第九十一章《胡适英文论著：中国文学》（3）

  第九十一章《胡适英文论著：中国文学》（3）

  What are the collector's pieces in the Gest collection ?
  Broadly speaking, of the 100,000 volumes that form the original Gest Library, at least forty per cent—about 40,000 volumes—may be properly described as collector's pieces. They are grouped here for the purpose of giving a general picture of the Gest treasures:
  Volumes
  1.Books printed from wood blocks cut in the years
  A.D. 1232-1272 (Sung dynasty editions)700
  2.Books printed from wood blocks cut in the years
  1297-1322 (Yüan dynasty editions)1,700
  3.Books printed in the Ming dynasty (1368-1644)24,500
  4.Manuscripts3,000
  a. Manuscripts copied before 16022,150
  b. Other manuscripts850
  5.The complete set of the 1728 edition of the
  Imperial Encyclopedia in movable copper type5,020
  6.Two copies of The Imperial Palace Movable (wood)
  Type Reprint Series”1,412
  7.A complete set of the Palace Wood-block Edition of the
  Twenty-four Dynastic Histories (printed between 1739 and 1784)754
  8.First editions, rare Palace editions, rare reprints of Sung,
  Yüan, and Ming editions, made in recent centuries (1644-1920)2,000
  9.The Kanjur (translations of Buddhist Scriptures) in
  Mongol (translations from the Tibetan Kanjur), 1772-1790109
  10.Books on medicine and materia medica2,000
  Total41,195
  In short, more than forty per cent of the original Gest-Gillis collection was gathered with the express desire to make it a collector's library—a library of rare and exquisite manuscripts, of notable examples of fine and early printing, and of the oldest and best editions of well-known works of reference, history, religion, philosophy, and literature.
  What is a rare book to the traditional Chinese booklover and collector ?
  To begin with, there are no Chinese incunabula in the European sense of the term. This is, in the first place, because book printing from carved wood blocks began in China so long ago—undoubtedly as early as the eighth century A.D. and very probably even earlier—and its evolution from seal-cutting and from rubbings of stone inscriptions was so gradual that its exact beginnings were never particularly noticed or recorded. And, secondly, because block printing probably had a rather lowly birth, being born of a superstitious belief cultivated by the medieval religions of Buddhism and Taoism that duplicating and distributing a charm or a sacred scripture meant an accumulation of great merit toward the salvation of the soul (either of one's self or of his beloved parents or relatives) after death. Thus the earliest extant specimens of block printing are the meaningless Buddhist tantric charms (The Muku Joko Sutra ) printed in one million copies about A.D. 769 by order of the Japanese Empress Shotoku. And the oldest extant Chinese printed book is the Buddhist Diamond Sutra with a charm at the end, printed in May, 868, by Wang Chieh for free general distribution in memory of his father and mother. Such short pieces, printed in tens of thousands of copies for distribution for religious merit, were usually ignored by early collectors of books, to whom books meant only the beautifully and carefully handwritten scrolls of respectable literature, and to whom such cheaply duplicated short pieces were no books at all. The old prejudice was still strong that good books were for the enjoyment of the qualified few, and not for the masses. There were bookshops recorded as early as the first century A.D. But poets and philosophers never wrote for the general public—and they certainly never intended to have their writings sold for money.
  The famous poet Yüan-chên (A.D. 779-831) wrote in 824 that, for twenty years past, some of his poems and songs and those of his friend Po Chü-i (772-846) had been hand-copied and printed without authorization from wood blocks by enterprising persons in Soochow and Hangchow and that the printed copies were sold in the market place for money or in exchange for wine and food. It was a tribute to the popularity of those two great poets, who, however, would never have thought of condescending so low as to have their collected works cut on wood blocks and printed for sale.
  Indeed, Po Chü-i had his complete poetic and prose works carefully copied in four copies, one to be kept in his own home, three to be deposited in three Buddhist monasteries (one in Lushan in the South, and two in the two capitals of Ch'ang-an and Lo-yang in the North). These four complete copies were made in the years 835, 836, and 839. In 840 he had his latest writings similarly copied for deposit. He was apparently not attracted by the new art of book printing, which had certainly already been making great progress in the preceding century.
  The printed Diamond Sutra of 868, which consists of six sheets of text and one sheet of a woodcut illustration, all neatly pasted together to form a continuous roll sixteen feet long, shows already a highly advanced technique: the text was written in good calligraphy and the cutting and printing were both well done. A contemporary poet and statesman, Ssu-k'ung T'u (who died a martyr's death in 908), wrote a circular letter soliciting funds for block printing a large learned work by a monk-lecturer of Lo-yang. The new art of printing was already attracting the attention of the Buddhist monasteries and lay Buddhists by the second half of the ninth century.
