Petzholdt's adoption of a classified arrangement required him to be very careful in assigning books to categories and to provide abundant cross-references. As we have seen, his subdivisions of categories are not clearly marked and may escape the notice of an experienced user of bibliographies. For example, a bibliography of "Programme" (learned essays issued with the annual reports of German secondary schools) appears (p. 293) properly enough among the bibliographies of German and Swiss publications but few will find it. A few pages later (pp. 298-299) Petzholdt lists bibliographies of German and Swiss journals. Since these two categories are not named in the table of contents or the index, the information is almost completely buried. "Prognostica" or prophecies of future events--a genre of writings that was very popular in the late Middle Ages and the Renaissance--gave Petzholdt trouble. Some of these are listed as pseudo-philosophy (see p. 467: Heuschling), and others are perhaps more appropriately found with almanacs ("Calenderliteratur," pp. 539-540). A bibliography of Swedish almanacs (p. 399) appears in the section for Swedish literature without a cross-reference to or from the bibliography of almanacs. "Loosbücher" or books telling how to interpret omens are in the section for psychology (p. 467), and this is a heading under "Philosophische Litteratur." Examples are wearisome, and I shall give no more. A classified bibliography must have an exhaustive table of contents, a full index of authors, and an adequate alphabetical subject index. Petzholdt's _Bibliotheca bibliographica_ is probably as carefully made as any such book can be made, but its table of contents is a scanty recapitulation of the very largest headings, its index of authors is incomplete, and a subject index is lacking. I have already expressed sincere admiration for the book and feel all the more keenly the presence of these defects. Petzholdt's frequent disparaging remarks show that he did not esteem highly the bibliographical achievements of the sixteenth and seventeenth centuries. We need not defend them here, but we must recognize that his low opinion of them explains many omissions of early bibliographies in his own work. His admirable survey of books that a scholar would find useful in 1866 gives no adequate account of the historical development of bibliography or of the wealth of bibliographical work before 1750. His very convenient chronological arrangement of titles in the various categories does often suggest the historical development and at times his choice of older books is generous.