In June 1966 the system was improved so that the input could be coded with an expanded char- acter set to produce output with both upper- and lowercase letters and the re- quired diacritical marks. The output for the published volumes continued to be produced by photo offset from a com- puter printout until further improve- ments in the system made it possible, late in 1969, to produce ·graphic arts quality printer's copy in double columns by computerized photocomposition tech- niques. The evolution of the output for- mat is virtually complete; all further im- provements will be in the input, process- ing systems, and development of new products. To date, more than 600,000 entries of the total 1.6 million in the shelflist have been converted. Twenty-two volumes have been published in the Widener Li- brary Shelflist series and a dozen more are scheduled for publication in 1970. An estimated twenty-five to thirty ad- . ditional volumes will be required to com- plete the series. Several of the classes that were initially keypunched in the limited uppercase format have been con- verted by a combination of computer program and manual editing to the new standard expanded character set and format, while the rest will be completed by the end of the year. Thus, all the records in the system will soon be in a single uniform and compatible machine format. The master files are arranged in classified or shelflist sequence on magnetic tape. Widener call numbers are machine processable and, since the numbers are unique, they also serve as identification numbers for the machine records. The entries in the old manuscript 320 I College & Research Libraries • September 1970 shelflists are not bibliographically com- plete. They were limited to call num- ber, volume count, author, title, place, and date of publication. Frequently the author's forenames were not spelled out and the titles were shortened. Notes, added entries, and subject headings were not included. The strategy of the con- version project is to accept the entry es- sentially as it is with some few excep- tions; obvious errors are corrected, au- thors' full names are added when easily obtainable, abbreviations in titles are spelled out, and a language code and a code distinguishing serials from mono- graphs are added. All elements present are tagged so as to permit machine ma- nipulation. The average number of char- acters per record is 100, while full LC records are estimated at 350-450 char- acters. This enforced limitation on the quality and completeness of the records is unfortunate for many reasons, but it has made the conversion and publication projects economically and technically feasible. Had the shelflist contained complete bibliographical records, the project would not have been attempted, for various reasons. Since clean and accurate copies of the Widener classification schedules are a necessary prerequisite for the prepara- tion of the published shelflists, a major program was undertaken in 1966 to re- vise and edit the schedules. The sched- ules are being converted to machine readable form , and a computer program nsed to facilitate editing as well as to format them into the two distinct forms that are required for the published lists. All shelflist conversion and editorial work is done in the library with regular library funds by a staff of eight nonpro- fessionals. It has become a routine ac- tivity of the Data Processing Division and funds for the completion of the project within the next several years seem assured. The design and programming of the system has been accomplished entirely by librarians trained as systems analysts. The routine computer work has been done for the most part on an IBM 1401 which has 8,000 positions of core storage and four tape drives and is located in the library. In 1970 the 1401 will be phased out after the entire system has been re- designed and reprogrammed to run on an IBM 360-65 located in the U niver- sity's Computing Center. The system conversion will be done by the library's data processing staff. The occasion will be seized to convert the local shelflist system into a more permanent and stan- dardized system based on the MARC II format. When the present system was designed, the MARC II standard format for bibliographic entries in ma- chine readable form did not exist. That format has now been completed and widely accepted internationally, and programs are b eing written at several centers to manipulate bibliographic data in that format in various wavs and for various purposes. Although Harvard shelflist entries are not as complete as full LC MARC II en- tries, the elements that are present can be tagged and put into the format , and those that are not can be left blank. When the library develops a system to input its current cataloging in the MARC II fonnat, those entries can be integrated into the new shelflist system, since the machine format of the two kinds of en- tries will b e compatible even though they differ in the amount of data in- cluded. In the more distant future it is ex- pected that the present brief shelflist entries will be superseded by standard bibliographical records in MARC II for- mat. Given the growing interest in retro- spective conversion at the national level, it is reasonable to foresee that a central bibliographical agency will convert and distribute these entries and that Harvard may be able to substitute them for its own incomplete entries .4 But this is a distant and as yet uncertain possibility. I """" Conversion and Publication Program I 321 Meanwhile, Harvard will have realized a satisfactory return on its investment in converting its abbreviated shelflist en- tries. The nature and extent of that re- turn is the subject of the remainder of this paper. uSES OF THE MACIDNE READABLE DATA BASE The present and potential uses of the Widener shelflist data base fall into two broad categories. The one involves creat- ing and publishing new or special list- ings of the holdings of the library for the use