In the historical development of bibliographies of bibliographies two aspects become especially prominent after 1900. Periodical surveys become a characteristic form of publication and cooperation in the making of bibliographies becomes more frequent or is at least more frequently called for. I shall speak only briefly about periodical publications because they aim at completeness, if they make such an effort at all, only for annual or other limited periods of time. Julius Petzholdt published lists of bibliographies that came to his attention around the middle of the nineteenth century in the Neuer Anzeiger fÜr Bibliographie und Bibliothekswissenschaft. Editors of other journals for bibliography, library science, the book trade, and related fields have published similar lists and have in some instances endeavored more or less successfully to convert them into surveys of current bibliographies of current bibliographical publications. A rapid growth of periodical bibliographies of special fields is characteristic of nineteenth-century scholarship. At the end of the century bibliographers advanced to the stage of compiling annual bibliographies of bibliographies. Such periodical surveys had long been established in fields like theology and classical literature and were now somewhat tardily created for bibliography itself.

The first annual survey of current bibliographical publications seems to be the Bibliographia bibliographica, which appeared in six volumes between 1898 and 1903. Librarians inspired and guided this cooperative enterprise. The list, which includes bibliographies published in non-bibliographical works, is arranged according to the decimal system of classification and was no doubt handicapped by this fact. Since the editors offer a brief outline of the decimal classification in place of a table of contents and provide no alphabetical index of subjects, the Bibliographia bibliographica is not easy to use. The lack of an author index was remedied by the publication of an index for the first two volumes that appeared at the end of the second volume. The Bibliographia bibliographica aroused very little interest among librarians and bibliographers. I have found no reviews of it in the contemporary journals for bibliography and library science. Harvard University Library purchased only the first two issues and these were so little used that, after the lapse of fifty years, they are still unbound.

A second annual survey of the current output of bibliographies is the Bibliographie des Bibliotheks- und Buchwesens, edited by Adalbert Hortzschansky from 1905 to 1925 with an interruption of eight years from 1913 to 1921. This supplement to the Zentralblatt fÜr Bibliothekswesen had a longer and more successful life than its predecessor. It surveyed all publications that fell within the field of the journal and therefore included much more than the bibliography of bibliographies. In 1926 it became an independent publication with a slightly different title but with no change in the subjects reported upon. This Internationale Bibliographie des Buch- und Bibliothekswesens continued to be issued down to the outbreak of war in 1939. An enterprise of somewhat similar scope, the Literarisches Beiblatt der Zeitschrift (later: zum Jahrbuch) des deutschen Vereins fÜr Buchwesen und Schrifttum began to appear in 1924 and continued to 1939. Since these annual surveys include more than the bibliography of bibliographies, I shall not discuss them further.

Three reviews of contemporary bibliographical work have appeared during the last twenty-five years. One of them is limited to bibliographies of a particular kind, and the other two are more or less complete periodical surveys of bibliographical writings. I mention them here as the last examples of the development of periodical bibliographies of bibliographies and as a means by which one can estimate the task of any modern compiler of a bibliography of bibliographies. The first of these, the Index bibliographicus, which first appeared in 1925, offers an interesting example of specialization within the field of bibliographies of bibliographies. The Index bibliographicus is general in scope but cites only bibliographies of bibliographies that appear as current serial publications. In the six years between its first appearance in 1925 and its republication in enlarged and improved form in 1931 the number of currently appearing serial bibliographies rose from 1025 to 1900. Some of these had been overlooked in 1925, but many of the additions concerned bibliographies of bibliographies that had been established during the six years between the two editions. The Index bibliographicus, which was compiled with the assistance of the League of Nations, assumed a more definitely international and cooperative aspect when Joris Vorstius joined Marcel Godet as editor.[182] A third edition of the Index made by Theodore Besterman in 1952 is still larger than either of its predecessors.

The Internationaler Jahresbericht der Bibliographie, which flourished from 1930 to 1940 under the editorship of Joris Vorstius, enables us to survey quickly the current annual production of bibliographies. Critical comments attached to the titles make it one of the most readable bibliographies of bibliographies. Like caviar, the genre is digestible only by those who have acquired a taste for it. The organization of the Internationaler Jahresbericht is skillful, and the comments are judicious and instructive. Since Vorstius was editor of the previously mentioned Internationale Bibliographie des Buch- und Bibliothekswesens as well as the Zentralblatt fÜr Bibliothekswesen, he saw a very large number of bibliographies. He was compelled to hold very carefully to the definitions of his closely related and very similar tasks. The Internationaler Jahresbericht of course lists only bibliographies.

