College and Research Libraries By C H A R L E S F. G O S N E L L Books of Exceptional Age AN alert college purchasing officer once A J L asked a college librarian why he paid premium prices for out-of-print books. T h e purchasing-officer's logic was simple. A n out-of-print book got that way because there was not enough demand to enable the pub- lisher to continue manufacturing and selling the book. If there was lack of demand, why did the local college library want a book which others did not w a n t ? Librarians and booksellers have many answers to this seemingly simple logic. W e all know how desperately a f e w copies of an out-of-print book may be needed. T h e first printing filled most of the demand, and there is not demand enough to justify a second printing. But the simple supply and demand logic is still tantalizing. T h e writer has made certain quantitative studies of book obsolescence in an attempt to show h o w rapidly books go out of date.1 Here he presents a f e w case studies in an effort to determine what is the nature of books of exceptional longevity. T h e difference between the methods (and sometimes the results) of detailed case studies and broad statistical analyses are too well-known to need recounting. T h e emphasis in the previous studies has been on statistical analysis of the ages or imprint dates of titles included in the three select lists of books compiled by Shaw and M o h r - hardt.2 These compilers set out to list good 1 " O b s o l e s c e n c e of Books in College L i b r a r i e s . " Col- lege and Research Libraries 5 : 1 1 5 - 2 5 , M a r c h 1944. " O b s o l e t e L i b r a r y B o o k s . " Scientific Monthly 64:421-27, M a y 1947. 2 S h a w , C h a r l e s B. A List of Books for College Libraries. A . L . A . , C h i c a g o , 1931. . Supplement 1931-1938. A . L . A . , Chicago, 1940. M o h r h a r d t , Foster E . A List of Books for Junior College Libraries. A . L . A . , Chicago, 1937. books for college and junior college libraries. Analysis of the distribution of imprint dates showed a marked preference for new titles and recent editions. T h e Shaw list, however, contains twelve titles ( T a b l e I ) which were over a hun- dred years of age at the time the list was issued. A l l were out-of-print and obtain- able only in the secondhand market; hence, they may be regarded as relatively difficult to secure for a library which did not have them and certainly limited in supply. T h e writer will make no attempt to solve the problem of h o w such a limited supply might be spread among so many institutions seek- ing to meet the specifications of the list. H e can only offer suggestions and specula- tions as to why the titles were included at all. T o a third question, as to why, if these titles are so important and needed by so many libraries, they have not been revised and published in new editions, there seems to be no final answer. By design, all titles in the M o h r h a r d t list were in print at the time of compila- tion. T h e oldest, which was eighty-six years old at the time of compilation in 1936, was M a d a m e de Stael's De VAllemagne, 1850 ( " F r e n c h , " p. 1 1 5 ) . Next in age was Smith's English-Latin Dictionary, 1871 ("Classics," p. 1 1 ) . T h i r d was M i l l ' s Three Essays on Religion, 1874 ( " P h i - losophy," p. 2 3 6 ) . A l l other titles were less than sixty years old, with imprint dates of 1877 or later. Five subjects had no titles more than thirty years old, viz., "Chemistry," " H e a l t h and Physical Edu- cation," " M a t h e m a t i c s , " "Physics," and " P s y c h o l o g y . " JANUARY, 1948 33 T A B L E I Titles in the Shaw List 100 Years or M o r e O l d Subject Page Author Title Date English 146 Fuller History of the Worthies 1811 156 Johnson Rambler 1825 188 Bartram Observations 1751 190 Barlow Columbiad 1813 190 Dennie Lay Preacher 1796 204 Sedgwick Redwood 1824 General 3 i i Camoes Lusiad, tr. by Musgrave 1826 History 402 Roper Life of More 1822 455 Long History of Jamaica 1774 Philosophy 504 Hutcheson Moral Philosophy 1755 Romance Languages 621 Fontenelle Entretiens 1821 624 LeSage Oeuvres 1810 N o n e of the great classics such as the Bible, H o m e r , o r Shakespeare is included a m o n g these older titles. I t is not because the classics are not included but because they are included in later editions. Be- cause they are classics, there is a steady flow of textual revision and commentary, and