of scholars, bibliographers, and li- brarians at Harvard and elsewhere. The prototype is the published shelflist se- ries; this series and its possible future variations will be discussed first. The other involves using the machine read- able data base to improve or facilitate certain library operations such as shelf- list maintenance, circulation control, col- lection building, and the generation of statistical and other information for man- agement and analysis purposes. Publications and Reader Services The publication of the Library's shelf- list was one of the principal justifications for converting the shelflist to machine- readable form. The rationale is stated succinctly in the preface to the pub- lished volumes: In the absence of a classified catalog, the shelflist has long been used by librarians and experienced library users as a means of systematically surveying the library's holdings in a particular subject. When pe- rusing the shelflist one sees all the titles that have been classified in a given area, and not merely those which happen to be on the shelves and whose spine lettering is legible. In addition, one can take in at a glance the essential bibliographical de- scription of a book-author, title, place and date of publication. However, the potential usefulness to readers of the Widener sheaf shelflist in manuscript form has never been realized because it existed in only one copy. Moreover, it was kept in a relatively inac- cessible area, was awkward to read and frequently difficult to interpret. Computer technology has made it possible to enlarge the concept and to expand the uses of the shelflist while improving the techniques of maintaining it and making it available to readers .... The development and publi- cation of the shelflist in this form is an at- tempt to equip the serious reader with a copy of the classification scheme that has been used to organize the collection, to- gether with lists in classified, alphabetical, and chronological order of the books and journals in each class. After each class and its corresponding classification schedule have been con- verted to machine readable form, a three-part catalog of the holdings in the class is published in the Widener Li- brary Shelflist series. The first part con- tains the classification schedule and a list of the entries in the class in call number (i.e., classification) sequence with subclass headings (derived by pro- gram from the machine readable classi- fication schedule) interspersed through- out the list. The second part is an alpha- betical listing by author and by title and is obtained by a programmed computer sort of the original entries, and the third part lists each entry again chronological- ly by date of publication. Thus , each entry is listed four times. The first twenty volumes in the series were produced by photo-offset from pho- tographically reduced computer print- outs and averaged about seventy entries per single-column page. Beginning with volume 21, all page copy has been set in 6-point Times Roman type in double columns by a computerized photocompo- sition technique, with approximately 140 entries per page. Volumes are 8~~, x 11", printed on durable paper, and cloth bound. The library is the publisher. The published volumes are extensive- ly used in the Harvard libraries in a variety of ways by both readers and staff. Sets of the entire series are locat~d in reading rooms for reference and in the stacks for circulation to readers. Copies 322 I College & Research Libraries • September 1970 of the volumes covering particular classes are located in special boxes attached to the end panels of the stacks in which the class is located and are used by readers as browsing guides and as convenient finding lists. The availability of the series also tends to reduce somewhat the ob- jections to shelving infrequently used books by size in storage areas outside the library, because these titles are retained in the shelflist with a symbol showing the actual location of the book in stor- age. The volumes are also used by book selectors in building collections as well as by inter-library loan staff, both at Harvard and in other libraries. Since the shelflist volumes form subject catalogs of specific portions of the collection, and since, unlike book catalogs of entire li- braries, they can be purchased separate- ly, many individual scholars acquire per- sonal copies of the volumes covering their field. All costs of the shelflist conversion project, including systems development, conversion, editing and machine costs, have been borne entirely by the library from regularly budgeted funds. All costs incurred in the actual publication of the series, including final computer sorts, photocomposition, printing, binding and distribution, are met from sales receipts. Within this framework the published se- ries has been self-supporting from its in- ception. The rationalization for this large expenditure of library funds is that con- version of the old manuscript shelflist is a necessary improvement of the library's record-keeping operations and that the investment in conversion (an estimated thirty cents per entry) will be amply jus- tified by long-term savings in shelflist maintenance and other library manage- ment gains. Other savings-impossible to measure-are in the time and effort of readers and staff who use the printed shelflist catalogs in lieu of going to the card catalogs. James L. Dolby makes this point nicely in his recent book on com- puterized book catalogs: In particular, we claim that no careful study is necessary to show that a printed catalog on the desk of the user, or at least in the immediate vicinity of his office, is a sufficient advance over the present card catalog to provide a substantial time ad- vantage in his use of the catalog. At the very least, the user is saved a trip to the li- brary for all those searches that prove to be fruitless. Further, in an automated catalog it is feasible to produce many more differ- ent orderings of the catalog (and subsets thereof) than is feasible in a card system. This in turn increases the number of ac- cess points to the library collection and the over-all utility of the catalog to the user. It may be difficult to put a precise dollar fig- ure on the value of added access, but at the first level it is certainly sufficient to off- set minor cost increments in the cataloging operation. 