The H. W. Wilson Co. has published the most comprehensive of all periodical surveys of bibliography. Starting in 1938, the quarterly issues are cumulated in annual volumes and these are, in turn, cumulated in volumes for periods of variable length. A cumulation of the bibliographies published in the years between 1937 and 1942 appeared in 1945, and a second cumulation for the years 1943-1946 appeared in 1948. This is the first virtually complete account of current bibliographical production, and the picture is amazing. Between 1937 and 1942 some fifty thousand bibliographies were published. The editors of the Bibliographic Index have classified them in almost ten thousand categories.

The foregoing discussion of these annual or otherwise chronologically limited surveys of bibliography is incidental to the main historical purpose of this essay. Such surveys illustrate very effectively an emphasis which has become characteristic of much modern bibliographical work, and especially of bibliographies of bibliographies, since Gabriel Peignot's book of 1812. Being concerned solely with the current production of bibliographies, they have obviously had no occasion to deal historically with bibliographies or to cite bibliographies published before the limits that they set for themselves. This emphasis on currently useful bibliographical tools goes hand in hand with the cooperative aspect of making bibliographies that I shall stress in this chapter. Already in his RÉpertoire of 1812 Gabriel Peignot had reviewed eighteenth-century bibliography with occasional citations of earlier works that had not been superseded. In 1866 Julius Petzholdt had dealt somewhat more generously than Peignot with Renaissance and seventeenth-century bibliographies, but had scarcely included enough of them to give a picture satisfactory to a historian. Like these predecessors, Joseph Sabin, LÉon VallÉe, and Henri Stein had shown a marked preference for contemporary works. The uses which bibliographies of bibliographies ordinarily serve explain this preference and make it a reasonable one.

The history of mass-production methods in the making of bibliographies has not yet been written. I conjecture that it begins with bibliographies produced more than a century ago by the German publisher Wilhelm Engelmann. His firm continued and revised some bibliographies established by Johann Samuel Ersch (1766-1828), whose scholarly and bibliographical activity began in the eighteenth century. It is not entirely clear whether Ersch himself had already adopted something like mass-production methods. However this may be, the titles and the nature of many bibliographies produced by T. C. F. Enslin (1787-1851) and Wilhelm Engelmann (1808-1878), who made new editions of some of Enslin's bibliographies as well as many of his own, virtually imply such methods. Information about the making of these bibliographies and those of F. A. Wilhelm MÜldener (1830-1900), who seems to have worked in the same way, is difficult to obtain.[183] The compilation and publication of bibliographies by printers and publishers rather than scholars has been continued by such American firms as the Library Bureau (now no longer in existence), R. R. Bowker & Co., and H. W. Wilson Co. These firms have actively supported the making of bibliographies in this country for more than two generations.[184]

An admirable essay, Some Aspects of Bibliography (1900), by John Ferguson (d. 1916) suggests the cooperative aspect that is characteristic of bibliographical studies in the last two generations. Although it is not a full-length bibliography of bibliographies, it reviews the kinds of bibliographies that have been made and appeals to scholars to compile the bibliographies necessary to satisfy the most obvious needs. Ferguson's modest list of some four hundred bibliographies is intended to serve two purposes. It is an effort to show the great variety of bibliographies that have been made and it offers a supplement to Petzholdt's Bibliotheca bibliographica. Ferguson's clear and very instructive classification of bibliographies is as follows: bibliographies according to (1) date; (2) place; (3) printer; (4) material;[185] (5) type;[186] (6) size;[187] (7) illustrations; (8) language; (9) subject; (10) groups of authors;[188] (11) individuals; (12) single books;[189] (13) anonymous books;[190] (14) suppressed books;[191] (15) rare books; (16) general bibliographies.

Some categories of bibliographies might be added to this list. For example, he probably includes bibliographies of private presses in (3). He has no good place for bibliographies of translations, which do not fit easily in the ninth category of bibliographies according to subjects. Nor is there a convenient place for bibliographies of belles lettres according to genres like the novel, the essay, or the book review.

