from the present difficulty, yet ashamed to allow it to be such, we still more unthinkingly exclaim, We will class our books under their subjects.' Again we begin-on our new plan, and again all goes on well-for a time. At length we open a book which is neither theology, law, history, medicine, astronomy, geology, nor any other wholesome and definite ology. We ponder its contents, in the vain hope of being able to assign to them a local habitation and a name, until, baffled in our endeavours, and desperate of success, we put the volume under the head Miscellaneous;' a mode of procedure just as scientific as would be that of a botanist, who, doubtful of the class of a particular plant, should put it 6 among Weeds.'

Simple as this may seem, it affords a key to the meaning of much that has been done, and done in vain. It will account in some degree for the variety of plans which have been adopted in making alphabetical catalogues of books, and for the numerous abortive attempts at an analytical arrangement of their subjects. Numerous as are libraries, innumerable as are catalogues, scarcely can any two libraries be found which nearly agree in their arrangements, or which are catalogued upon the same principles; and wide indeed are the discrepancies between the various schemes which have been proposed for systematising the whole range of human knowledge. Turn to those Procrustean modes for forcing all the varied imaginations of man's heart within some definite limits the encyclopædias. The most bulky of these huge manuals, that which bears the name of Rees, purports to be alphabetical, yet in it whole sciences are placed under one word; whilst the best of those which profess to be based upon analysis, and which, in the composition of its details, may challenge a comparison with any other, the Encyclopædia Metropolitana,' has no less than six volumes, one-fifth part of its whole, as matter not within the compass of any of its great divisions, nor reducible to system, but arranged alphabetically, that is, as miscellaneous, or, to pursue our illustration, as Weeds.'

It is this residuum, this caput mortuum, which more than all things else has hitherto baffled, as it will continue to baffle, the bibliographers in their attempts at classification.

Of the number of works which have been printed since the year 1450 there exist no sufficient data to enable us to form a certain estimate; and, so far as we know, the Statistical Society have not yet grappled with the subject. The number of volumes claimed to be possessed by the twelve greatest libraries of Europe is as follows: The Bibliothèque du Roi, at Paris, 650,000; Munich, 500,000 (of which one-fifth at the least are duplicates); Copenha* We quote these numbers from the Appendix to the Report on the British Museum.

gen,

gen, 400,000; St. Petersburg, 400,000; Berlin, 320,000; Vienna, 300,000; the British Museum, 270,000; Dresden, 250,000; the Bibliothèque de l'Arsenal, at Paris, 200,000; the Bibliothèque de St. Geneviève, at Paris, 200,000; the Brera Library, at Milan, 200,000; Göttingen, 200,000. These are vague numbers, and, be it remembered, not of works, but of volumes. We may assume with certainty that each of these libraries contains a proportion of its number, perhaps one-tenth, which is not to be found in the other eleven; and we may assume with equal certainty that a vast number of works do not exist in any of the twelve which are to be found in the many libraries of Europe below the number of 200,000. If we take 2,500,000 of works or volumes, to express the number which have been printed—and in our opinion this is far below the actual truth-we find that no library contains much more than a quarter of the books which have issued from the press during the four centuries in which the art of printing has flourished. As there is no published catalogue of any one of these libraries which at all represents its actual state, it is not surprising that such an estimate as we have made should be so vague as it is; but it does surprise us that the amounts of their numbers should also be, as they in fact are, nearly as vague. Whatever difficulty there may be in ascertaining the literary contents of a library, one would suppose it to be a comparatively easy task to ascertain, with some degree of accuracy, the numerical amount of volumes,-a purely mechanical process. But such is not the case; and it is therefore very difficult to institute a positive comparison between any two libraries. At all times tens and hundreds of thousands have been spoken of familiarly. To what is said of the 700,000 volumes in the Alexandrian Library, founded by Ptolemy Philadelphus, we attach just so much faith as we do to the legend of the 11,000 virgins of Cologne. Göttingen Library has been quoted repeatedly by the number of 300,000.* We have now before us, in the writing of the librarian, Dr. Benecke, that in 1835, though it had 300,000 works, it had but 200,000 volumes, the number which we have used in our statement. The Bibliothèque du Roi, at Paris, professes to have 650,000 or 700,000 volumes. Now we have seen the rooms in the Rue Richelieu, from the ground-floor, where the books on vellum, the editiones principes, and the incunabula of the typographic art, are secluded from the profane eyes of vulgar readers, to the showrooms on the first floor, where the public wander and wonder, and the dismal garrets above, full of masses of unbound and uncatalogued books in dire confusion piled:' we have also seen the *In a very inaccurate puff of the Göttingen Library in a defunct publication of the Useful Knowledge Society, the Quarterly Journal of Education, No. iy. B 2

The

British

British Museum Library, and its well-packed and well-ordered shelves, and we find it difficult to reconcile the relative numbers given with the relative space of each library, and to believe that one is less than one-half of the other. Great allowance must be made for modes of enumeration. If every brochure and every pamphlet, and every volume of every novel, every German thesis, and every one of the 60,000 pamphlets on the French Revolution alone, which the British Museum contains, were severally enumerated, as we suspect to be the case in France and elsewhere, the number would be, perhaps, 400,000, an amount which, though large, is still vastly inferior to 700,000. We have lately seen in the newspapers an amusing statement, which we believe to be nearly accurate, that the printed books in the British Museum Library occupy ten miles of shelf. We are not about to give here the mileage, nor the superficial nor the cubic contents of the European libraries; for even if they were measured or squared, or cubed with tolerable accuracy, their relative length, or surface, or bulk, would be no criteria by which to judge of their relative value. Munich might well afford to part with its disposable 100,000 volumes, rejected even of America, for a portion of the collection of a private English gentleman, Mr. Grenville. Our purpose in mentioning these numerical details is, that our readers may be able to form some idea of what a catalogue of books on a large scale must really be. If the number of printed books and brochures in the British Museum be 400,000, the titles or entries would be at least 500,000. In the first volume of the new catalogue, of which we have given the title at the head of this article, we find about 1000 entries or titles under the single name of Aristotle.