the preference is f o r the latest. T h e n , too, every age demands reinterpretations of class- ics in its o w n terms. Case Studies T h e s e fifteen older titles seem to rep- resent, then, not the ultimate in eternal values, but the twilight period of obso- lescence. T h e y w i l l be examined in detail as f o l l o w s . First, the f u l l title as specified in the lists w i l l be cited. T h e n there w i l l be a brief bibliographical note on editions. T h e h o l d - ings of the title by the five libraries sur- veyed w i l l be indicated. In a brief comment, the investigator w i l l try to point out the generally accepted opinion as to the importance of the w o r k and w i l l dis- cuss the significance and relationship of the imprint date to the study as a w h o l e . Case J Barlow, Joel. The Columbiad. A poem. W i t h the last corrections of the author. Paris, Printed f o r F. Schoell, 1813. xl, 488p., 2 pi. (incl. f r o n t . ) , 2 port. O u t - o f - print. (8-3871) Note: First published 1807 in luxurious f o r - mat, sponsored by Robert Fulton. Although this 1813 edition is the latest listed by Sabin and the Cambridge History of American Literature, there was a lesser-known edition published in Washington, D . C . , in 1825. Parts have been reprinted in various anthologies. Holdings: None. Comment: Barlow (1754-1812) was poet, patriot, diplomat, and businessman. T h i s heroic epic, in elegant, grandiose, and some- times turgid phrasing, is typical of the rising spirit of nationalism at the time. It is now of interest as an outstanding contemporary document rather than as a lasting piece of literature. History is not complete without it, yet no scholar has yet seen fit to bring out a critical edition. Case 2 Bartram, John. Observations on the Inhabi- tants, Climate, Soil, Rivers, Productions, Animals, and Other Matters Worthy of Notice. Made by Mr. John Bartram, in his travels from Pensilvania to Onondago, Oswego and the Lake Ontario in Canada. To which is annex'd a curious account of the cataracts at Niagara. By M r . Peter Kalm. . . . London, Printed for J. W h i s - ton and B. W h i t e , 1751. viii, 94P., front, ( f o l d . plan). Out-of-print. (1-16152) Note: A reprint, in an edition of 300 copies, was issued at Geneva, N . Y . , in 1895. Holdings: B Comment: Bartram (1701-77) was a noted 34 COLLEGE AND RESEARCH LIBRARIES early American traveler and scientist, with an aptitude f o r clear and homely narrative. T h e title is an important item of early Americana but hardly a classic. In view of the reprint ( 1 8 9 5 ) , it is difficult to regard the earlier date (1751) as of great significance. In his sketch in the Dictionary of American Biog- raphy, D . C. Peattie records his regret that Bartram's letters and biography have been neglected. Case 3 Camoes, Luiz De. The Lusiad, an Epic Poem. T r . from the Portuguese by Thomas M o o r e Musgrave. London, M u r r a y , 1826. xxi, 585P. Out-of-print. (1-IOO94) Note: T h e r e are at least seven other trans- lations, several more recent. T h e latest to be published is that of Fanshaw (1608-66) edited by J. D . M . Ford ( H a r v a r d , 1940). Holdings: Library B has two other transla- tions. Comment: The Lusiad is pre-eminent in Portuguese literature. T h e reason for the Shaw list preference f o r Musgrave's transla- tion is not made clear. But if this translation is required, the volume of this date (1826) is the only one available. T h e format is hand- some and there are copious notes. Case 4 Dennie, Joseph. The Lay Preacher; or Short Sermons, for Idle Readers. Walpole, Newhampshire, Printed by D . Carlisle, jun., 1796. iv, I32p. Out-of-print. ( 7 - 3 7 1 0 ) Note: A second selection was issued in 1817 by John E. Hall. Dennie's Letters were issued in 1936 ( M a i n e University Press). Subsequent to the time this study was originally made (1942-43), a reprint combining the 1796 and 1817 editions was edited by Milton Ellis and published by Scholars' Facsimiles and Reprints in 1943. Holdings: None. Comment: Dennie (1768-1812) enjoyed great popularity during his lifetime and was sometimes called " T h e American Addison." But he was soon eclipsed by Washington I r v - ing. " H i s failure to get his books published, and the neglect of his early biographers have doomed him to an unmerited obscurity," ac- cording to G . W . Fuller in the Dictionary of American Biography. Continued interest seems to be based on his intrinsic merit