The first bibliography of bibliographies published in the twentieth century is A Register of National Bibliography with a selection of the chief bibliographical books and articles printed in other countries (3 v., 1905-1912) by W. P. Courtney (1845-1913). It is also the first effort of this sort to be made by an Englishman. Like the American Sabin, Courtney limits the bibliographies published in languages other than English to a selection. In the course of twenty years Courtney had accumulated a great many references and four years of work in preparation of the Register greatly increased the number. He acknowledges the assistance of G. L. Apperson, who later published a useful collection of English proverbs, and Robert A. Peddie, who wrote a very large subject bibliography. He has taken references from Henri Stein's Manuel, especially references in the Slavic languages.

He found that the vast number of bibliographies in print made necessary some limitations on the scope of his work. He excludes sale catalogues (although a few are cited), catalogues of manuscripts, and lists of maps and charts. Probably few will quarrel with his decision. He also omits many headings in the bibliography of geology, India, and other unspecified large fields. Here it would be helpful to know more accurately what these omissions were.

Courtney's Register lists some 30,000 titles in a main and two supplementary alphabets. The rapid growth of the material as he proceeded with his work explains this inconvenient division. The very numerous citations of bibliographies in non-bibliographical works contribute to this large figure. I cannot estimate closely the number of small headings, which run into the thousands. The bulk of the Register and the ease with which it can be used make it valuable. As examples of the wealth of information in it, I cite his references to thirteen bibliographies of bacteriology, seventeen bibliographies of hymns, nine bibliographies of insanity, and three bibliographies of swimming. One will rarely leave the Register empty-handed.

Courtney could have greatly reduced the size of his book without a sacrifice of convenience or the loss of significant references. His alphabetical arrangement makes it unnecessary to repeat the headings in the index. He could have profitably used the space saved to add a descriptive word that would differentiate the various works by one author. The dozen references to J. C. Pilling, who wrote bibliographies of American Indian languages, might, for example, have been identified by appropriate adjectives. Courtney could have reduced the size of his book substantially and without loss by omitting bibliographies printed in obvious places, to which one needs no reference. For example, he could have spared references to bibliographies in four editions of Beowulf. The poorly-organized longer articles in the Register are often burdened with miscellaneous or unnecessary information. The article "Bibliography" includes, for example, the universal bibliographies by Georg Draud and Theophil Georgi, (but not Conrad Gesner); Olphar Hamst, Aggravating Ladies, which is a list of pseudonymous books written by "A Lady";[192] the bibliographical journal La Bibliofilia;[193] and a bibliography of church history. This is not a display of good workmanship. The article "Libraries" is a similar farrago of Namur's bibliography of bibliographies, Wheatley's book on how to make a library, Edward Edwards' book on libraries and their founders, Meusel's biographical dictionary of German artists, and other books of as little pertinence. There is very useful information to be gleaned from Courtney's Register and one can easily find it in the chaff.

We are still too close to the latest bibliographies of bibliographies to see them in a true perspective.[194] Efforts to make a comprehensive bibliography of bibliographies continue and a new development that was foreshadowed in Sabin's restriction of his work to a single language with "more than a few" titles added to fill it out is apparent in some less extensive but very excellent bibliographies of bibliographies.

The death of Vilhelm Grundtvig (1866-1950) and the destruction of his collectanea during the war make it certain that the bibliography of bibliographies that he and Joris Vorstius planned will never appear. He had called for help in the enterprise in an article[195] published in 1926 and had obtained approval of it at the international meeting of librarians at Madrid in 1935. During the course of his work he succeeded in gaining the assistance of Joris Vorstius, whom we have learned to know as the editor of annual surveys of current bibliographies. He wrote in 1940 in a review of Theodore Besterman, A World Bibliography of Bibliographies, that his collectanea were arranged according to countries for submission to various workers who might criticize and supplement them.[196] The loss of this compilation is greatly to be regretted because the experience and good judgment of the two editors make it certain that the book would have been comprehensive, well-planned, and satisfactorily executed.