5 The selling price of the individual volumes ranges from $10 to $45 and is based on the number of pages, the esti- mated sales potential of the particular volume, the manufacturing cost, and in special cases, such as the Slavic class, the amount and cost of extra editorial work. Since the shelflist in this form was a new and unfamiliar kind of bibliograph- ical tool, and since the promotion efforts were deliberately limited, sales were initially slow and tended to be limited to the larger American research libraries, many of which placed standing orders for the series. Sales have increased as the series has become larger and better known and as the format has been im- proved. The market for volumes has ranged from four to eight hundred copies, depending on the subject cov- ered; while further improvement is pos- sible, it is unlikely that the sales of any volume will exceed a thousand copies. There has been a market for these vol- umes because they list the holdings of one of the world's great research libraries and as a result are valuable tools for li- brarians and scholars. To date the fol- lowing volumes have been published or are in preparation: Conversion and Publication Program I 323 Volumes Published: 1. Crusades. 1965. 82 pp., 1,202 ti- tles. $3: Out of Print. 2. Africa. 1965. 790 pp., 13,335 titles. $25. Out of Print. 3. Twentieth -Century Russian Literature. 1965. 428 pp., 9,430 titles. $20. Out of Print. 4. Russian History Since 1917. 1966. 698 pp., 13,722 titles. $30. Out of Print. 5-6. Latin America. 1966. 1,492 pp., 27,292 titles. 2 vols., $65. 7. Bibliography. 1966. 1,066 pp., 19,643 titles. $40. 8. Reference Collections. 1966. 187 pp., · 4,300 titles. $10. Out of Print. 9-13. American History. 1967. 4,087 pp., 83,867 titles. 5 vols., $175. 14. China, Japan and Korea. 1968. 494 pp. , 11,388 titles. $25. 15. Periodical Classes. 1968. 758 pp., 25,685 titles. $25. 16-17. Education. 1968. 1,610 pp., 32,- 722 titles. 2 vols., $60. 18. Literature: General and Com- parative. 1968. 189 pp., 5,065 titles. $10. 19. Southern Asia: Afghanistan, Bhutan, Burma, Cambodia , Ceylon, India, Laos, Malaya, Nepal, Pakistan, Sikkim, Singa- pore, Thailand, and Vietnam. 1968. 543 pp., 10,292 titles. $20. 20. Canadian History and Litera- ture. 1968. 411 pp., 10,212 ti- tles. $17.50. 21. Latin American Literature. 1969. 498 pp., 16,900 titles. $40. 22. Government. 1969. 263 pp., 7,190 titles. $20. Volumes in Preparation: 23-24. Economics. 1970. 1,800 pp. , 65,- 000 titles. 2 vols., $95. 25. Celtic Literature. 1970. 192 pp. , 7,500 titles. $25. 26-27. American Literature. 1970. 1,600 pp., 50,000 titles. 2 vols., $95. 28-31. Slavic History and Literatures. 1970. 2,700 pp., 93,000 titles. 4 vols. , $190. 32. General European and World History. 1970. 35,000 titles. $50. 33. Reference Collections. 1970. 160 pp., 5,000 titles. $10. 34. African History and Literatures. 1970. 500 pp., 16,000 titles. $35. English Literature. 1971. 108,- 000 titles. Finnish and Baltic History and Literatures. 1971. Spanish History and Literature. 1971. Note: Another 25 to 30 volumes will be required for the remaining significant classes which are scheduled for publication. Many minor and lesser-used classes will not be published in the se- ries. In 1968, after a thorough analysis of the cost and other factors, a decision was made to change the output system to produce printer's copy by a computer- ized photocomposition process and to discontinue using line-printer output for publication. The logic behind the de- cision was that the increased page den- sity of photocomposed text would reduce the num her of pages in a volume by ap- proximately one-third thus reducing printing costs by a similar amount, while increasing the quality and legibility of the book. Although the cost of creating a photocomposed page is several times the cost of a line-printer page, the in- crease would be more than offset by the reduction in printing costs. Experience proved that this was the case but the savings were not as great as anticipated because the cost of the additional com- puter time required to prepare the tapes for input into the photocomposition ma- chine were underestimated. It costs slightly more to produce the photocom- 324 I College & Research Libraries • September 1970 posed volume, but this added cost is justified because it improves the quality of the finished book immeasurably. The slight increase in cost for producing printer's copy in this manner is a tem- porary penalty only; a significant drop in photocomposition costs can be expect- ed in the next few years as the equip- ment improves, as the volume of busi- ness increases, and as the industry be- comes more competitive. Even at cur- rent prices, photocomposition is a minor cost breakthrough for the production of book catalogs, particularly in large edi- tions where the savings in printing and paper costs are important. The relatively new COM (computer- output-microfilm) technology may well provide the solution to the problem of producing small editions of book catalogs at acceptable costs. This process pro- duces output from a magnetic tape onto 16 or 35 mm microfilm at tape running speeds. 