  It was in the tenth century that book printing was taken up seriously and enthusiastically by individuals and by governments. The Imperial Government in the North began in 932 the big undertaking of block printing the Canonical Classics of Confucianism, which took twenty-one years (932-953) to complete. The text was based on the Stone Canon cut on stone in 837, the copies were written by selected calligraphers and checked by competent scholars of the National Academy, and the cutting was done by selected wood carvers. A prime minister of that government, Ho Ning (898-955), who was also a popular poet, copied his collected poetic and prose writings in his own fine calligraphy to be carved on blocks for printing. Several hundred complete sets of his collected works, each totaling 100 chüan in volume, were thus printed for presentation to friends as gifts. It was beneath his dignity to have his works printed for sale, although books printed by the government and the monasteries were for sale.
  Another scholar-statesman of the age, Mu Chao-i (died circa 960), left an instructive story which has become famous in the history of Chinese book printing. When he was a young struggling scholar, he suffered the humiliation of being refused the loan of a manuscript copy of a famous anthology of ancient and medieval poetry and prose. Thereupon, he made a vow to himself that, if he should achieve honors and wealth, he would have that voluminous anthology block printed for the benefit of scholars. He did achieve high honors and became prime minister in the western Kingdom of Shu. He fulfilled his vow by selecting his best students to copy and collate, not only the Wên-hsüan which had been refused him, but also a number of useful encyclopedias, and having them block printed in Chengtu for general circulation. His family preserved the blocks and continued to print from them long after the Kingdom of Shu was conquered by the Sung Empire (965) and it was said that, as late as the early decades of the eleventh century, his descendants still made a comfortable living from the printing and sale of books of the Mu impress, which were then being sold in all parts of the reunited Empire.
  This is the brief history, not of the earliest centuries of obscure and humble beginnings, but of the glorious development of the fine art of block printing in the tenth century.
  Books printed in the first four centuries, from the tenth to the thirteenth, roughly corresponding to the Sung dynasty (960-1276) , and therefore loosely called books of Sung editions, have always been eagerly sought by all lovers and collectors of Chinese books. Their passionate love and search for Sung editions is quite comparable to the Occidental collector's love and search for the European incunabula. And as Sung editions became rarer and rarer with the ravages of time, the collector's zeal for such books rose to even greater heights. When we read the many personal and intimate colophons which the famous collector Huang P'ei-lieh (1763-1825) wrote to each of his rare items of Sung printing, we can still catch the contagion of his enthusiasm and the enthusiasm of his age. Huang often signed himself as A Man Biased in Favor of Sung Editions, and named his library One-Hundred-Sung-Editions-in-One-House. Another famous collector, Lu Hsin-yüan (1834-1894), built a special library to house his Sung editions and named it Pi Sung Lou , The Library of Two Hundred Sung Editions. Such names attest to the zeal and pride of these collectors.
  Books printed in the next dynasty— the Yüan or Mongol dynasty (1277-1368) —were sought by collectors with almost equal fervor. In the nineteenth century, books of Sung and Yüan editions were often purchased at so many silver dollars per leaf.
  But this passion for Sung and Yüan editions (books printed in the tenth, eleventh, twelfth, thirteenth, and fourteenth centuries) constituted only one of the many phases of Chinese book collecting. The vast majority of collectors naturally could not afford to pay for those early editions. The general principle for the wise collector is to seek the best available edition (shan-pên ) for every work. A Sung edition based on a defective manuscript is not so good as a much later edition based on a more perfect text or collated by competent scholarship. As for works of the later ages, the practice has always been to search for the first or earliest available printed edition or an original manuscript copy. During the last three hundred years there has developed what may be called a science of textual collation and criticism. Scholarly collectors no longer regarded early editions as mere objects of antiquarian adoration, but as tools for textual comparison and emendation. They would borrow the best and earliest editions from the wealthy collectors and note down all textual variations on their own copies. Such collated texts are highly valued by booklovers, and very often new editions are made with detailed textual notes by scholarly editors.
  In the last century, the introduction of the photolithographic processes enabled collectors and publishers to reproduce many rare books in facsimile form at prices well within the reach of thousands of readers. Hundreds of very rare editions of classical literature, the voluminous dynastic histories, philosophical works, the dramas of Yüan and Ming dynasties, and others have been photolithographically preserved and made accessible to a larger public.
  I may sum up this brief sketch of the history of Chinese book collecting by a quotation from a contemporary bibliographer, Mr. Ku T'ing-lung, who in the preface to his Illustrated Descriptive Catalogue of Classified Specimens of Ming Editions (Ming-tai pan-pên t'u-lu , 1941) says: Even booklovers of our own day have not been entirely free from the old prejudice of treasuring Sung and Yüan editions far more than the books printed in the Ming [1368-1644] and Ch'ing [1644-1911] periods. … They do not seem to realize that most of the Sung and Yüan editions have been either exactly reproduced in wood-block editions of recent centuries or photolithographically reprinted in facsimile editions, or reprinted with textual emendations and supplements, or indirectly preserved by having their textual variations recorded and published by scholars. How few of them still remain hidden and unknown?