On several occasions Grundtvig stated briefly the task of making a bibliography of bibliographies. His ability to see clearly its difficulties, his wide reading, and his recognition of the many types of bibliographical works that must be considered make these preliminary statements valuable. A long article of 1903 entitled "Gedanken Über Bibliographie"[197] was suggested by A. G. S. Josephson's pamphlet, which will be mentioned later. Here Grundtvig points out the varieties of existing bibliographies and their defects and gives examples of unfamiliar or neglected varieties. For example, he comments (pp. 415-417) on the lists of antiquarian catalogues and catalogues of private libraries. Although we now have more information about these catalogues than Grundtvig found in 1903, we still have no critical bibliographies of them. He points out (p. 418) the unsatisfactory quality of bibliographies of ephemeral publications (chapbooks and the like) and surveys the available lists. He has, to be sure, overlooked one of the earliest of such lists—Giovanni Cinelli Calvoli, Della biblioteca volante (Padua, 1677-1716; 2d ed., 1734-1747. ICN [2d ed.])—but it is rare and virtually unknown. He comments incisively on the lists of collective biographies (pp. 420, 441) and suggests the need for a more critical survey of them.[198] Although Grundtvig's article is not easy reading, it is a very stimulating survey of bibliographies. Any writer of a bibliography of bibliographies should read it attentively.

Grundtvig's pamphlet of 1919, entitled Om Bibliografi og Bibliografier, is much more conveniently arranged than the article of 1903. It is a review of the bibliographical chapter in Svend Dahl, Haandbog i Bibliotekskundskab. He finds it very unsatisfactory and shows how it might have been written. He gives a brief survey of bibliographies in general (pp. 8-10), comments on the making of collectanea and their arrangement (pp. 10-13) and the varieties of bibliographies including those in non-bibliographical works (pp. 14-19), and surveys bibliographies of bibliographies (pp. 19-23), international bibliographies (pp. 23-25), and national bibliographies (pp. 25-29). In keeping with his purpose, he names only the most obvious works.

We come now to the largest of all bibliographies of bibliographies: Theodore Besterman, A World Bibliography of Bibliographies (1939-1940; 2d ed., 1947-1949). Since the two editions do not differ essentially in character, I have found it convenient and probably more helpful to the reader to cite illustrations of Besterman's method from the second edition and to conclude my remarks with brief comment on the changes and improvements made in this edition. The plan of Besterman's book is novel in many details. It is, like Peignot's RÉpertoire of 1812, an alphabetical list of many small headings. Courtney had adopted the same plan in his Register (1905-1912), but few others have seen the great merits of this arrangement. In giving bibliographical details Besterman goes far beyond anything that had been previously attempted. He gives more complete collations than any of his predecessors except Petzholdt had given, and in several regards surpasses Petzholdt. He describes carefully such long sets as the Catalogue of Books in the Library of the Surgeon Generals Office (cols. 1866-1867), the many national bibliographies made by booksellers or librarians, and the annual bibliographies of special fields like The Record of Zoological Literature and its continuation (cols. 3187-3189). He goes beyond Petzholdt and most other bibliographers by estimating the number of titles cited in the books that he lists. He ranges farther afield than any of his predecessors. He includes lists of maps and printed music, registers of documents and charters, and indexes of laws and patents. He includes catalogues of manuscripts and specialized catalogues of books in institutional libraries but not general catalogues of books owned by the same libraries. He includes a generous selection of specialized catalogues of private libraries. He brings more Finnish, Hungarian, and Slavic titles than anyone before him and regrets his inability to include books in Oriental languages. In the first edition he intended to cite all bibliographies printed as books that had appeared before 1936 and succeeded in picking up a large number of those printed between 1936 and 1939. He says that he has cited three times as many bibliographies printed before 1860 as Petzholdt had found. This comparison gives an idea of the amazing extent of Besterman's work.

Besterman sees clearly the difficulties inherent in his choice of an alphabetical arrangement of many small headings and finds perhaps the only answer. It is to offer an abundance of cross-references. Although these headings can be found in special dictionaries that cite synonyms and related words, it remains to be seen how well they will stand the test of time. A suggestion of what may happen is perhaps already to be found in the general unfamiliarity of scholars with these dictionaries. Librarians know and use them, but scholars do not. The time may come when only a specialist and indeed only a specialist acquainted with the history of his discipline will know the meaning of many headings. The headings used in the seventeenth and eighteenth centuries are obsolete. As we have seen, Petzholdt did not recognize the meaning of philosophy that was accepted in the early seventeenth century. I have remarked upon the unfamiliarity of modern students of theology with Peignot's term thÉologie positive. Can we expect future scholars to perceive readily the difference between psychiatry and psychoanalysis? Unless they do, they will not find it easy to use a bibliography of bibliographies arranged as an alphabetical dictionary of many small headings. There is, as a matter of fact, no better illustration of these difficulties than the word "bibliography" itself. This very interesting word needs a historical and lexicographical investigation that will continue Pierre Frieden's article, "Bibliographie. Etymologie et histoire du mot," Revue de synthÈse, VII (1934), 45-52. During this century "bibliography" has been used more and more often to refer to either a study of a book as a physical object or to a list of titles having some common quality. In such titles as Theodore Besterman, A World Bibliography of Bibliographies, and Norman E. Binns, An Introduction to Historical Bibliography (London, 1953), the word has two quite different meanings.

