Essays in Librarianship and Bibliography. Richard Garnett

volumes of the manuscript catalogue having been selected to be combined in a volume of print, they undergo in the first place a literary revision. Queries respecting headings, authorship, and date are raised and settled, mistranscriptions and wrong punctuation corrected, and the catalogue is weeded of its practically duplicate entries by cutting these down to the mere phrase "another edition; another copy," as the case may require. A second and more troublesome revision then becomes necessary, for the order of the entries frequently admits of great improvement. The titles having been incorporated by a variety of persons, and the process of insertion having now gone on for more than thirty years, many errors and inconsistencies have inevitably crept in, and these require to be rectified by an assistant of especial ability and experience in this department of work, whose researches frequently originate a new set of catalogue queries. At last, however, the copy goes to press, the proof is promptly returned and corrected (we are content with a single revise), and the three or four bulky volumes of manuscript are condensed into a single handy and portable volume of type, printed in double columns and on ordinary paper for subscribers, but for reading-room use in single column on a strong vellum paper, adapted to bear rough handling, the opposite column being left blank for insertions, and the book supplied with guards to allow of interleaving. There have hitherto been on the average 220 columns or 110 folios to a volume. On the average of twenty entries to a column, which is rather under the mark, this gives 4400 titles to each volume. The blank space left for insertions and the provision for interleaving would allow of this number of titles being quadrupled, but the weight of the paper prescribes a limit which it would be inconvenient to transgress. Supposing that each volume will take 9000 titles only, then, as the Reading Room will accommodate 2000 volumes of catalogue without encroachment on the reference library, sufficient space will have been provided for eighteen millions of titles, or for three centuries' accumulations at the present annual rate of increase. A year or two ago we were at an utter loss how to accommodate less than three million titles. Several volumes are now (September 1882) in hand in various stages of progress. The number fully completed and placed in the Reading Room is twenty-two, which comprise the contents of about 70 manuscript volumes, including, with many others, all in letter A after the article Aristotle to the end. They have cost, in round figures, £2450, or about £110 each. Arrangements lately completed will diminish this cost by nearly a sixth, and the sum economised will be available for additional printing. It ought to be stated that all the extra work entailed by printing has been performed by the ordinary Museum staff, with no addition to its resources, except an arrangement by which two gentlemen work two or three hours' overtime.