  On the other hand, the books printed in the Ming period [1368-1644] deserve our attention no less than the earlier editions. We are today as far removed in time from the Ming dynasty as the Mings were from the early Sung age. Many Sung editions of important works of classical learning were very well preserved in exact reproductions in the Ming period. And there were important Ming works of great historical value which have never been reprinted and are available only in their original Ming editions. Many of these works (because they happened to contain uncomplimentary references to the Manchus) were often destroyed or suppressed by the Manchu rulers. Surviving copies must be collected and preserved in the interest of history. And, finally, many of the Ming editions, because of their conscious emulation of the best printing of a preceding era, are themselves of such artistic excellence as to deserve the admiration and care of all booklovers.
  What are the distinctive features of this collection made by an American naval officer who trained himself to be a Chinese bibliographer and collector ?
  There is clear evidence that Mr. Gillis made a careful study of the published descriptive catalogues of a number of the well-known libraries and collections, catalogues of books which were to be destroyed or suppressed by order of the Manchu government, indices to the Ts'ung-shu (collectanea), and bibliographical works by recognized scholars. In his Title Index , Gillis listed about two hundred such catalogues and bibliographies, but he indicates that only about twenty-eight of these were frequently consulted by him in his Notes and Title Index .
  Being a shrewd Yankee, Gillis could readily see that it would be sheer madness for him, with his limited allowance from Mr. Gest, to try to compete with Chinese and Japanese collectors in the game of hunting for Sung and Yüan editions, which, by the early decades of the twentieth century, were almost no longer obtainable except through elaborate negotiations with private owners and at prohibitive prices. Yeh Tê-hui (1864-1927) reported in 1911 that he knew of a Hunan collector who paid three thousand silver dollars (at that time equivalent to about three thousand American dollars) for an incomplete copy of a Sung edition of Su Shih's Poems ! So Mr. Gillis made the very wise decision to concentrate his limited resources on collecting Ming (1368-1644) editions, which have the double advantage of being old enough to be rare books, but not old enough (according to Chinese collectors' standards) to command very high prices. This explains why the Gest collection contains about 24,500 volumes of Ming editions. That is to say, almost one quarter of the original Gest-Gillis collection consists of books printed between 1368 and 1644. This is the largest collection of Ming editions outside of China and Japan, and one of the best in the world.
  It is not easy to single out particular items from this huge collection of Ming editions for special mention. A few significant points of general interest may be stressed here.
  In the first place, the collection contains representative specimens of the development of the art of book printing throughout the 276 years of the Ming dynasty. More than a tenth of this Ming collection consists of books (mostly government editions of the Confucianist Canon and Buddhist Scriptures in several different editions) which were printed before the end of the Ching-t'ai reign (1450-1456)—that is, before the Gutenberg Bible. Of this group, there are seventeen volumes of punctuated Buddhist texts printed from blocks cut in 1399, which are unusual rarities in being the earliest Buddhist books using punctuations (punctuated Confucianist Canonical works having begun as early as the twelfth century), and in being probably the only known clearly dated books printed from blocks cut under the reign of Emperor Chien-wên (1399-1402), which reign, having been overthrown by the Emperor Yung-lo (1403-1424), was deliberately obliterated from all documents and publications. These seventeen volumes show that the reign-name was also obliterated from the blocks, but the date year one and the cycle-number chi-mao combine to identify the year beyond question. Such unmistakable marks showing how the tabooed reign-name was cut from the blocks explains why collectors of Ming editions could never find books printed during that ill-fated reign. Mr. Wang Chung-min has also pointed out that there are Ming books in the Gest collection which Gillis acquired without knowing that they might be the only extant copies in the whole world. The work Mr. Wang cited is a collection of poems by Hsü Chung-hsing of the sixteenth century.
  Secondly, this collection includes practically all types of book printing of the Ming period: Imperial Palace editions, editions made by the many royal princes who were well-known as patrons of learning and literature, editions by the two National Academies at Peking and Nanking, editions by provincial and local governments, and books printed by private families and by commercial printers. Of the Imperial Government editions, mention may be made of the most beautifully printed copy (1595) of the Yüeh-lü chüan-shu (Treatises on Music) by Prince Chu Tsai-yü, one of the most learned men of the sixteenth century. Of the books printed by private families and commercial printers, special mention should be given to the many works in the Gest collection which were printed by the great bibliophile and publisher Mao Chin (1599-1659), who, as a private individual living in an age of war, foreign invasion, and change of dynasties, undertook to print and publish in his lifetime a total of about six hundred works, including a large number of very voluminous works, such as the Thirteen Classics with Standard Commentaries and the Seventeen Dynastic Histories. The printing of these books, says Mr. Fang Chao-ying in his brief biography of Mao Chin in Eminent Chinese of the Ch'ing Period , aggregating more than 200,000 double pages, required a large quantity of paper which Mao Chin purchased in Kiangsi province in two varieties—a rather heavy kind known as mao-pien and a thinner kind called mao-t'ai . Both names retain the surname Mao, and the papers…are still so designated in publishing circles.
