To what degree do the list and per-page prices of popular fiction and nonfiction vary by category, binding, and presence of illustrations? To study this question, two issues of Booklist, one from 1991 and one from 1992, were used to generate a sample of 343 and 352 items. Two hundred seven publishers were represented. About 5% of the books were starred by Booklist as being of unusual value. Between 81% and 87% of the books were cloth editions, and 46% to 50% were illustrated. The mean price for an adult book was $21.95 in 1991 and $20.98 in 1992. Per-page prices were easily calculated. The mean price per page for an adult book was 15 cents in both 1991 and 1992. Between 8% and 10% of the books in each year had per-page prices far beyond the mean. Binding appears to be the best predictor of list and per-page price. Relationships between list price and per-page price and category, binding, illustration, audience, and publisher are discussed.
The syndetic structure of subject heading lists, in particular the broader/na¬rower term references, constitutes a hidden classification that may be converted to an explicit tree structure. Such a structure may be used to examine the hierarchy of LC subject headings (LCSH) and to compare them with that of Library of Congress Classification (LCC). Joseph Galron's compilation, Library of Congress Subject Headings in Jewish Studies (199I), was analyzed for several features relating to the hierarchy of terms, and tree structures were built for the deepest hierarchies: Jews, Judaism, Hebrew language, and Israel. These were compared with the corresponding LC classes. A hierarchy in LCSH may have more levels than the corresponding schedule in LCC. It is concluded that the conversion of the broader-term narrower-term references of subject heading lists into tree structures is a useful tool for examining the correctness of a hierarchy. Display of subject headings in tree-structure format can assist users in grasping the hierarchy of subject headings and in navigating online catalogs.
A project currently is being conducted at Cornell University to test the feasibility of using digital image technology to preserve and improve access to deteriorating library materials. The findings of an experiment to produce microfilm from digital images by means of an electron beam recorder are reported in this article. Cornell evaluated the quality of the microfilm, computed its "digital resolution" based on a formula recently developed by an Association for Information and Image Management (AIIM) technical committee, and compared it to printers' type sizes used by publishers during the period 1800-1950). Based an these analyses, Cornell has concluded that a scanning resolution of 600 dots per inch is sufficient to produce digital computer output microfilm (COM) that meets American National Standards Institute/Association for Information and Image Management (ANSI/AIIM) standards for image quality for virtually all books published during the period of paper's greatest brittleness. Although the Cornell experiment demonstrated the technical feasibility of producing preservation microfilm, some of the issues surrounding quality, processing, costs, and vendor services associated with the conversion process have yet to he resolved. Nonetheless, digital technology's primary preservation use in the immediate future could be to serge as an alternative means for capturing material and producing microfilm that meets preservation standards, while also allowing for the flexibility in storage, distribution, and access associated with the technology.
The feasibility of using machine links based upon the presence of standard record control numbers in linking entries to cluster successive-entry serial bibliographic records was studied. Two potential scenarios were considered for clustering serial successive entries in a database of predominantly CONSER authenticated records. A sample of 400 title changes was studied to determine the occurrences of OCLC control numbers, ISSNs, and LCCNs in linking fields. In all, 71.29% of title-change record sets would be linked if an approach is used that takes into account the presence of all of any one of the types of standard record control numbers in the required linking fields.