At what period catalogues, such as we now mean by the word, were first formed, it would be difficult to determine. Some early inventories of the MSS. possessed by monasteries and by sovereigns serve chiefly to show how small an amount of real learning was within their means, and how few of the great authors of antiquity were known to them in any purity of text. When we read of Chaucer's clerk that

him was lever han at his beddes hed A twenty bokes, clothed in black and red, Of Aristotle and his philosophie,

Than robes riche, or fidel, or sautrie,'

we must remember that they were but translations of translations. The chief purpose of these inventories was the identification of the MSS., which were treasures of great value; and a simple but effectual means to this end was afforded by the practice of quoting a few words from the commencement of the second leaf. The invention

invention of typography, which multiplied copies closely resembling each other, rendered this precaution unnecessary. There was little value attached to a book printed upon paper in rude black type, with coarsely rubricated initial letters, written by a common scribe, as compared with the elegant writing upon fine vellum, gorgeously illuminated in gold and colours, of such magnificent volumes as yet remain to indicate the high point to which the art had reached. There is no dearth of catalogues now— they may be reckoned by thousands.

The rapid multiplication of printed books was followed by various attempts to arrange them, until at length bibliography arose into the rank of a science, and De Bure, Panzer, Maittaire, Placcius, Morelli, Audiffredi, Fontanini, Zeno, Gamba, Barbier, Brunet, Ebert, Watt, Hain, and a host of other writers, have become familiar to all who claim to be bibliographers. The variety of modes into which the authors whom we have just named have cast their labours will somewhat illustrate our point.

De Bure's Bibliographie Instructive is a classed catalogue raisonné of the best books. Brunet's Manuel du Libraire is an alphabetical catalogue raisonné of books selected for their intrinsic value, their rarity, or curiosity, with a classed index. Panzer's Annales Artis Typographica contain an account of early-printed books, arranged chronologically under the places where they were printed; whilst Hain's Repertorium Bibliographicum is a descriptive catalogue of all books printed before the year 1500, arranged in one alphabetical series. The works of Placcius and Barbier are dictionaries of anonymous and pseudonymous books; that of Barbier is arranged on a plan which at first sight appears to outrage common sense, but which has nevertheless deservedly obtained the sanction of recent bibliographers, many of whom have adopted it as the most generally useful. The credit of originating this simple plan belongs to Audiffredi. He arranges anonymous books alphabetically-not under the subject matter, nor under any prominent word, but under the first word of the title-page, whatever it may be, articles and prepositions excepted. Watt's Bibliotheca Britannica is a general alphabetical catalogue of authors, with an index of matters; but it is not, as the title would lead us to suppose, confined to Britain or British authors.

Not

the least remarkable circumstance attending it is this, that it was begun and completed at Glasgow, far from any large public library. It is founded upon the catalogues of the British Museum, the Bodleian, and Sion College libraries, with some others; to which are added many modern books and names from periodical publications, and from catalogues of sales. Such a work repeats of course all the errors of its sources, with the addition of many others

others arising from the mode of compilation. We shall never forget the horror expressed in the countenance and words of the revered Bishop Jebb, on finding himself in Watt's Bibliotheca identified with his uncle, the Unitarian writer. It is, however, one of the most useful books of its class. We have been credibly informed that the labour bestowed upon it proved fatal to the two authors, father and son.

And here we may as well say, that the difficulties necessarily attendant upon the compilation of an alphabetical catalogue of a large public library cannot be estimated by the compiler of a list of an elegant selection' of books. His judgment, if founded upon his own experience, would be as valuable as that of the musician who declined to go to the commemoration of Handel: 'He had played all Handel, choruses and all, upon his flute!' To such we recommend a consideration of the ninety-one Rules for the Compilation of the Catalogue' prefixed to the first volume of the new work published by order of the trustees of the British Museum. Into the consideration of these rules we cannot enter in the course of a brief and rapid review; but they will convince, we think, every one who reads them, that the compilation of a large catalogue requires rare and various knowledge, and much critical discrimination. We are not about to trouble our readers with the beauties of bibliography in the form of quotations from the Catalogue of Printed Books in the British Museum,' nor to make elegant extracts from the some 20,000 or 30,000 title-pages beginning with letter A; but upon a first examination of this volume we were certainly startled by various things which appeared strange to our ideas. We remembered, however, that in order to judge fairly of an author's work, his preface should be considered; and accordingly, on turning to the Rules' which are prefixed, we found that what we so noticed was in conformity with them, and that all objections, except as to merely clerical errors or mistakes in execution, must be brought against the ninety-one rules, and not against the workmen. What most strikes us as objectionable, is the collocation, under one general head, viz., Academies,' of all acts, memoirs, transactions, journals, minutes, &c. of academies, institutes, associations, universities, or societies, learned, scientific, or literary, by whatever name known or designated-as well as works by various hands, forming part of a series of volumes edited by any such society,' arranged alphabetically under places. This produces strange results: it brings together under one head Oxford and Cambridge, the Army Medical Board, the Record Commission, and all the Book Clubsthe Académie des Inscriptions, and the Camden Society, which prints Latin rubbish from manuscripts which it cannot read, with

English