Besterman's enormous compilation does not include all the available bibliographies. Vilhelm Grundtvig found some four hundred titles that had escaped Besterman in the first edition and came regretfully to the conclusion that the hope for a wholly satisfactory international bibliography of bibliographies was now unlikely to be realized.[199] His judgment is severe and the second edition has no doubt gone a long way toward removing these defects. A third edition of the World Bibliography (1955) is now in process, with the first volume already off the press. It will contain some 80,000 titles, an increase of one-third over the second edition.

A few difficulties in Besterman's bibliography concern what have been called "linked" books. These are works having their own title pages but issued in conjunction with another book.[200] Unless they have been catalogued as separate works, they are virtually impossible to identify. We have already seen what annoyance a title of this sort can cause in the case of LabbÉ's bibliography of bibliographies published in 1653.

I cannot reach a decision altogether satisfactory to myself regarding Besterman's inclusion of "abridgments of patent specifications" (I, p. xv). These contain bibliographical information not readily obtainable from any other source, and my disposition is inclined toward generosity. By including them Besterman offers a much more adequate representation of scientific and technological bibliography than would otherwise have been possible. As he correctly says (I, p. xvii), an interest in the fields of humane studies has been predominant in earlier bibliographies of bibliographies. On the other hand, Besterman seems, in including these abridgments, to have stretched to the breaking point his rule for the exclusion of bibliographies contained in non-bibliographical works. If these abridgments are to be included, then one is tempted to call attention to the fact that many German doctoral dissertations offer good bibliographies of small subjects and can be very useful on occasion.[201]

Any definition of a bibliography is difficult to formulate and even more difficult to adhere to. I cite only one more illustration of the problems that arise. Besterman cites (col. 1040) Antti Aarne's catalogue of printed and manuscript versions of Finnish tales. This is clearly within his definition of a bibliography. Perhaps a score of similar catalogues for the tales of countries from Iceland to Rumania are in existence and might equally well have been cited. Although it is not made according to Aarne's pattern, the Typen tÜrkischer VolksmÄrchen (Wiesbaden, 1953) by Wolfram Eberhard and P. N. Boratav is a similar catalogue of tales. How far shall one go in seeking out such extremely technical reference aids as these? The specialist will know them, and few others can use them with any comfort. Students of folklore have done a great deal of indexing and cataloguing and have produced works that can only be separated from bibliographies with difficulty. What shall one say of John Meier, Kunstlieder im Volksmunde (Halle, 1906)? This is a catalogue of German songs that have been heard in oral tradition but can be traced back to known authors. Meier gives full references to the sources in both the printed works of the authors and the collections of folksongs. To return to tales once more, I mention two books of an apparently wholly bibliographical nature that Besterman does not mention. A. C. Lee, The Decameron. Its Sources and Analogues (London, 1909) is, as its title indicates, a compilation of tales related to those in the Decameron. The Anmerkungen zu den Kinder- und HausmÄrchen der BrÜder Grimm (5 v.; Leipzig, 1913-1932) by Johannes Bolte and Georg PolÍvka is a bibliography of parallels to the Household Tales.

Some details in Besterman's work call for comment or correction, but their number is negligible in view of the vast number of titles with which he deals. There are instances in which I should disagree with him in the classification of titles. For example, the Augsburg library catalogue of 1600 (col. 626) is not a bibliography of classical literature but a general catalogue of books and manuscripts in the Augsburg municipal library. Furthermore, its inclusion contradicts the principle stated in the Introduction (I, p. xiv) according to which general catalogues of institutional libraries are omitted. J. B. Mencken's Gelehrten-Lexicon with the second and third editions by C. G. JÖcher is cited (col. 331) as a universal bibliography, but JÖcher's later and much larger revision with its continuations is cited (col. 343) in a different category as a select universal bibliography. Errors in names, place names, and dates appear to be very few. I note that Thomas Cremius (col. 340) should be Thomas Crenius. Such details scarcely call for comment, and their lack of importance is itself a characterization of Besterman's skill. I could wish that the Preface to volume III had explained at greater length the alphabetization of anonymous titles beginning with such words as "Catalogue" or "Index." For example, I cannot find the International Catalogue of Scientific Literature in the Index, although Section N for Zoology is cited in col. 3188. Nor is the International Catalogue in the article "Science" (cols. 2725-2749), where a List of Journals connected with it is cited (col. 2727).