6 The cost of producing the film is considerably less than line-printer out- put, and the quality of the print image is somewhat superior to that of the line printer. However, it does not compare with photocomposed copy, which is sig- nificantly better but several times more costly. The COM output can either be used in microfilm or automatically enlarged to full-sized master copy for reproduc- tion in small editions. Because of the poorer quality product and other uncer- tainties, the COM process is not being considered as a possible alternative to the present photocomposition process. However, it is being considered as a means of maintaining the official shelf- list and more will be said about this later. In the longer range, and particularly after the entire shelflist has been con- verted, COM will offer many interesting possibilities for exploiting the shelflist data base so that a whole variety of list- ings in different sequences and for dif- ferent purposes can be published in small, inexpensive microform or even full-sized editions depending on the need and use to be made of them. The problem of issuing supplements or revised editions of the volumes in the current Shelflist series is a difficult one. The publication and distribution of sup- plements to the individual volumes is questionable from the point of view of both costs and usability. It has been re- jected in favor of issuing new and en- larged editions when the basic volume has become seriously outdated, general- ly after five or more years. Thus, the contents of the first volume, Crusades, will be included as part of the General European and World History volume; volume 2, Africa, which was published in 1965, will be revised, enlarged and reissued in 1970 in the new photocom- posed format; other early volumes in the series will be treated in a similar man- ner. In the future , the problem of pub- lishing subsequent editions may well be solved by advances in technology and improvements in the economics of pub- lishing. COM and reductions in the cost of photocomposition and computing are reasonable expectations in the near fu- ture. As has already been suggested, these developments may make possible the publication of special or even custom listings of great usefulness, but of rela- tively limited demand. For example, upon completion of conversion of the entire shelflist it might be desirable and feasible to produce, by COM at an ac- ceptable cost, an up-to-date microform edition of the entire file in classified, author and title, and chronological se- quence. Listings by language would also be possible as would a listing of all seri- als and journals in the collection ar- ranged in a single alphabetical sequence. Current accessions lists would be anoth- er useful product. The technique of merging several re- lated classes into a single sequence has already been accomplished with excel- Conversion and Publication Program I 325 lent results and could be further ex- ploited. An example of this would be to expand the Slavic History and Litera- ture class into a comprehensive Slavic area studies catalog by adding the Slavic titles from other classes such as Educa- tion, Folklore, Philology, Sociology, Gov- ernment, etc. The technique could be applied to other areas such as Africa and Latin America. New shelflist-type cata- logs of Judaica and other subjects might be created by pulling together the bib- liographic entries that are located in the various country and literature classes as well as in Sociology, Folklore, etc. Mis- cellaneous scattered titles might be lo- cated by searching the tapes for certain key words in titles. The results would have to be edited to eliminate false drops but the process might be useful as a first pass. Similar techniques could be used to search the data base and create special or custom listings for individual ·scholars or groups on request. When it becomes economically feasi- ble to store such a large data file in a direct access device and to search and manipulate it from a cathode ray tube console, the possibilities for making in- teresting and novel uses of the data will be expanded enormously. While mass storage and on-line direct access is an operational technology today, it will probably be several years before it will be economically feasible in the research library environment. 7 It seems idle,' therefore, to speculate about these in- teresting but relatively remote possibil- ities in an article set in the context of current economic realities in libraries. Recent experience indicates that im- provements in computer and photo- graphic technology are occurring at an ever-accelerating rate, and the possibil- ity of dramatic advances and cost break- throughs in the next few years should not be discounted. A long range but still realistic idea is the possibility of turning the conven- tional library shelflist into a kind of clas- sified catalog once it has been converted and is maintained in machine readable form. The basic difference between a conventional shelflist and a classified catalog is that the shelflist treats a book as a single physical object and records it only once, no matter how many sub- jects it covers, while a classified catalog records the book in as many places as its subject requires. With a computer- ized shelflist, the reason for this limita- tion no longer exists; a book can be given one number to record its physical loca- tion, and several other class or base num- bers to indicate facets of content. Thus , a single book could appear several times and in various classes. The two types of call numbers would be distinguished by a symbol or other means, and these add- ed entries could be printed or sup- pressed depending on the use to be made of the list. The introduction of this innovation in shelflisting can only be done after conversion has been complet- ed and it has, therefore, not yet been proposed for the Widener shelflist. Library Management Uses The present system for adding en- tries to the official copy of the computer- produced printout shelflists is identical with the system for adding to the old manuscript shelflists. The machine lists are printed with five blank lines between the entries in order to leave space for writing in new additions. Periodically, and as the pages b ecome crowded, all new entries and changes in the list are keypunched and added to the master tape file and a new printout replaces the old one. The inefficiencies of this proce- dure are obvious, but they were tolerat- ed in the early stages of the conversion project on the grounds that it was pref- erable to have a single shelflisting pro- cedure for both manuscript and ma- chine-produced shelflists until such time as the proportion of machine lists in- creased to a point where a second sys- tem would yield significant savings. 