and his commanding position in early American letters. T h i s interest has resulted in the re- print noted and may someday result in a com- prehensive and critical collection of his essays and other works. W h e n such a volume or volumes are published, the title listed and the reprint will immediately become quite obso- lete. Case 5 Fontenelle, Bernard Le Bovier D e . Entre- tiens sur la pluralite des mondes, augmentes des Dialogues des morts. . . . Nouv. ed., Paris, Bossange, 1821. xvi, 4iop., fold, front. Out-of-print. (5-14163) Note: T h e Bibliotheque Nationale Cata- logue lists 34 editions of the Entretiens up to 1911. M a n y of them lack the Dialogues, however. There are 12 separate editions of the Dialogues listed. T h e r e have been several English translations. T h e Entretiens were not listed as in print in La Librairie Fran- qaise in 1931. T h e Library of Congress de- pository catalog shows the 1821 edition only. Holdings: None. Comment: Fontenelle (1657-1757) was im- portant as an "occasional" writer early in the eighteenth century, and the title in hand is an early scientific work illustrative of the time. It is difficult to attribute special importance to the edition of 1821. T h e Dialogues, which are not available in some of the other editions listed by the Bibliotheque Nationale, are in- cluded by Shaw as a separate title on the same page. Selection of this date may be an accidental one, due to the availability of the Library of Congress card. Case 6 Fuller, Thomas. The History of the IVorthies of England. . . . A new ed., with a few explanatory notes, by John Nichols. London, Printed f o r Rivington, 1811. 2v., front, (port.) Out-of-print. (23-5747) Note: T h e first edition was issued in 1662, the second in 1811, and the last in 1840 ( L o n - don; T e g g , 3 v o l s ) . M o d e r n Selections are available ( O x f o r d , 1928). T h e Library of Congress depository catalog includes cards for the 1811 and 1840 editions. Holdings: A. JANUARY, 1948 35 Comment: Fuller (1608-61) was a royalist and an anglican divine. T h e Worthies is a gazetteer of England, with many short bio- graphical sketches and lists of officials. T h e edition of 1811 is in two large quarto volumes, well-printed. T h e edition of 1840 is in three volumes of more convenient size but inferior paper. Both have indices and footnotes. T h e interest and importance of the Worthies today are aptly characterized, and an interesting sidelight on obsolescence in gen- eral is given in the following words from an introduction to the Selections by E. K. Broadus ( O x f o r d , 1928, p.vii) : Fuller's folios have long since been broken up into numerous volumes; but, even in this more usable form, they still make what Lamb called "massy reading," and to the pleasure-seeker their titles are as forbidding as their bulk. What reader would turn for delectation to six volumes of a Church History or three stout volumes (the Worthies) filled with half-page biographies of persons whose candle flickered out more than three centuries ago? Or who would seek lively reading in a collection of "Characters," illustrat- ing "The Good Wife," "The Good Husband," "The Good Widow," "The Good Landlord," and more, under the heading of The Holy State? Surely these must be (the words are Fuller's) mere "Auxiliary books, only to be repaired to on occasions," or even "such as are mere pieces of formality, so that if you look on them you look through them; and he that peeps through the casement of the index, sees as much as if he were in the house." Case 7 Hutcheson, Francis. A System of Moral Philosophy; in Three Books; Pub. from the Original Mss. by his Son Francis Hutcheson. To Which Is Prefixed Some Account of the Life, Writings, and Char- acter of the Author by the Rev. William Leechman. London, Sold by A . Miller, 1775*. 