Besterman's bibliography deserves special praise for the fact that it rests on a personal inspection of virtually all the works cited. It needs scarcely to be said that he could have executed this task only at the British Museum or in a few other very great libraries. We can probably infer that Conrad Gesner handled the bibliographies that he cited in 1548, but few later workers have been equally successful in seeing all the books that they name. In this regard Gabriel Peignot made a long step in advance in the RÉpertoire of 1812. The character of his comments makes it clear that he had a firsthand knowledge of the books that he cites. In 1838 his successor Pie Namur yielded countless times to the temptation to cite books from secondary sources, which moreover he does not name. In 1866 Julius Petzholdt established the standards of bibliographical accuracy that ought to be observed and indicated the books that he had not consulted, but his successors have not in general approached these standards. It is therefore altogether gratifying to praise Besterman's attention to such bibliographical details as collations, the citation of editions, and the identification of pseudonyms. In addition to these merits the correctness of the Index calls for particular mention. The bibliographies named in this essay have not always required an index, but those which do contain one have usually served their readers poorly. Philippe LabbÉ was the first but by no means the last of our bibliographers to offer his reader an unsatisfactory index. Even Julius Petzholdt left much to be desired in this regard. The index to Henri Stein's Manuel is conspicuous for its faults. In comparison with his predecessors Besterman's success in the making of an index is all the more meritorious.

In a second edition published between 1947 and 1949, Besterman revised his bibliography to include books printed as late as 1944 and 1945. Since I have referred to this and not to the first edition in my comments, it is sufficient to quote Besterman's statement of the differences between the two editions. The improvements and changes are important to every user of the book but do not affect its nature in any fundamental way.

Large parts of the field have been surveyed anew, the text has been minutely revised throughout, and improvements made. The number of cross-references has been multiplied. Most important, however, are the new entries, which make this edition over 55 per centum bigger than the first. The number of volumes recorded and separately collated is now about 65,000. Nearly all intermediate editions [between the first and last] have now been deleted; they have only been retained, in fact, for bibliographies first published before 1800, and for those of special interest or importance.[202]

And now, five years after the completion of the second edition, Besterman has begun to print a third edition, which he declares to be a "final" edition. As he writes in a letter of July 27, 1954, the new edition will show "the normal increase in size due to the passage of time." A systematic check of Library of Congress holdings has enabled him to strengthen considerably the coverage of American bibliographical publication, north and south. He has also made renewed efforts to improve the representation of scientific and Slavic books. The first volume, which is now in proof, will appear early in 1955 and three more volumes will follow. One lays A World Bibliography of Bibliographies aside with astonishment that one man had the courage to conceive the task and the strength to complete it. In its conception of universality and its success in approaching completeness Besterman's book is a climax in this history of bibliographies of bibliographies.

An emphasis on the current usefulness of the works cited is characteristic of the last four bibliographies of bibliographies that I shall name. These compilations by Bohatta, Funke, and Hodes; by Collison; by MalclÈs; and by Totok and Weitzel are intentionally selective in nature and will therefore require only brief comment. Incidentally, they do not owe their origin to Besterman's suggestion: "it will no doubt eventually become necessary to publish a general bibliography of best bibliographies" (I, p. xvii), although they serve this purpose more or less adequately. The practical emphasis in these four bibliographies betrays the training of their authors in librarianship.