326 I College & Research Libraries • September 1970 Now that more than a third of the shelflist, including many of the most ac- tive classes, is in machine form, the con- ceptual design of a machine based sys- tem for maintaining official copy has been developed and is being consid- ered. It can be briefly described as fol- lows: all classes in machine format would be updated and produced in an efficient single-spaced format on micro- film or microfiche using a COM (com- puter-output-microfilm) technique. This film would serve as the official shelflist copy along with a temporary card sup- plement. Book numbers for new books would be assigned by consulting both the film and the card supplement. The number would be preempted by making a temporary slip for it in the supplement, and this slip would be replaced by a unit card after it had been produced. Periodically the contents of the card supplement would be converted to ma- chine form, merged with the master tape, a new cumulated official film or fiche version would be produced by COM, and a new card supplement would be started. This procedure could be further simplified after developing and implementing a system to input cur- rent cataloging into machine readable form, but even in the interim the sav- ings would be substantial. Assuming a cost of five cents per frame of microfilm containing 80 entries, the entire shelflist of 1.6 million entries would require 20,- 000 frames and could be produced for about $1,000 on approximately 12 reels of film. A microfiche version would re- quire only 250 4" x 6" fiche. Computer printing and other costs would be substantially less than in the present system. Shelflisting now requires a staff of four persons and an area of 600 square feet. It could probably be reduced to a single work station located in the cataloging room where it logical- ly belongs, while reference copies could be maintained in other locations. The completed shelflist file can be made to serve many of the purposes of a central bibliographical record in ma- chine form. By running call numbers against this file a variety of products could be produced such as machine readable book cards for an automated circulation system, lists of overdue books, missing books, and books to be replaced or purchased in duplicate. In short, any list of call numbers could be expanded into full shelflist type entries by simply key-punching them and matching them with the data file by the aid of a pro- gram. The records of the one and one-quarter million circulation transactions made in Widener since 1965, when the machine system was installed, have been pre- served on five reels of magnetic tape and constitute an invaluable and unique data base from which statistical analyses of the use of the collection have been made. 8 One of the chief limitations of this file comes from the fact that the bibliographic data in the charge records is limited to call numbers. This limita- tion can be overcome by using the call numbers to extract the complete entries from the shelflist file. Thus, for example, listings of the most frequently used ti- tles could be obtained by sorting the charges in the order of frequency of use and using the resulting call numbers to obtain a listing of the bibliographical en- tries from the master shelflist file. De- cisions about where to locate material in the library and which material to send to deposit collections can be made on the basis of these statistics. Such potentially useful management information has nev- er before been available to library ad- ministrators. Another area of statistical analysis that is opened by the existence of the master shelflist file is the analysis of the collections themselves, their make-up, their rate of growth over the years and in various subject areas. Detailed and accurate counts can be obtained of the individual classes and of the collection SAL 5611 . 14.100 SAL 5723.6.100 SAL 7723.1.35 SAL 9125.19 . 100 SAL 4870.18 . 120 SAL 5657.1.32 SAL 7113.35 . 120 SAL 9171.50.120 SAL 427.82.100 SAL 9149 .99.41 SAL 9272.51.140 SAL 321.1.15 . 15 SAL 4409 .90 .320 SAL 1526 .6. 110 SAL 922 .55 . 100 SAL 5679.86 SAL 7720.41 .200 SAL 1740.25.51 SAL 7627.3.100 SAL 5513.12.120 SAL 4545.77.100 SAL 5206.65.100 SAL 7189.14.31 SAL 7780.3.41 SAL 9189.14.220 SAL 9121.5.120 SAL 8076.10.100 SAL 4327 .66.1100 SAL 1731.33 . 100 SAL 296.41.130 SAL 9226.41 .2800 SAL 9173.5 . 110 SAL 9173.5.2100 SAL 9173.5.100 SAL 4569.3.100 SAL 4310.23 , 11)() :;AL 365.LJ Dialogos en "La rein a del mar" . (Garcia, J .J .) Bogota, 1965. Dialogos en otros mundos. (Restrepo, Felix.) Manizales, 1936. Dialogos limpicos. v .1-2. (Reyles, C.) Buenos Aires. 1919. Dialogos no espelho . (Bandeira, Antonio Rangel.) Sao Paulo, 1968. Dialogos y "cuentos de mi paisaje". (Medinaceli, Carlos .) La Paz, 1963. Dialogos y juguetes escenicos. (Leon Gomez, Adolf.) Bogota, 1909. Diamantes y pedernales. (Arguedas, Jose M .) Lima, 1954. Diamantina. (Delfino, Aldo.) Bello Horizonte, 1914. El diametro y lo estero . (Rossardi, Orlando .) Madrid, 1964. Diario. Capa de Carlos Penafiel. (Cardoso, Lucio.) Rio de Janeiro, 1960? Diario critico de Sergio Milliet (1940- 1943) . (Milliet, S.) Sao Paulo, 1944-45. 