  In the third place, this collection of Ming editions includes a number of works which the Manchu rulers, for political and racial reasons, had ordered burned and completely prohibited, or partially deleted and banned. At the height of the power of the dynasty, notably in the eighteenth century, this despotic destruction and prohibition of such books was harshly effective, violators being actually punished by capital punishment, and hundreds of books were probably irretrievably lost. Commander Gillis, in common with Chinese collectors of his day, took special interest in collecting items on the official lists of Books to be Burned and Prohibited. Of this group, I like to mention the Ch'u hsüeh chi (first series of collected works) of Ch'ien Ch'ien-i (1582-1664), who was a great scholar and intellectual leader of his age as well as one of the most famous collectors of rare books, but whose works came to be intensely hated by the Manchu Emperor Ch'ien-lung (1736-1795) and were ordered to be burned and destroyed wherever found. The Emperor's wrath was so strong that works of any of Ch'ien's contemporaries containing a preface by him, or correspondence with him, or even a complimentary reference to him, were ordered to be either partially or wholly banned or destroyed. This edition of Ch'ien's earlier collected works was made in 1643, the year before the fall of the Ming dynasty, and, having been copied by the best calligraphers and carved by a well-known artist, is considered by all experts as a most perfect example of the art of block printing of the Ming period. That such a voluminous work in 110 chüan , which could have been secretly kept by a collector only at the most dangerous peril to his life and to his family, has been preserved to this day in perfect condition, is an eloquent tribute to the great courage which the love of good books can inspire in the true bibliophile.
  But Gillis the detective and collector could not suppress his ambition to hunt for the Sung and Yüan editions. He bided his time and waited for his opportunity to satisfy his long-cherished desire. In the meantime, he was training himself to be able to recognize and acquire such early and rare specimens of Chinese printing without having to pay exorbitant prices for them. In other words, he expected to find his treasures among the large loads of books that were often offered at junk prices by impoverished noble families and dilapidated temples and monasteries in Peking and its vicinity.
  His great opportunity came when he found and bought an incomplete set of the Buddhist Canon of Scriptures in apparently an admixture of various printed editions and hand-copied supplementary volumes. The total number of folded volumes was about 5,348. This purchase was probably made in 1926 or 1927. I have not found Gillis' correspondence or notes on the details of this most remarkable find. But Dr. Berthold Laufer, in his preliminary report on the Gest collection, which he had examined at McGill University on July 11-12, 1929, said that the consignment of this Buddhist collection had just reached Vancouver, B.C., and that Mr. Gest had shown him photostats of a number of pages. Dr. Laufer had also been told by Mr. Gest that this collection included 698 volumes printed under the Sung in A.D. 1246 [and] 1,635 volumes printed under the Yüan (14th century) mostly in A.D. 1306. … All this shows that this collection must have been acquired a few years before July, 1929, and that Gillis had already had time to examine and determine these dates.
  It is very interesting that Dr. Laufer was informed that these Buddhist Scriptures were obtained in a remote part of China. Gillis called this collection Ta-pei-ssǔ ching (the Buddhist Scriptures in the Ta Pei Monastery). I have not found any document of his describing the location and history of this monastery, but, judging from the colophons at the end of a number of the manuscript volumes, I have no doubt that this must be one of the monasteries in the city of Peking, which was the capital of the Mongols from 1264 to 1368 and of the Ming dynasty from 1421 to its downfall in 1644. It was quite possible that Mr. Gillis had his reasons for not revealing to anyone where and how he acquired this historic collection. At any rate, he found it expedient not to consult his learned Chinese friends about this precious find, and had to work all by himself on this collection and to rely on only one of his two hundred catalogues for guidance in making his own catalogue of the 5,348 volumes. It happened that Chinese collectors and bibliographers of the old school were not interested in the Buddhist Canons, which were usually too voluminous for the private library and which orthodox Chinese scholars were supposed to view with disdain and contempt. So the one catalogue Mr. Gillis consulted did not give him much information or guidance. He did not even realize that this edition of the Tripitaka, like all other editions, was arranged according to the order of the characters in the Thousand-Character Primer, which every Chinese child could recite by heart. So in his catalogue he followed the order or disorder of the 580 packages that were haphazardly numbered by the monks who sold him the collection! And nobody corrected him.