There is a significant difference between the idea of a bibliography as a complete record of books on a particular subject or of a particular kind and that of bibliographies of bibliographies as they have been ordinarily made. To some extent every bibliography serves a practical need, but bibliographies of bibliographies have at first served this need somewhat unconsciously and have served it more and more deliberately as time has passed. As my comments in this essay have shown, the bibliography of bibliographies has been characteristically a tool having immediate practical usefulness. There are few exceptions to this rule: Namur's careless book of 1838, Besterman's book that we have just examined, and Josephson's bibliography of bibliographies of bibliographies that we shall mention at the end of this chapter include works that the compilers regarded as having historical interest rather than practical value. Conrad Gesner named in 1548 a considerable number of classical Greek and Latin works because they seemed to his contemporaries to serve their needs. This aspect of immediate contemporary usefulness has remained characteristic of bibliographies of bibliographies down to the present time. References to classical authorities had still a certain degree of practical value for LabbÉ (1664) and Teissier (1686, 1705). They have disappeared completely or almost completely from later bibliographies of bibliographies. During the last century this emphasis on immediate contemporary usefulness has perhaps expressed itself more clearly in acts than in words. For example, Peignot in 1812 is already looking in this direction. Although subsequent bibliographers may include outmoded books, their eyes turn, as Petzholdt's did in 1866, more and more consciously to modern writings. The four most recent bibliographies of bibliographies recognize fully that they intend to be primarily guides to the best modern sources of information.

In spite of its brave title, Internationale Bibliographie der Bibliographien (1939-1950), this book by Hanns Bohatta (1864-1950), Walter Funke, and Franz Hodes belongs on the level of bibliographies by Durey de Noinville and Michael de San JosÉ. It will only rarely aid either the beginner or the more advanced scholar. It is a selective bibliography and the choice of titles will satisfy no one. Obvious books are lacking[203] and worthless compilations are present.[204] The authors pay little attention to the categories that they set up.[205] The references are incomplete and inaccurate.[206] The comments are often misleading or erroneous. For example, the remark that Giuseppe Fumagalli, La bibliografia, is much less complete than Giuseppe Ottino and Giuseppe Fumagalli, Biblioteca bibliografica italiana, is a fundamental misapprehension of both works. The first is a handbook of general bibliography; the second is a bibliography of bibliographies written in Italian or concerned with Italy. The second book belongs elsewhere and the supplements to it should be cited at length because they were written, in part, by other authors. A comparison of the two books is without point. With all its faults this disorderly Internationale Bibliographie der Bibliographien yields useful information.[207]

The most ambitious and the best of these four modern selective bibliographies is L.-N. MalclÈs, Les Sources du travail bibliographique, I (Geneva, 1950). This deals with general works and cites almost exclusively bibliographies. A second volume (2 pts., Geneva, 1952), which deals with the humanities, has recently appeared. A third volume, which will deal with the sciences, is promised. An abbreviated edition has been published even though volume 3 has not yet been issued. Both the second and the promised third volume are subject bibliographies and therefore need no mention here. The first volume contains some information about books that are not bibliographies, although they are somewhat similar in nature to bibliographies. There is, for example, a very interesting chapter on encyclopedias (pp. 213-224) and another chapter (pp. 225-237) on collective biographies. The wholly practical spirit of Mlle. MalclÈs's endeavor appears clearly in her list of German encyclopedias. She is right in thinking that a modern worker will rarely look at a German encyclopedia older than Ersch and Gruber ("1818-1889, 97 vol. 4^o [A-Z]"). The reference is, incidentally, not quite accurate, since large portions of the alphabet were never written. We hear nothing of the early German encyclopedist J. H. Alsted, who lived and wrote two generations before Louis MorÉri (he is mentioned as the first French author of an encyclopedia on p. 219), or of KrÜnitz and Zedler, who wrote vast encyclopedias almost two centuries ago. Such German works are not appropriate to Mlle. MalclÈs's purpose, but their absence means that her book does not serve a student who wishes to inform himself about the historical development of encyclopedias. In other words, Mlle. MalclÈs has deliberately and successfully satisfied the needs of French scholars.

Mlle. MalclÈs's admirably organized and very rich list of currently useful bibliographies is, as Joris Vorstius says in his review, indispensable to every librarian. Particularly interesting are the introductory remarks in each chapter. These describe the general nature of the works listed and offer comparisons and critical comment on the value and purpose of the different works. This excellent orientation supplements the brief descriptive remarks attached to the titles. As I have already implied, Les Sources du travail bibliographique has been written for French reference librarians. For this reason Mlle. MalclÈs is often content to cite secondary authorities for bibliographies not written in French or concerned with subjects of minor interest to French students. This admirable book stands at the peak of selective bibliographies of bibliographies and is therefore a companion to Besterman's comprehensive work.