2v. Diario de amor. (Gomez de Avellaneda, G.) Madrid, 1928? El diario de Gabriel Quiroga. (Galvez, Manuel.) Buenos Aires, 1910. El diario de Jose Toledo . Nove Ia. I a ed . (Barbachano Ponce , Miguel.) Mexico, 1964. Diario de Ia montana, 1957-1960. (Ramirez de Arellano de Nolla, Olga.) San Juan, 1967 . Diario de Lecumherri. (Mutis, Alvaro.) Xalapa, 1960 . Diario de naje a Paris . (Quiroga, H.) Montevideo, 1950 . Diario de nuestro viaje a los Estados Unidos . V. 2. (Sierra, Justo.) Mexico, 1938 . Diario de Roberto. (Halegua, 1.) Montevideo, 1960. Diario de un peat6n. (Arciniegas, German.) Bogota, 1936. Diario de un solteron penitente. (Speroni , Miguel Angel.) Buenos Aires, 1940. Diario de una "patinadora ". (Frontaura, Rafael.) Santiago de Chile, 1963? Diario de viaje . (Escobar, Alberto.) Lima, 1958 . Diario de viaje. (Viana, Francisco Javier de.) Montevideo, I~ Diario do Exercit SAL 5060.70 SAL 5060.71 SAL 5022.16 SAL 5137.1.82 SAL 5218.19 .820 SAL 5020.17 SAL 5173.2.21 SAL 5068.2 SAL 5022.15 SAL 1573.4 . 100 SAL 5173.5.100 SAL 5573.2.31 SAL 5573.2.35 SAL 1573 .6. 120 SAL 1573 .6. 100 SAL 1573.6.12 SAL 1573.6 .1 10 SAL 6142.48.800 SAL 9065.7 SAL 1573 .7.100 SAL 1573 .7. 110 Diaz Arrieta, Hernan. Las cien mejores poesias chilenas. 3a ed . Santiago de Chile, 1957. Diaz Arrieta, Hernan. Los cien mejores poesias chilenas. 4a ed . Santiago de Chile, 1962 . Diaz Arrieta, Hernan. Los cuatro grandes de Ia literatura chilena. Santiago de Chile, 1963 . Diaz Arrieta, Hernan. Don Alberto Blest Gana . Santiago de Chile, 1940. Diaz Arrieta, Hernan . Gabriela Mistral. Santiago de Chile, 1946 . Diaz Arrieta, Hernan. Historia personal de Ia literatura chilena. Santiago de Chile, 1954 . Diaz Arrieta, Hernan. Leer y escribir. Santiago de Chile, 1962. Diaz Arrieta, Hernan. Memorialistas chilenos. Santiago de . Chile, 1960. Diaz Arrieta, Hernan . Panorama de Ia literatura chilena durante el siglo XX . Santiago de Chile, 1931. Diaz Bolio, J . El mayab resplandeciente. Mexico , 1934. Diaz Casanueva, H. Requiem . 3a ed. Santiago de Chile, 1958. Diaz Castro, Eugenio. Manuela. Paris, 1889. 2v. Diaz Castro, Eugenio. Una ronda de don Ventura Ahumada y otros cuadros. Bogota, 1936. Diaz Covarrubias, Juan. Gil Gomez el insurge nte; novela historica . Mexico, 1919 . Diaz Covarrubias, Juan . lmpresiones y sentimientos . Mexico, 1857 . Diaz Covarrubias, Juan. Obras completas. I a ed. Mexico, 1954. Diaz Covarrubias, Juan . Pagi nas del coraz6n; poesias . Mexico, 1859. Diaz Cuera, Miguel. Bibliografia de fra y Vi-.ente Solano. Cuenca, 1965 . Diaz da Cruz, E. Pequeria antologi a de cuentos brasileiios. Buenos Aires, 1946. Diaz de Leon, Rafael. Por los pobres. Hermosillo, 1921. Diaz de Leon, Rafael. Semidiosa, prosas romanticas . Mexico. 1923 . Heroe de San Mateo. Pasto, 1914 . de.) Rio de Ja Diario I; dificil e ' Janeiro, 1962. WIDENER LIBRARY SHELFLIST, 21 Teatro: El pretor. Bogota, 1967 . Teatro . V. 1,3-4- Bogo ta, 1963- )v. itmos de libertad ; sonetos . m. La voz del torrente . 2a ed. Santia go Diario intimo. (M Diario poetico, ab Aires, 1961. AUTHOR AND TITLE LISTING fipo s de Ia Habana . Habana, 1895 . Jonzalez. 0 . Collection of play s. 9 pam . io. Cr6nicas de Altocerro . Santo Diario semanario Xalapa, 1961 . Diarios de viajes. , . Dias, Fernando Cerreia. Jose Alphonsus: tempo e modo . Belo Horizonte, 1965. Dias, Je6filo . Poesias escolhidas. Sao Paulo, 1960. Dias, Oscar. Perfil com movimento. Goiania, 1966. Dias, Theophilo. 0 comedia dos deuses; poema . Sao Paulo, 1887 . Los dias. (Uribe, B.) Buenos Aires, 1959 . Dias ajenos. (Antenta, E.) Buenos Aires, 1962. Dias amargos; poesias . (Malde Vizoso, E.) Habana, 1915 . SAL 3173.5.700 SAL 3173.5 . 102 SAL 3173.5.100 SAL 4705.89 .800 SAL 4773.23 . 120 SAL 4773.23.100 Diaz Lozano, Argentina . Enriqueta and I. N.Y., 1944 . Diaz Lozano, Argentina . Peregrinaje. 2a ed. Guatemala, 1955. Diaz Lozano, Argentina. Peregrinaje . Santiago de Chile, 1944 . Diaz Machado, P. Nataniel Aguirre. Buenos Aires, 1945 . Diaz Machicao, Porfirio . Cauce de palabras . La Paz. 1967. Diaz Machicao, Porfirio. Cronica de cronicas. La Paz, 1963. SAL 4231.5 SAL 4231.8 SAL4231 . 10 SAL 4231.12 SAL 4231.12 . 12 SAL 4231.12.15 SAL4231.13 SAL4231.14 SAL 4231.15 Echague, Juan P. Teatro argentino . Madrid, 1917. Bianchi, A .A . Teatro nacional. Buenos Aires, 1920. Rodriguez Acasuso , L. Del teatro al libro . Buenos Aires, 1920. · Echague, Juan P. Un teatro en formaci6n. Buenos Aires, 1919. Echague, Juan P. Una epoca del teatro argentino 1904- 18. 2a ed. Buen os Aires, 1926 . Echague, Juan P. Le theatre argentin. Paris, 1927 . Ordaz, L. Breve historia del teatro argentino. Buenos Aires, 1962-63. 8v. Bastardi , Francisco . Yo tambien con mi s memorias . Buenos Aires, 1963 . Blanco Amares de Pagella, Angela . Nuevos temas en el teatro a rgentino . Buenos Aires, 1965 . SAL 4232 Latin American literature in Spanish- Argentina- History of special forms - Drama - Local SAL 4232 . 1 Bosch, M .G . Te atro antiguo de Buenos Aires. Buenos SAL 4232. 1.3 SAL 4232 . 1.1 0 SAL 4232 .2 SAL 4232.2. 3 SAL 423 2.3 SAL 423 2.5 SAL 4232 .7 SAL 4232 .8 Aires, 1904 . Bosch, M.G. His to ria del teatro en Buenos Aires. Buenos Aires, 1910 . Castillo, C. Hi storia de los origenes del teatro nacional argentino y Ia epoca de Pablo Podesta . Buenos Aires, 1929 . Rela , Walter. El mito Santos Vega en el teatro del Rio de Ia Plata . Montevideo, 1958 . Rel a, Walter. El mito Santos Vega en el teatro del Rio de Ia Plata . 3a ed . Montevideo, 1966 . Castagnin o , R . H . Contribucion documental a Ia historia del teatro en Buenos Aires. Buenos Aires, 1944. Beltran, O . R . Las origenes del teatro argentino. Buenos Aires, 1934 . Bianchi, A.A . Veintiuno aflos de teatro nacional. Buenos Aires, 1927. M a rtinez, 0. Historia del teatro en Bahia Blan ca . Bahia Blanca? 191 3 SAL 4233 Latin American literature in Spani SAL 4240 .24 SAL 4240 .26 de Ia i:poca de Ia Revoluc10n de Mayo. Buenos Aires, 196- 5v. Fernandez, Juan Romulo . Civilizaci6n argentina; Ia obra de La Prensa en 50 aflos. Buenos Aires, 1919 . Gimenez, Jorge C. Los delitos de imprenta . Buenos Aires, 1899 . SAL 4251 Latin American literature in Spanish- Argentina- General anthologies, collections - General works - Other general SAL 4251.1 Martinez, B.T . Antologia argentina, trozos historicos . SAL4251. 2 SAL 4251.5 .2 SAL 4251.5.10 SAL 4251.10 SAL 4251.11 SAL 4251.30 SAL 4251.35 SAL 4251 .35 .2 SAL 4251.40 SAL 4251.42 SAL 4251.44 SAL 4251.46 Buenos Aires, 1890. 