  Moreover, his secret find was made a few years before the Chinese discovery in 1931 of an almost complete copy of the Chi Sha edition of the Buddhist Canon housed in two Buddhist monasteries in Sian, in Shensi province. General Chu Ch'ing-lan, a leader in Chinese famine relief work and a lay believer in Buddhism, found this collection on one of his relief missions to Shensi. He immediately reported it to his Buddhist friends in Shanghai. A society was organized in October, 1931, to plan the work of reprinting this entire Canon in photolithographic reproduction in greatly reduced size. An editorial committee of this society took up the work of photographing the collection and checking the missing volumes and pages, for which corresponding volumes and pages from other old editions were photographed as replacements. The committee by that time was able to consult the three-volume Catalogue of Catalogues of All Editions of Buddhist Canons, which was published in Japan in 1929-1934 after the completion of the Japanese publication of the Taishō Tripitaka , and which includes a Catalogue of the Buddhist Tripitaka of the Yen-shêng Monastery at Chi Sha in P'ing-chiang-fu—originally printed in A.D. 1234. The editorial committee found that the catalogue of 1234, having been printed at the beginning of the great undertaking, contains only 548 han (cases) numbered by 548 characters, while the Canon at the completion of printing in 1322 included forty-three more han , making a total of 591 numbered cases, comprising 1,532 works in 6,362 chüan , originally bound in about 5,910 folded volumes. A new catalogue of the photolithographic edition was made and similarly printed when the reprinting of the whole Canon was completed in December, 1935.
  As a result of this checking and replacement, this set seems to have been completed about 1602, but it came to Mr. Gillis incomplete. The Canon should have about 5,910 volumes. According to Gillis' count, the Gest copy has only:
  Volumes
  Sung editions697 ?
  Yüan editions1,632 ?
  Ming editions868
  Manuscript recopies2,150
  Total5,348
  In his Notes (No. 2198), Gillis gives a detailed list of all these 5,348 volumes, each volume being marked by S (Sung), Y (Yüan), M (Ming), or W (white paper manuscript). In the same Notes , he makes another catalogue in which all the dated volumes are recorded with the years and months in which the blocks were made. There are fifty-five Sung volumes dated from 1232 to 1271 (the lithographic reprint edition contains one volume dated 1231 and three volumes dated 1272). There are 124 Yüan volumes dated from 1297 to 1315 (the lithographic reprint edition contains two volumes dated 1322).
  With these dated volumes as criteria, Gillis was able to judge 697 ? volumes as of Sung editions and 1,632 ? volumes as of Yüan editions. The remaining printed volumes were of Ming editions. Gillis' judgments are valid in practically all the Sung and Yüan volumes, with only very few cases of oversight or miscount. I have found six volumes dated 1233, 1235, 1238, 1239, 1240, and 1241 respectively which were not noted in Gillis' catalogue of dated Sung volumes. But these overlooked dates only confirm Gillis' judgment that they were of the Sung editions. These nearly 700 volumes of Sung editions and nearly 1,700 volumes of Yüan editions (1,632 ? volumes of the Chi Sha Canon plus a set of a Yüan cyclopedia in sixty volumes printed from blocks made in 1307) make the Gest Library richer in block printed books of the thirteenth and early fourteenth centuries than any other library in Europe or America.
  Although we are still ignorant about where and how Gillis acquired this collection, and although he did not fully understand what he had found, it is time for us to announce that the Gest Library owns one of the two incomplete copies of the now famous Chi Sha Tripitaka of 1231-1322 known to be extant. For the Sian copy, from which the 1931-1935 photolithographic edition was made, is also incomplete, with about 600 volumes originally replaced by corresponding volumes in a Huchow edition of 1332, in addition to other missing volumes and pages which have been replaced by the editorial committee in Shanghai.
  But the editors of the reprint edition informed us in their introductory notes that there were apparently eleven volumes which could not be replaced because their titles were unknown. These are:
  Vols. 3, 4, 9, and 10 of Case 568
  Vols. 1, 2, and 3 of Case 571
  Vols. 7 and 8 of Case 576
  Vols. 8 and 9 of Case 585
  We have checked the Gest copy and can now inform the editors and owners of the reprint edition that, of the eleven missing volumes, the Gest copy has the following seven:
  Vols. 3 and 4 of Case 568
  Vols. 1, 2, and 3 of Case 571
  Vols. 7 and 8 of Case 576
  And the Gest copy also has Volumes 9 and 10 of Case 568, which are the same as Volumes 11 and 12 respectively of the reprint edition. They are not missing, but only wrongly numbered. The only volumes which are missing in both the Gest copy and the Sian copy are Volumes 8 and 9 of Case 585. At a more propitious time, these missing volumes found complete in the Gest copy should be made accessible to China and Japan through photostatic reproductions.
