Of course it will not do to catalogue these severally under Incipit, Hoc, Historia, Ludus, Romans, Sept, Siete, Hienach, Hystorie, Hystory, Sevin, and Siu. In this and other prose and poetical romances of the Middle Ages the heading must be taken in general from the subject of the romance; the name appearing of course in the original language, with all necessary references from other forms. In the present case all the editions would be collected under Septem sapientes,[[32]] with references from Ludus, Sept sages, Siete sabios, Hienach, and Seven, provided the library has so many editions.
Somewhat similarly collections of papers known by the name of a principal contributor or a previous owner or of the house where they were found should be entered under such name, or, if they must be entered under the name of an editor, should have a reference from such name; ex., Dudley papers, Winthrop papers, etc.
A title like “The modern Plutarch” does not mean to imply that the work is written by Plutarchus; such a book would be treated as anonymous, unless it had an editor.
[30] A catalogue of authors alone finds the entry of its anonymous books a source of incongruity. The dictionary catalogue has no such trouble. It does not attempt to enter them in the author-catalogue until the author’s name is known.
[31] For a smaller catalogue this may read “except anonymous works relating to a person, city, or other subject distinctly mentioned in the title, which are to be put under the name of the person, city, or subject.” In the catalogue of a larger library where more exactness (“red tape,” “pedantry”) is indispensable, biography should be the only exception, the place of entry under subjects and under large cities being {40} too doubtful. And in planning a manuscript catalogue, it should be remembered that a small library may grow into a large one, and that if the catalogue is made in the best way at first there will be no need of alteration.
If a book’s title-page is lost, and it is impossible to ascertain what it was from other copies or other editions, or from catalogues or bibliographies, use the half-title or the running title, stating the fact; if it has neither, manufacture a title, within brackets. Such an entry will require many references.
[32] Since this was in type I have come to the conclusion that all these should be entered under Sandabad (Lat. Syntipas), the reputed author of the original Indian romance. But the example will still serve to show the great variety in mediæval titles, and the inconvenience of following a strict first-word rule.
69. A single inscription by an unknown author needs no title-entry, but should have subject-entry under the subject of which it treats, or the name of the place where it is found, or both.
70. When the author’s name is known, it will be enough for Medium to make not an entry under the first word, but a reference from it to the author. The shelf-mark or class-mark should always be given with this reference, that the man who merely wishes to get the book need not have to look in two places for the mark. If there are several editions all the marks should be given, which is not satisfactory unless the imprints are also given, that is, unless an entry is made and not merely a reference.
The entry (or reference) for an anonymous work should be made, even if the author’s name is given in another edition.