Page:Encyclopædia Britannica, Ninth Edition, v. 3.djvu/680

662 F. Rostgaard published in 1697 a Projet tfune nouvdle methode pour dresser un catalogue, which serves to illustrate the difference between arranging a catalogue and a library. Naude finds fault with the far-fetched refinements of his predecessors, and his divisions have the merit of being more simple and precise. &quot;It is certain,&quot; says Mr Edwards, &quot; that a good catalogue will require a much more minute classification than would be either useful or practicable in the presses of a library.&quot; A confusion between these two distinct objects has largely pervaded the &quot;system&quot; of even later writers, who have supposed the same nicety and exactness to be equally necessary and equally practicable in both. Where there is a classed catalogue, the grand objects of a systematic arrangement are sufficiently provided for, independently of the location of the books on which so much fanciful lore has been expended. If there be no classed catalogue, it is tolerably clear that, for purposes of convenient and ready reference, a minute classification of books on the shelves, however accurate, may tend only to bewilder and confuse. Simplicity is the readiest means to aid the memory and abridge the labour of the librarian ; and this object can be attained by a much more elementary division of books than could be tolerated in any classified catalogue extending to details. These remarks apply largely to the ordinary system of modern French bibliographers, the origin of which is variously ascribed to Bouillaud, Garnier, and Martin. Priority of date appears to belong to Bouillaud, for his Bibliotheca Thuana, or sale- catalogue of the famous library of De Thou, had existed in MS. some time before it was edited by Quesnel in 1679. His system embraces five classes, theology, jurisprudence, history, philosophy, and literature, the last including heterodox and miscellaneous works. The catalogue by P. Garnier of the library of Clermont, entitled Systema Bibliotheca: Collegii Parisiensis Societatis Jesu, was published two years previously. The headings embrace 461 subdivisions, of which 74 belong to theology, 88 to philosophy (a class clumsily and confusedly arranged), 227 to history, and 72 to jurisprudence. In 1709 appeared Prosper Marchand s system, developed in his Catalogus bibliothecce Joachimi Faultrier. In his preface he attacks the system of Naude, and, after treating of the different methods, viz., the order of nature, of nations, of languages, of time, and alphabetical, sums up his divisions into theology, or divine knowledge; philosophy, or human knowledge, separated into belles lettres and sciences; and history, or the knowledge of events. Bouillaud s system, as modified by Marchand, was adopted by Gabriel Martin in most of the catalogues, amounting to nearly 150, which he published between 1711 and 1760, and, afterwards, with some enlargement of subdivisions, by De Bure in his Bibliographic Instructive. The result of their successive labours, which is known as &quot; the system of the Paris booksellers,&quot; is the one commonly adopted in France, and consists of theology, jurisprudence, sciences and arts, literature, and history. Some changes, it is true, were afterwards proposed. M. Ameilhon, in a paper published in 1799 in the Memoirs of the French Institute, suggested as primitive classes, grammar, logic, morals, jurisprudence, metaphysics, physics, arts, belles lettres, and history, his Revolutionary sympathies induc ing him to discard theology from the list. But the system, finally elaborated by Martin, survived to govern the classification of the principal libraries in his country. Of the various innovations, the system of Daunou in his Mcmoire sur la Classification des Livres d une grande Bibliotheque, 1800, is frequently cited as the best. Since then the Paris scheme has been modified by bibliographeis like Barbier, Achard, and Brunet ; by M. Merlin in his catalogue of the library of Baron Silvestre de Sacy (1842) ; and by M. Albert, in his Rccherches sur les principes fonda- mentaux de la classification d une Bibliotheque, Paris, 1847. Olenin s system (1808), for the Imperial library at St Petersburg, separated sciences from arts, and introduced philology as a distinct class. Dr Conyers Middleton in 1723 submitted a scheme to the senate of Cambridge for the classification of the university library ; the classes pro posed by him being these theology, history, jurisprudence, philosophy, mathematics, natural history, medicine, belles lettres ^(literce humaniores), and miscellaneous. Hartwell Home s Outlines for the Classification of a Library, based on the Paris system, were submitted about the same time to the Trustees of the British Museum. A serviceable &quot; Scheme for Town Libraries &quot; is embodied in the chapter of Mr Edward s book previously quoted. Of the more general &quot; systems,&quot; based on a survey of the field of human knowledge, and not immediately directed to the requirements of a library, a brief notice must suffice. The earliest system, in this sense of the word, is com monly ascribed to Conrad Gesner, the founder, as Dibdin calls him, of pure bibliography. Yet he was, in fact, pre ceded, however feebly, by Alexo Yanegas, whose work, published at Toledo in 1540, forms the first imperfect type of future efforts of that kind. His divisions are fourfold, viz.: &quot;Original of the harmony between predestination and free will ; Natural of the philosophy of the visible world ; Rational of the function and use of reason ; Revealed of the authority of the Scriptures.&quot; Gesner s, however, was the first comprehensive attempt at a general encyclopedia of literature, constructed in the form of a catalogue. His system was first published in 1548 as an index of matters to his Bibliotheca Universalis, under the title of Pandcctarum sive Partitionum Universaliiim Libri XXI. Florian Trefler, a Bavarian Benedictine, published at Augsburg in 1560 a Methode de classer les Litres, which Peignot describes as &quot;plus que mediocre.&quot; In 1587 ap peared the Tableaux accomplis de tous les arts libcraux, by Christofle de Savigny, which Brunet asserts was the model of Bacon s &quot; Encyclopaedical Tree,&quot; but which was substan tially the system of Gesner. The well-known speculations of Bacon as to the genealogy of knowledge were embodied by D Alembert in his Discours preliminaire a V Encyclo pedic Methodique, Amst., 1767. They were also made the basis of other schemes by llegnault-Warin, Laire, Ferrario, and especially Peignot, whose system was divided into three primitive classes, viz., history, philosophy, and imagination, with the addition of bibliography, as an intro ductory class. Girard s system was embodied in an Encyclopedic ou Dictionnaire raisonne des Sciences, des Arts, et des Metiers, edited in 1751 by Diderot and d Alembert, the latter of whom undertook the part relating to mathe matics. Camus in 1798 took man in a state of nature, and then classed his library in the order in which this &quot; man of nature &quot; is supposed to form his impressions of the universe. The divisions of Thiebaut in 1802 comprised (1), Con- naissances instrumental ; (2), Connaissances essentielles ; and (3), Connaissances de convenances, and were founded on a somewhat similar principle to that adopted in 1822 by the Marquis Fortia d Urban, in his Nouveau Systbme de Bibliographic alphabetique, who prefaced his classes with 