Robert L. Collison, Bibliographies Subject and National. A Guide to their contents, arrangement and use (London, 1951) is a pleasant little book containing the information promised in its title. It is a rare example of a bibliography written in a descriptive style that relieves the tedium of a list. The author has intended to offer no more than a brief handlist of currently useful works with some interpretative comments. He has succeeded well in his purpose.[208]

A recently published German counterpart to MalclÈs and Collison is Wilhelm Totok and Rolf Weitzel, Handbuch der bibliographischen Nachschlagewerke (1954). Less comprehensive than the French book and much richer than the English one, it is a meritorious compendium of currently useful bibliographies in all fields. The authors list bibliographies, library catalogues, biographical and biobibliographical handbooks, general and specialized encyclopedias, and treatises of various sorts that contain bibliographical information. Historical and descriptive remarks that are often very instructive introduce the chapters and sections and critical comments usually are appended to the titles cited. The choice of titles will, as the authors no doubt intended, serve best German readers. For example, no Spanish, Latin American, or Russian dictionaries of anonyma and pseudonyma are mentioned (pp. 70-73). I should scarcely agree with the opinion (p. 70) that interest in dictionaries of this sort subsided after the first decades of this century. An emphasis on modern writing often leads the authors to overlook earlier bibliographies that have not lost their usefulness. For example, Mundt's incomplete list (extending only to R) of European dissertations published before 1900 (p. 75) is not "the only means of identifying older university publications (dissertations)." The Catalogus dissertationum academicarum quibus nuper aucta est Bibliotheca Bodleiana MDCCCXXXII (Oxford, 1834) will serve this purpose very well and extends to the end of the alphabet. Bibliographies of university dissertations were, moreover, published in the early eighteenth century. I cannot understand why the authors chose to omit John Meier's enormous bibliography of German folklore in Hermann Paul, ed., Grundriss der germanischen Philologie, III (2d ed., Strassburg, 1909) or why they preferred Wilhelm Pessler's handbook of German folklore to the exclusion of the convenient bibliography in Adolf Spamer, Die deutsche Volkskunde (2d ed. [unchanged], Berlin, 1934-1935). Suggestions of this sort occur readily enough to any attentive reader and are intended to characterize the book rather than to point out its deficiencies. In my opinion, the authors have succeeded well in their intention which was to write a book occupying a position between a bulky guide to information and a beginner's handbook ("Vorwort," p. v).

We have come finally to the last bibliography of all. Its date (1901) entitles it to the first place in this chapter, but it stands last because it is an even more specialized compilation than a bibliography of bibliographies. This bibliography of bibliographies of bibliographies, that is to say, a bibliography in the third degree, is entitled Bibliographies of Bibliographies. The author is Aksel G. S. Josephson, a former member of the staff of the John Crerar Library. It is a chronological list of one hundred and fifty-six bibliographies of bibliographies. The conception is not new, but this pamphlet is the first separate publication of such a list. Similar lists are found of course in Peignot's RÉpertoire of 1812, Petzholdt's Bibliotheca bibliographica of 1866, and a great variety of other reference works. The pertinent sections in handbooks of library science, bibliography, and the like are usually of little interest or value, but Josephson lists them carefully. Perhaps forty titles that he names are significant. He has chosen the strange plan of a chronological arrangement of titles and adds to its inconvenience by providing neither an author nor a subject index. He has yielded to the temptation to include titles of no pertinence like treatises on systems of cataloguing (Nos. 41, 43),[209] H. B. Wheatley's What Is an Index? (No. 59), a guide for making a pastor's library (No. 77), a list of fictitious books (No. 80), and guides to the use of a library (Nos. 67, 69, 104). The many references to bibliographies in the Neuer Anzeiger fÜr Bibliographie und Bibliothekswesen are no doubt pertinent but are scarcely as important as they are numerous. He has probably more references to bibliographical lists published in journals of library science than any other source of information.[210] The value of Josephson's pamphlet lies in an arrangement that makes apparent the historical development and emphasizes the growth of bibliographical lists in journals. Mistakes seem to be few.[211]

In making a second edition of the Bibliography of Bibliographies Josephson profited greatly from the long criticism by Vilhelm Grundtvig that we have already discussed. He replaced the chronological arrangement by a classified arrangement, within which he arranged titles chronologically. He added many new titles that he had found or had excerpted from Grundtvig's criticism. His retirement from active duty and long delays in publication greatly handicapped him in producing a satisfactory piece of work.