2v. Out-of-print. (10-32132) Note: Published posthumously and never reprinted. Holdings: None. Comment: Hutcheson (1694-1746) holds a prominent place in the history of Scottish philosophy. H e is not of great interest at present compared, f o r example, with Bishop Butler (1692-1752) whose words are avail- able in modern editions. * T y p o g r a p h i c a l error in S h a w list f o r 1755. Case 8 Johnson, Samuel. The Rambler; a Periodi- cal Paper, Published in 1750, 1751, 1752. London, Jones, 1825. vii, 35ip., front. (port.) (British essayists). Out-of-print. (5-16682) Note: T h e Rambler has been reprinted more than twenty times. T h e latest edition in the British Museum Catalogue is 1876 (2 vols, with a sketch of the author's life by Sir W a l t e r Scott. Selections are available and one is specified by Mohrhardt (ed. by W . H . White, O x f o r d , 1907). Holdings: B, C, have complete editions other than 1825. Comment: T h e importance of Johnson (1709-84) and his Rambler is unquestioned. So many editions, in such quantities, have been issued that sets are easy to obtain, although technically the title is out-of-print. Choice of an edition may rest on format, including legibility, and on availability. T h e date of 1825 is without particular significance but is indicative of the period from which a set must be chosen. Case 9 LeSage, Alain Rene. Oeuvres. . . . Paris, Leblanc, 1810. i6v., fig. $32. ( 6 d.) Note: There are other collections as f o l - l o w s : 1783, 15 vols; 1821, 12 vols; 1823, 16 vols ( o r 32 vols) ; 1828, 12 vols. Complete details are given by H . Cordier, Essai Biblio- graphique, Paris, 1910, and in the Catalogue Generale of the Bibliotheque Nationale. There is no recent or definitive edition such as in the series, Les Grands Ecrivains. M a n y individual titles are available in various edi- tions. Holdings: A has Oeuvres, 1783. Comment: T h e importance of LeSage (1668-1747) in early eighteenth century French literature is unquestioned. A library which wishes a set of his works must get one at least a century old. T h a t a more re- cent and definitive edition of his works has not been published is due doubtless to a com- bination of factors including ( a ) the bulk of his writings; ( b ) the frequent issue of his more popular titles as contrasted with the lack of interest in others; and ( c ) his failure to achieve first rank in French letters. This entry in the Shaw list is unusual in that it is not reproduced from a Library of 36 COLLEGE AND RESEARCH LIBRARIES Congress card and no card number is cited. T h e Library of Congress depository catalog contains cards for the edition of 1821 (Paris, Renouard) and a reprint of 1828 (Paris, L e d o u x ) . Case 10 Long, Edward. The History of Jamaica; or, General Survey of the Antient and Modern State of That Island: with Reflections on its Situation, Settlements, Inhabitants, Cli- mate, Products, Commerce, Laws, and Government. Illustrated with copper plates. London, Lowndes, 1774. 3v. fold, fronts. ( v . 1 - 2 ) . plates (1 fold.) fold, maps, plan. Out-of-print. (2-11573) Note: Never reissued, although the author prepared revision. N o other full history of Jamaica has since been published. Holdings: None. Comment: Long (1734-1813) was highly praised for the work at the time but later condemned it and prepared a revision. Sev- eral brief histories have since appeared, but the island is less the center of interest it once was. Shaw also lists A History of Jamaica . . .to the Year 1872 by W . J. Gardner, new ed., N e w Y o r k , Appleton, 1909, 5iop. For a full account of the history of Jamaica, Long's w o r k is undoubtedly necessary. But, although Sabin in his Dictionary of Books Relating to America describes it as not " e x - cessively rare," it is certainly not generally available. There is need for a comprehensive modern work which in part might be based on Long as a primary source. Case 11 M i l l , John Stuart. Three Essays on Religion. Ed. by Helen Taylor. Longmans, 1874. xi, 302p. $1.80. E16-62. Note: Posthumously published and still in print. T h e original issue was dated 1874. Various impressions have been made since that date. Columbia University has a copy dated 1885, with the designation " T h i r d edi- tion." A copy was purchased from Long- mans Green in N e w Y o r k by Queens College in 1938. T h e binder's title on the spine as well as the half-title is Three Essays on Reli- gion. T h e title page reads Nature, the Utility of Religion and Theism. . . . N e w impression reprinted from the third edition. London, Longmans, 1923. T h e plates with which it is printed are obviously quite worn and appear to be the same as those from which earlier impressions have been taken. Early Ameri- can impressions published by H o l t appear to be from the same plates but have a fourth part, viz., "Berkeley's Life and Writings." Holdings: A (1923) ; B (1884, with "Berke- ley's Life and Writings," p.261-302). Comment: There has been no great oc- casion for revision of this work, and the steady sale has justified occasional new