2v. Gutierrez, J.M . Pensamientos, maximas, sentencias. Buenos Aires , 1859. Benavento, G .O . Letras. 2a ed. Buenos Aires, 1930. Benavento, G.O. Letras. 6a ed . Buenos Aires, 1935. lbarguren, C . Le paysage et !'arne argentins. Buenos Aires, 1938. Cobello , Jose. Primera antologia lunfarda . Buenos Aires, 1961. Berdiales, G. Maestros del idioma . 9a ed. Buenos Aires, 1941? Bullrich Palenque, Sylvina. El compadrito, su destino . Buenos Aires, 1945 . Bullrich Palenque, Sylvina. El compadrito. 2a ed. Buenos Aires, 1968 . Villanueva, A. El ombu Ia civilizaci6n. Santa Fe, 1955 . Garganigo, John F. Antologia de Ia literatura gauchesca y criollista . Montevideo, 1967. Prieto , Adolfo. El peri6dico Martin Fierro. Buenos Aires, 1968 . Ruffinelli, Jorge. La revista Caras y caretas. Buenos Aires , 1968 . SAL 4255 Latin American literature in Spanish- Argentina- General anthologies, collections - Special periods SAL 4 255.10 1-l e!".riquez, U.P. Antologia clasica de Ia literatura aT!! enti •'" ~a ed Buenos Aires, 193- ,a de Ia Presidencia. Mensajes a Ia .1. n.p ., 1949 . of special forms - Drama - Special topics SAL 4233 .5 Agilda, E. El ~ WIDENER LIBRARY SHELFLIST, 21 ; ias y poesia . Buenos A ires, 1961. Los conversadores. Buenos Aires, 1942 . tologia de Boedo y Florida. I a ed . Aires, 1960 . SAL 4233 . 10 Gallo Bias, Ra. SAL 4 233 . 15 Aires, 1959 . Capde vila, Artl Aires, 1951. CLASSIFIED LISTING BY CALL NUMBER udo. Los fundadores. Buenos SAL 4234.1 - .99 Latin American literatu! Argentina - History of special forms - Drama - Biographies of actors, directors, etc. - Collected SAL 4234.57 Munoz, Andres. 30 vidas de arti stas argentinos. Buen os Aires, 1940 . SAL 4235 Latin America11 literature in Spanish- Argentina- History of special forms - Fiction - General works SAL 4235.5 G a rcia, German . La novela argentina; un itinerari o. Buenos Aires, 1952 . SAL 4235 . I 0 Co bo, A .J. Adonde va Ia literatura a rgentina? Buenos Aires, 1954. SAL 4 235 . 12 Ghiano, J .C. Testimonio de Ia no ve Ia argentina . Bueno s Aires, 1956 . udo . La literatura virreinal. Buenos ~AL 4258 Latin Arnt!.;.:an liler .. l.-h: . .. _,pamsh ., Argentina- General anthologies, collections - Local SAL 4258.5 Celebraci6n nacional. .. juegos florales. Tucum a n, 1916. SAL 4260 Latin American literature in Spanish-. Argentina- Anthologies of special forms - Poetry SAL 4260.1 Poesias arjentinas. Buenos Aires, 1869. 3 pam . SAL 4260 .2 Poesias arjentinas. Buenos Aires, 1875. 4 pam . SAL 4260.3 Poesias arjentinas . Buenos Aires, 1877. 4 pam. SAL 4260.4 Ascasubi, H. Aniceto el gallo ... y otras poesias. Paris, 1872. SAL 4260.5 Pamphlet vol. Poetry . Argentine Republic . 20 pam . SAL 4260 .7 La lira argentina . Buenos Aires, 1824 . 1 SAL 316.1 .31 SAL321.1 .6 SAL 321.1 . 15.10 SAL 321 . I .14 SAL 328.1 .7 SAL 362.41.100 SAL 7664 . 1.4 SAL 496 .2.11.2 SAL 7851.7 SAL 160 .25 SAL 1597 .50.106 SAL 356.66.100 SAL 4534 . 1.8 SAL 5269.1.4 SAL 11567 .54 . 120 SAL 155 .20 SAL 1460.12 SAL 388.4 I . I 00 SAL 1433. 21-7 SAL 1726 .2 .3 SAL5757.1 .5 SAL 5137.2.2.3 SAL 250.17.100 SAL 7164.1.2 SAL 9196.77 . 100 SAL 1609 .I .I 0 SAL4467.1.3 SAL 60.25 SAL 375.83.100 SAL 1708 . 1.3 SAL 1729.41.100 SAL 475.9.100 SAL 7119.41.100 SAL 481.1.8 SAL 5134.1.4 SAL 304 . 1.6 Gimenez, J. Bibiana, o los terremotos de Cuha. Cuba, 1852. Gomez de Avellaneda, G . El heroe de Bailen ; loa. Madrid, 1852. Gomez de Avellaneda, G . La hija de las flores, 6 Todos estan locos. Madrid, 1852 . Gomez de Avellaneda, G . La verdad vence apariencias. Madrid, 1852. Heredia, Jose M . Poesfas . La Hahana, 1852 . Lorie, Antonio M . El vie rnes 20 de agosto en Cuba. Cuba, 1852. M agaririos Cervantes, A. Celiar, lcyenda americana en va riedad de metros . Madrid, 1852. Zequefra y Arango, Manuel. Poesfas . 2a ed . llabana, 1852. 1853 Aguinaldo religioso del 'Co rreo de Caracas'. Caracas, 1853. Cuatro laudes por Ramon Zambrana . Habana, 1853 . Fernandez Lizardi, J.J. El periquillo sa rnient o. Sa ed . Mexico, 1853. 4v. Leon: J .S . de. Flores silvestres. Habana, 1853 . 2 pam . Martmez Vlllergas, J. Sarmiento 6 a mal Sarmiento. Pari s, 1853. Matta, G. Cuentos en ve rso. Santiago de Chile, 1853 . Merlin , Mercedes Jaruco da. Memorias y recuerdos de Ia senora Condesa de Merlin . V.l-2. Habana, 1853 . Murmurios del canto. Cuba, 1853 . a_va rro, }.R. Guirnalda p oe tica. Mejico, 1853 . Oc1o, J ose de Jesus de. I Clavellinas I, poesfas. Matanzas, 1853. Re g lamento para los teatros de Mexico. Mexico 1853 . Rodri g uez, J.M . E l anahuac. Mexico, 1853. ' Torres Caicedo, J . M. Coleccion de poesfas orijinales . N .Y ., 1853 . SAL 8011.2.5 SAL 9220.1 .41 SAL 9220 . 1 .37 SAL 361.5 .31 SAL 363.3.29 SAL 9264.3.25 SAL 372.50 .20 SAL 398. 1.36 SAL 410.13 . 100 SAL 5732 .69 . 12 SAL 5333.1.6 SAL 5333.1.7 SAL 456.51.20 SAL 475.1.1.5 SAL 475.1.21 SAL 202.1.31 SAL 7908. I .5 SAL 224.1.9 SAL 5531.58.100 SAL 232 .3.32 SAL 237 . 1.2 SAL 253 .69.22 SAL 160 .89 Caracas, I !!57 . Garcia de Quevedo. J.H. Leyendas . Caracas, 1!!57 . . Gon~;alves Dias, Antonio . Cantos; collec~;ao de p oes 1as . 3a ed . Leipzig, 1857 . . Gon~;alves Dias, Antonio. Os Tymhicas; poema amer:cano . Leipzig, 1857 . . . . Lorna Ossorio, J.M . Flores sin fruto, o msp trac1ones poeticas. Rev. ed. Hahana, 1857 . Luaces, Joaquin L. Poe sias . Hahana. 1857 . . Macedo, Joaquim M . de . A nehulosa; poema. R10 de Janeiro, 1857 . Mill an, Jose Agustin . Obras dram£1ticas. V. 1-2 . Hahana, 1857. Otero, Rafael. Un cohurgo. Matanzas, 1857 . Pichardo, E. Villaclara romanti ca. Villaclara, 1857. Samper, Jose Maria. Coleccion de piezas dramatlcas . Bogo ta, 1857 . . . Sanfuentes, S. Ricardo i Lucia. Santiago de Chile, 1857 . 2v . Sanfuentes, S. Teudo. Santiago de Chile, 1857 . Talon, Mi g uel T. Flores i espinas . Matanzas, 1857 . 3 pam. Valdes, G. de Ia C. Poes ias completas . Paris, 1857 . Valdes, G. de Ia C. Poesias de Placid o . 3a ed . N.Y., 1857. 2v. 185S Acosta, Ignacio M. de. Romance historico y geog rafico de Ia isla de Cuba. Matanzas, 1858 . Alfonso, R.M . Las quejos de Venezuela. Caracas, 1858 . Balmaseda, F.J. Fabulas morales. Habana, 1858 . Bernal 0 ., R . Vierie por mi i carga con usted . Bogota, 1.858. 