  I have used The Imperial Palace Movable Type Reprint Series to illustrate the scientific technique of Commander Gillis in his book collecting, which is one of the distinctive features of the Gest-Gillis collection. That same microscopic analytical technique he applied also to his identification of the volumes in the Chi Sha Tripitaka and of the tens of thousands of volumes in his huge collection of Ming editions.
  As an Occidental, Gillis was more deeply interested in the movable type editions than is the traditional Chinese bibliophile. It was this interest which led him to make the large collection of Chinese books printed with movable type of wood, lead, and copper (which were used before the invention of modern machine-made metal type). This collection, the largest in the world, includes a number of Ming books printed with movable type as well as the two copies of the wood type Imperial Palace Reprint Series in 1,412 volumes, and the great Encyclopedia (the Ku-chin t'u-shu chi-ch'êng ) of 1728 in 5,020 volumes, all in the original copper type edition. Only sixty-four sets of this Encyclopedia were printed in the original edition. They were given by the Emperor Ch'ien-lung and his successors only to families and persons as rewards for special services to the state. In 1884-1888 a new edition of this Encyclopedia was made in Shanghai with much smaller modern lead type, and 1,500 sets were printed as a private commercial enterprise. In 1890 the Chinese government ordered a lithographic reproduction made of the original edition. One hundred sets were printed of this edition, which was completed in 1898. But it was Mr. Gillis' ambition to get together a complete set of the original copper type edition for the Gest Library. It is said that it took him years of search and minute checking to complete the Gest copy. It is one of the three or four sets of the original edition extant in the world. Because this set in 5,020 volumes was made up of volumes from a number of incomplete lots, Gillis probably paid very little money to have this complete copy—with many spare volumes for possible replacement of missing ones!
  The patient picking of a complete set of this huge Encyclopedia out of numerous scattered and inexpensive volumes illustrates another of the distinctive features of Mr. Gillis' book collecting. The Chinese bibliographer and collector Mr. Miao Ch'üan-sun had spent years patiently selecting odd items of The Imperial Palace Movable Type Reprint Series and in the end made a complete set of it in 812 volumes, all of the original edition. The shrewd Yankee collector could readily recognize in such an enviable collection the principle that odd and inexpensive books could become valuable when they were made parts of a collection selected with a far-sighted design or objective.
  For instance, many of the voluminous works published by the Manchu government and often printed on very good paper were usually ignored by Chinese collectors and their market-price has never been high. Even in the old days, incomplete sets of such bulky works were often sold to dealers by weight, and the pages were sometimes used for paddings in rebinding of old and rare books. (In recent years, such Palace and government editions are being collected in Red China to be delivered to state-controlled paper factories to make pulp.) But Mr. Gillis' love for good editions and his modern sense of values (and the fact that his wife was a Manchu lady) led him to collect for the Gest Library one of the excellent collections of books published by the Imperial Palace and government, including official histories of major military campaigns, collected works of all the Manchu emperors, scientific works in mathematical and astronomical fields, collections of imperial edicts in both Chinese and Manchu, and hundreds of volumes of Manchu translations of Chinese classics, moral philosophy, and popular literature. He also made a complete collection of the Palace edition of the Twenty-four Dynastic Histories in 754 volumes, the block printing of which spread over a period of forty-five years (1739-1784). Mr. Gillis in his early days of book collecting was often criticized for his fondness for Palace editions. I believe that the time is already coming when his vast collection of these official publications of the Manchu dynasty will be classed as valuable rarities of historical importance; and in particular his large collection of Manchu translations of Chinese works (among them the famous pornographic novel Chin p'ing mei —not a Palace or government edition, of course) will be found a very important source of material for the student of the language and literature of this once great people, as well as for historians and anthropologists interested in problems of cultural contact and diffusion.
  Of the large collection of manuscript volumes, there are many items of interest. Our oldest manuscript is a copy of three chapters of a Buddhist sutra, which is one of the many thousands of ancient manuscripts hidden for nearly a thousand years in a cave-library in the desert region of Tun-huang. The manuscript can be conservatively dated as of the sixth century A.D. It has a hemp-cloth wrapper bearing the official seal of the local collection of Buddhist Scriptures. The woman who did the sewing of the wrapper had the inspiration to sign her name, with the date which corresponds to A.D. 685. Another old manuscript—one of the few items bought by Mr. Gest himself in Japan—is a copy of a Buddhist text bearing a colophon which tells us that the copying was ordered by a Japanese empress in A.D. 740.
  Of the 2,000-odd manuscript volumes copied in 1600-1602 for replacements of missing volumes in the Chi Sha Tripitaka, I may cite one colophon of human interest:
  I, Mrs. Chao, née Shen, a devout believer, give fifteen taels of silver for the purpose of copying 100 of the missing volumes of the Sacred Tripitaka with the most devout prayer that my husband, Chao Chih-kao, the Grand Secretary of the Chien-chi Palace [i.e., Prime Minister of the Empire], may be blessed with improved health, that his hands and feet may be restored to smooth functioning, and that our young son, Chao Feng-ko, may be free from all calamities and be blessed with long life and happiness.