impressions. It is not unusual for a publisher to disguise a new impression from old plates with a new date and sometimes a new title. As M o h r - hardt specifically indicates that all his titles are in print, the 1923 issue described obviously meets his intent, although it does not have the fourth essay. T h e date of 1874 comes from the Library of Congress card. Some- where in the process of checking with the publishers as to whether the book was still in print, there was failure to record the new title and date. T h e date of 1874 appears to be without significance for this study. It might just as well have been 1923. Case 12 Roper, William. The Life of Sir Thomas More, by his Son-in-law, with Notes, and an Appendix of Letters. A new ed., rev. and cor., by S. W . Singer. Chiswick, From the press of C. Whittingham, for R. T r i p - hook, London, 1822. I 9 5 P - , front, (port.) N o w published by Blackie (Library of Golden P r o s e ) , (5-969) Note: As the Shaw entry indicates, the 1822 edition is not required. T h e Life is short and has frequently been reprinted, some- times as part of the introduction to the Utopia. ' It is also included in some editions (among them Everyman's) of Anne M a n - ning's adaptation, The Household of Sir Thomas More. It has also been published under the title, The Mirrour of Vertue in Worldly Greatness. T h e Early English T e x t Society issued a copiously annotated and care- fully collated text in 1935. Holdings: A , in Everyman, no date, and 1935. B, in Sampson's edition of the Utopia, 1910. E, in Selections, 1924. Comment: T h e Life by Roper (1496-1578) was first circulated in manuscript form, and consequently there are the usual problems JANUARY, 1948 37 of textual variations. T h e Guide to Histori- cal Literature ( N e w Y o r k , 1931) lists only the first edition of 1626 (Item L 4 7 1 ) and new editions of 1903 and 1924. T h e Cambridge Bibliography of English Literature lists the editions by Rhys in 1890 and Gollancz in 1902 and 1903, reprinted 1910. T h e 1935 edition should settle most textual problems and of course supersede the 1822 edition by Singer. T h e date of 1822 cannot be regarded as of great significance, although it represents the first issue of one of the better texts. It is not recommended by the leading guides in history and English literature. W i t h the coming of the 1935 edition, the 1822 edition may be considered quite obsolete f o r the college li- brary. T h e designation of the date 1822 may have been an accidental one, incident to the selec- tion of the Library of Congress card, or the other editions as parts of other works, may have been overlooked. Case 13 Sedgwick, Catharine M a r i a . Redwood; a Tale. Bliss & W h i t e , 1824. 2v. O u t - o f - print. (8-11246) Note: Later reprints, with some new ma- terial and author's revisions, appeared in 1850 and 1856. Holdings: None of this title, but some of other titles by the same author. Comment: A conventional romantic moral- istic novel of the day. Sedgwick (1789-1867) was once the most popular authoress in the United States. T h i s title is typical of her w o r k and times. Whether the editions of 1850 and 1856 are superior to that of 1824 is of relatively slight consequence. N o modern editions are avail- able and there is little interest in her w o r k today. Case 14 Smith, Sir William. A Copious and Critical English-Latin Dictionary. By W i l l i a m Smith and Theophilus D . Hall. T o which is added a dictionary of proper names. American Book Co., 1871. xi, 964P., il., 709-54P. $6. [1-4278] Note: T h e problem of this date is similar to that of Mill's Three Essays (Case 11). T h e preface to the dictionary was signed in 1870 and the book published and copyrighted in 1871. So far as is known, there have been few if any changes in the plates since that time. Various impressions have carried later dates, but recent copies sold by the publishers have carried no date except 1870 in the preface and 1871 in the copyright claim. Holdings: A , B, E. Comment: Smith (1813-93) was the pro- digious compiler of a number of encyclopedias and dictionaries on classical subjects. T h e y are standard reference works which have never been superseded. T h e imprint date of 1871 is of doubtful significance for this study. It does indicate the first issue. It does not mark any important distinction in usefulness f r o m a vol- ume with a later date. Specification of this date conflicts with the