2 pam . . . Betancourt, Jose R. de. Una fena de Ia candad en 183. 2a ed. Habana, 1858. Blanchet, Emilio. Versos y prosa. Matanzas, 1858 . Cepero, Belen . Ayes del corazon. Habana, 1858-65. _3 ~am. Corona funebre a Ia memoria de Ia virtuosa y bella se nonta :s Fernandez de Castro y Trelles. )8. Los hijos naturales; drama en tres aetas . Ble st Gana, G. Ca rdenas y Ch: Corpancho, M . Ferreira, Franc Gallego, Juan Marmo!, Jose . Mendive, R . M Montaos y Rol WIDENER LIBRARY SHELFLIST, 21 do, J.H. Versos. Caracas, 1858. CHRONOLOGICAL LISTING Matanzas, I Pesado, J.J. L.s-s dLu.;ca · u . ovl<.:An.v, ,.., _, ... Rui seco, Tomas. El ge mu cspariol. Veracruz, 1854 . Valdes Mendoza, M . Poesias. Habana, 1854 . 1855 Avila, Manuel. Re vol uci o n del Peru . Lima, 1855 . Baudry, Libreria Europea. Obras de Vinageras. Juici o critico. Pari s, 1855. Bilbao, Francisco. Coleccion. n.p., 1855 . 6 pam . Fornaris, Jose . Poesias . Habana , 1855 . J T \ L"'t\.JJ . VJ.IVV SAL 5707.23.100 SAL 5707.25.115 SAL 420.3.31 SAL 427.4.21 SAL 481.65.100 M . A ly ra gemedora; poesias . Ri o de . Obra del primer hombre . n.p., 1858 . as. 2a ed. Madrid, 1858. 2v. leccion de articulos y poesias . i8. , ..-vn uu.-rz.anr. Poesias. Matanzas, 1858. Pcrei~a Gamba, P. Akimen-Zaque 6 La conquista de Tunja. Bogo ta, 1858 . Perez, Felipe. Jilm a 6 continuaci6n de Los Pizarros. Bogo ta, 1858. Quintana, A.M. El retrato del diablo. Matanzas, 1858-59. Romay, Tomas. Ohras escogidas. Habana, 1858-61. 2v. Vipolas, Pedro . El coraz6n de una actriz 6 Suerio Y realidad . Habana , 1858 . 330 I College & Research Libraries • September 1970 as a whole, e.g., counts by class, by lan- guage, by place of publication, by date, as well as counts of serials, monographs, and volumes. Many of these statistics have already been obtained from the converted classes and used for manage- ment purposes. The general research value of the bibliographical data contained in large research library catalogs has already been recognized and exploited to some extent by Dolby, Forsyth, and Resnikofi. 9 They have used data from one of the published volumes of the Widener shelf- list and are currently working with the computer tapes of other classes. 10 Their views on the statistical uses of catalogs in machine readable form have been summarized as follows: Library catalogs contain a wealth of infor- mation about the historic development of the many fields of human endeavor and the interrelations that bind these activities. Mechanization of the catalog permits ex- ploitation of this information by workers in many fields of research. Analysis of the same information can greatly assist librar- ians in studying their own collections and jn managing the acquisition of materials for the library. Many studies of this type can be conducted on random samples of the catalog, though more detailed work re- quires access to the entire collection in ma- chine-readable form. 11 CONCLUSION In 1968 this author concluded a de- scription of the shelflist conversion proj- ect with this statement: As it now stands, the Widener shelflist program, like many other present library computer systems, is regarded as an interim system designed to extract the maximum return from a simple existing bibliographi- cal record of the contents of the Library. It is expected that in time the system will be- come obsolete and the imperfect shelflist entries will be superseded by standard bib- liographical records in the emerging Li- brary of Congress MARC II format ... . The expectation is that a central biblio- graphical agency will convert and distrib- ute these entries. It seems reasonable to suppose, however, that this conversion ef- fort is still some years in the future and that, in the meantime, Harvard will have realized a satisfactory return on its invest- ment in converting an abbreviated biblio- graphical record.1 2 Developments during the two years that have passed since that statement was made only serve to confirm this brief as- sessment of the program. REFERENCES 1. Richard De Gennaro, "A Computer Produced Shelflist," College & Re- search Libraries 26, 4:311-315, 353 (July 1965). 2. ---, "Automation in the Harvard Col- lege Library," Harvard Library Bulle- tin 16:217-236 (July 1968). 3. Foster M. Palmer, "Conversion of Exist- ing Records in Large Libraries, with Special Reference to the Widener Shelf- list," in The Brasenose Conference on the Automatwn of Libraries, Proceed- ings of the Anglo-Ame1·ican Conference on the M echanizatian of Libraries held at Oxford ... 30 June-3 July 1966. (London and Chicago: Mansell, 1967 ), p.57-80. 4. Conversion of Retrospective Catalog Records to Machine Readabl~ Form , A Study of a Natwnal Bibliog1·aphical Service; prepared by the RECON Working Task Force, Henriette D. Av- ram, Chairman. (Washington, D.C.: Li- brary of Congress, 1969); see also> Conversion and Publication Program I 331 Richard De Gennaro, "A National Bib- liographical Data Base in Machine- Readable Form: Progress and Pros- pects," Library Trends vol. 18, no. 4: 537-50 (April1970). 5. James L. Dolby, V. J. Forsyth, and H. L. Resnikoff, Computerized Library Catalogs: Their Growth, Cost and Util- ity (Cambridge, Mass.: M.I.T. Press, 1969), p.25. 6. Don M. Avedon, Computer Output Mi- crofilm (NMA Monograph No.4). (An- napolis, Md.: National Microfilm Asso- ciation, 1969). 7. The Institute of Library Research at the University of California, Berkeley, is operating such a system in a re- search environment, as are several oth- er groups. 8. Foster M. Palmer, Widener Library Circulation Statistics 1965-1969: Book Use and Stack Space (Unpublished working paper. March 1970). 18p. 9. Dolby, Computerized Library Catalogs (Chapter 6: "On Economic Growth of Nations and Archival Collections"). p.115-33. 10. Ibid., (Chapter 1), p.1-19. 11. Ibid., p.17. 12. De Gennaro, "Automation in the Har- vard College Library," p.229.