  The copies were completed on the sixth day of the Sixth Moon of the 28th year of Wan-li [1600].
  Her husband (whose biography appears in chapter 219 of the Ming Shih ) was Prime Minister from 1594 until his death in 1601, but was confined to his sickbed for about four years before he died. Her pious vow will interest the student of history, of religion, and of the development of book printing. It was the same belief in the merit of duplicating and spreading sacred scriptures—the belief that had been responsible for the origin of block printing in China—that made Madame Chao contribute money for copying the missing volumes. And it will interest the economic historian to know that fifteen taels of silver in 1600 was sufficient to pay the scribes for making careful and exact hand copies of one hundred volumes. The scribe got 0.15 of a tael of silver for copying each volume, which meant at least two whole days' labor. According to contemporary records, the official rate in 1606 was 690 copper cash for one tael of silver, but the market rate was only 450 cash for one tael. So the scribe got about seventy copper cash for two days' labor!
  And, finally, what, after all, is the value of this collection—in nucleus a collector's library—to the trained research scholar in the field of Chinese and Oriental history and culture ?
  In much that I have said above, I have already tried to answer some such question, not with abstract discussions of the real worth of a collector's library to the research scholar, but with concrete examples of how difficult research problems can sometimes be solved with the help of rare and authentic documents and books found only in some collector's library.
  It is unnecessary to defend or apologize for the Gest collection with the often repeated statement that it is really more than a collector's library. Of course it is much more than a collector's library, especially with the thousands of additional volumes acquired in recent decades. But there is nothing wrong for any learned institution to have and take pride in having as many collector's libraries as can be had. Such libraries may remain for months and years without being consulted or utilized. But it is always the dream of a research scholar to find someday in some collector's library the very item for which he has long been hunting in vain. As an old Chinese proverb says: An army is maintained for a thousand days so that it may be used on one particular morning. That is the luxury and the utility of libraries of rare and very rare items.
  The thousands of Tun-huang manuscripts, for example, had been lying idle for decades at the British Museum and the Bibliothèque Nationale. Then there came a Chinese professor who found among them a great many pieces of Buddhist and secular stories retold in popular rhymed recitals hitherto unknown to the historian of Chinese literature. Then came another Chinese professor who found among these manuscripts long rolls containing the recorded discourses of Shên-hui (died A.D. 760), a great Buddhist monk of the eighth century who was the real founder of Ch'an (Zen) Buddhism but whose works had long been lost. He also found in the London and Paris collections other long-lost documents of great historical significance in understanding the development of Buddhism. Such finds have resulted in the rewriting of the history of medieval Chinese literature and the history of Chinese Buddhism.
  It is my sincere belief that the Gest Oriental Library will become a place to which scholars will resort more and more in their hunt for rare and authentic materials needed in their historical researches in various specialized fields. A few of those fields may be mentioned here.
  In Buddhist literature, the Gest Library not only owns 2,300 volumes of Sung and Yüan editions of the original Chi Sha Tripitaka and its Ming replacements, but also 4,000 volumes of the Ming Northern Canon (Pei tsang ), 800 volumes of the Ming Southern Canon (Nan tsang ) and hundreds of volumes in other editions of the Ming period. Serious scholars not satisfied with such modern Japanese reprint editions as the Taishō Tripitaka (a set of which is in the Gest Library) may find it necessary to turn to these Gest volumes for textual collation and authentication. And any student interested in the historical development of Buddhist printing will certainly find it profitable to make use of the Gest Buddhist collections.
  The Gest collection of books on Chinese medicine and materia medica will surely interest scholars investigating the history of Chinese science in general and Chinese medicine in particular. In fact, this collection is now being utilized by an American scholar, Professor W. A. Lessa of the University of California, who is interested in Chinese works on anatomy and physiognomy.
  The Gest Library has an incomplete copy of the Ta-ming shih-lu (Veritable Records of the Ming Dynasty) in 1,492 ch'üan and 173 volumes. This copy, though incomplete, is a good copy once owned by the bibliophile Sung Yün (1681-1760). Another more complete copy of this work in 343 volumes, belonging to the National Library of Peiping, has been microfilmed by the Library of Congress, and positive copies of it are now accessible in a few leading university libraries. A copy of the Ta-ming shih-lu in Nanking was reproduced photolithographically and published in 500 volumes by the puppet regime in Nanking during the Japanese war in China. As this work constitutes one of the most important sources for the study of the history of the entire Ming period (1368-1644), the time will come when trained scholars will want to make a detailed textual comparison and collation of the published set against such available manuscript copies as the Gest copy and the microfilm copy of the National Library of Peiping. (Other manuscript copies are to be found at the Academia Sinica now in Taipei, Taiwan, at the University of Cambridge, and in private collections.) Such a huge work which has remained in manuscript copies for many centuries cannot be safely used without the necessary work of textual collation by competent scholarship.