general Mohrhardt specification that all books listed be in print. T h e book which is in print and now on sale does not carry this date. T h a t the publishers have seen fit to keep this title on the market f o r over seventy years does have implications for the study of obsolescence. T h i s w o r k has no close com- petitors. Its subject is one of declining in- terest, but the volume is indispensable on occasions. A n extensive revision, to incorpo- rate new words in English, if not in Latin, is possible but not likely to be successful commercially. T h a t this title has lived so long unchanged is due in part to its intrinsic merit and in part to diminished interest in the field. T h e lack of interest is insufficient support for competition and revision. Case 15 Stael-Holstein, Anne Louise Germaine ( N e c k e r ) D e . De I'Allemagne. Paris, Firmin-Didot, 1850. 592p., front, (port.) 12 Fr. [36-1364] Note: First published in 1810. T h e r e are many other editions of this title. Several of them, including one published by Firmin- Didot, are listed as in print in La Librairie Francaise, 1931. T h e Firmin-Didot volume is a reissue, with a new introduction, or an edition first brought out by that firm in 1846. Holdings: A and C have another edition (Paris, G a m i e r , 1894). B has (Paris, Charpentier, 1839). Comment: M a d a m e de Stael (1766-1817) rendered her greatest service in introducing German literature and philosophy to the 38 COLLEGE AND RESEARCH LIBRARIES French, and her work is of interest to stu- dents of both French and German. T h e Shaw list calls for a complete set of her works in 17 volumes (1820-21). It is difficult to attach exclusive importance to the date and edition cited here. Others of later date were available and in print when the Mohrhardt list was made. Hence, this date cannot be regarded as of great signifi- cance. Results of Case Studies T h e distinction between the specific im- print date of a given volume or edition and the timelessness of certain authors and their works, has often been emphasized in this study. T h i s distinction is emphasized by the cases which have been examined, and at the same time observation of this distinc- tion is necessary to thorough analysis of the cases themselves. Six of the oldest items in the Shaw list may be said to fall into a group of literary works in a twilight zone of interest, viz., Barlow, Bartram, Musgrave's translation of Camoes, Dennie, Fuller, and Sedgwick. Johnson's Rambler may be included in this group also, although Johnson is clearly of much greater interest today than the six others. It is easy to believe that had the Rambler not already been printed so many times, a modern edition might be available. O f the six, the dates of Fuller and Sedg- wick are open to question within the space of a f e w decades (likewise the Rambler), but whatever the edition chosen, they would still be relatively old. T h e set of LeSage falls into a similar category. T h e case of Fontenelle is similar but the possibility of variation between the imprint dates of 1821 and 1911 is much wider. In their respective fields of " H i s t o r y " and "Philosophy," the works of L o n g and Hutcheson are likewise of minor interest. T h e case of Roper's Life is clearly open to question. Selection of an edition with so early a date may have been an oversight. W i t h i n a f e w years of publication of the Shaw list, however, the edition specified was clearly superseded by a new one. A l l three of the oldest items in the M o h r h a r d t list are puzzling in bibliographi- cal detail. T h e y are cited with their earliest dates of issue, but the books to be bought f r o m the publishers are obviously new im- pressions. T h a t no substantial changes have been made since they were first pub- lished is equally evident. A l l three, es- pecially Smith's Dictionary, are staples by authors of more widely known works. T h e y are in fields and on subjects in which the occasion and demand f o r revision or new works are relatively slight. T h e English-Latin Dictionary is, as Fuller has said (Case 6 ) , one of the "auxiliary books, only to be repaired to on occasions," but when those occasions arise it must be at hand. Holdings of the Libraries Surveyed T h e fifteen titles were not generally found in five college libraries surveyed. Library B was first, with t w o titles as speci- fied and variant editions of five titles. Library A held t w o