  I shall conclude this survey by mentioning a few of the intellectual delights which I have personally derived from the collections of the Gest Library. One of the very recent delights was my discovery of a perfectly preserved copy of the first edition of Liao-chai chih-i (which Herbert A. Giles partially translated into English in 1908 under the title of Strange Stories from a Chinese Studio ), by P'u Sung-ling (1640-1715), one of China's greatest writers and story-tellers. As a devoted student of the life and works of this author and as one who once wrote a fifty thousand-word introduction to establish his authorship of an anonymously published great novel, Hsing-shih yin-yüan chuan (A Married Life that Would Awaken the World), I had been searching everywhere without success for a copy of this first edition of his short stories, which was not printed until 1766, fifty years after his death. It was therefore an entirely unexpected pleasure to find it here.
  In 1943 I undertook the retrial of a celebrated historical case involving three great men of the eighteenth century. This piece of historical research took me five years to finish (1943-1948). One of the controversial points required my examination of a manuscript copy of a work by Chao I-ch'ing (1709-1764), one of the three men involved—and I had to have it in the manuscript form as it was copied into the Imperial Manuscript Library of Emperor Ch'ien-lung . I was then in the United States and fully realized that it was absolutely impossible for me to have access to any of the four surviving copies of this manuscript library in 36,000 volumes. It was a great and happy surprise to me when I learned in 1944 that I could borrow from the Gest Library at Princeton a complete exact copy of Chao's work which was originally copied in the early nineteenth century from the Imperial Manuscript Library at Yangchow by the well-known bibliophile Ts'en Yung of that city. I was permitted to keep this copy for a time and compare its twenty thick volumes in detail with the printed editions. The controversial question was settled satisfactorily with the help of this copy, the only one which was available to me in those war years.
  In connection with the same research work, I also wanted to find out the date of a poem which the Emperor Ch'ien-lung wrote in praise of a manuscript copy of the newly emended text of the Shui-ching chu supposedly collated by the scholar and philosopher Tai Chên (1724-1777), one of three great men involved in the case I was retrying. But in the whole United States there was no complete copy of the collected poems of that emperor—except in the Gest Library. I appealed to my friends Dr. Nancy Lee Swann and Dr. T'ung Yiu at the Gest Library, who were kind enough to find the date I wanted to ascertain. The date turned out to be March 27 or 28, 1774, which, however, was beyond even my boldest hypothesis, for I had expected it to be some time in 1775 or very late in 1774. The poem was written eight months before Tai Chên's collated text was presented to the throne in November, 1774! I was therefore forced to conclude that the poem was in praise, not of Tai Chên's text, but of a text collated and edited earlier by someone else. Tai Chên did not arrive in Peking until September, 1773, and could not have completed before March, 1774, the detailed collation and editing of a work which, minus textual notes, comprises some 345,000 words.
  That was in 1945. Many years afterward, I was myself working at the Gest Library and one day I came upon the collected poems of Emperor Ch'ien-lung. I re-examined the poem and its short preface. To my great surprise, I found that the Emperor's preface to the poem contains three important words totally different from the corresponding words on the first page of Tai Chên's text as it was published in 1775 in the Palace Movable Type edition. The preface in the Emperor's collected poems praised the editor for having supplied many missing words in the text—from a few words to as many as eighty or ninety words . That was apparently the original wording in His Majesty's own copy which nobody had the opportunity to change. But in the Palace Movable Type edition published in 1775, the words eighty or ninety words were changed into over four hundred words! The chief editors of the Manuscript Library had apparently taken the liberty to change these words in order that the eulogistic poem might appear to apply to the later and much better collated text.
  So my early conclusion was more strongly verified that the text eulogized by the Emperor in March, 1774, was not the text collated by Tai Chên and presented in November, 1774. The great Emperor, the most conceited and omniscient dictator as he was, never knew the difference! My hero, Tai Chên, apparently so much detested this action of his powerful superiors that he published in the same year his own text in a private edition at his own expense, without the textual notes and without the Emperor's eulogistic poem , as a silent protest. That conclusion solved one of the most baffling mysteries in the celebrated controversy. For those evidences leading to its solution, I gladly give my hearty thanks to Mr. Gillis, who had the good sense to collect the complete works of an emperor who in his long life (1711-1799) and long reign (1736-1796) composed some 42,000 poems which were so poor that no Chinese bibliophile and no other library in the United States seemed to care to collect the six huge collections of his verse totaling 454 chüan .
  Rabindranath Tagore in China
  Free China Review
  Aug., 1961. Vol. 11. No. 8. pp. 19-21.