titles as specified and four in other editions. Library C held t w o titles in different editions. Library E held one specified and one variant title. Library D held none. Six of the titles were not found in any of the libraries. T h e y are among the rarer ones and are not always found in some of the largest research libraries. T w o titles were found once but in different editions. T w o were found in t w o libraries each but in different editions. T w o were found to be held by three libraries ( a total of six copies) but in different editions. T w o titles were found in one library each, in the specified editions. O n e title was held by three libraries. JANUARY, 1948 39 Theoretically, the three four-year col- leges ( A , B, and C ) might have had all of the specified titles in the Shaw list or 36 holdings in all. Actually there were 2, or 5.6 per cent, plus 6, or 16.7 per cent, in different editions. T h u s , in a total of 22.2 per cent of the possible holdings, the title specified or a reasonable substitute was available. These colleges made a better showing in the M o h r h a r d t titles, having t w o out of a possible nine or 22.2 per cent as specified, and 5, or 55.5 per cent, in different editions. O n e of the junior colleges had none in either list, while the other had one title in the M o h r h a r d t list and one f r o m the Shaw list in a different edition. T h a t more of the titles were not found in the libraries surveyed is itself testimony that they are not indispensable. N o n e of the largest research libraries in N e w Y o r k City had them all. Indications of Obsolescence T h e first sign of obsolescence as re- corded in the studies previously referred to was the extremely short life of many books (as indicated by the l o w mean ages) ; the second is, paradoxically, apparent lon- gevity. Books in the first group die quickly and are forgotten. Books in the second live long because there is not enough interest in them to revise them and bring them up- to-date or because there is not enough interest in the field to displace them. T h i s may serve as a partial explanation of the l o w rate of obsolescence in the "Classics," where developments have been slow and interest has been waning. It is fundamental that for a college li- brary books may become obsolete while their authors may not. Repeated new edi- tions are a sign that books may become obsolete while an author and the essence of his works may not. Early editions of Shakespeare are not suitable f o r the use of the undergraduate student. H e is ex- pected to have a modern edition with re- visions in text and orthography and with glossaries and notes. Conversely, though an obsolete book does not make a forgotten author, a forgotten author does make an obsolete title. Some of the oldest titles in the Shaw list fall in the twilight zone of oblivion. Such a case was that of Dennie, whose Lay Preacher essays existed only in two meager selec- tions. A s has been indicated (Case 4 ) , his failure to get more published has doomed him to obscurity. H e was not entirely forgotten, however, as evidenced by the re- print issued in 1943. If this reprint is substituted, this author and title jump from an imprint date of 1796 to 1943 or a mat- ter of 147 years. T h i s later date is testi- mony to greater interest than the earlier date; it confirms what the purchasing agent said about the relation of supply and de- mand in out-of-print books. T h a t the Shakespeare titles were all more recently published than the fifteen titles studied in detail is not an indication that the fifteen have shown greater resist- ance to obsolescence, but quite the reverse. A n d this leads to the proposition that, bar- ring extraordinary commercial vicissitudes, out-of-print volumes are relatively obsolete for the purposes of the college library. W i t h f e w exceptions, if they are not soon revived in newer editions or reprints, they will pass away altogether. Being out-of- print reflects to a degree lack of sales and, consequently, lack of interest. A n d being out-of-print is itself a physical barrier to full expression of interest. Revival of a title in a reprint or new edition, beyond removing this barrier, will not per se do much to keep a dying w o r k alive. Rather the revival must be the result of renewed or continued interest in the author himself and in the lasting appeal of the content of his w o r k . 40 COLLEGE AND RESEARCH LIBRARIES