An alphabetical index to the Chinese encyclopaedia
AN ALPHABETICAL INDEX
LIONEL GILES, M.A.
ASSISTANT IN THE DEPARTMENT OF ORIENTAL PRINTED BOOKS AND MSS. AT THE BRITISH MUSEUM
PRINTED BY ORDER OF THE TRUSTEES OF THE BRITISH MUSEUM
SOLD AT THE BRITISH MUSEUM
And by Longmans and Co., 39 Paternoster Row; Bernard Quaritch, 11 Grafton Street, New Bond Street, W.
Asher and Co., 14 Bedford Street, Covent Garden
and Henry Frowde, Oxford University Press, Amen Corner
[All rights reserved]
HORACE HART, M.A.
PRINTER TO THE UNIVERSITY
The Trustees of the British Museum, seeing the need of a full alphabetical index to the great Chinese Encyclopaedia acquired by them in 1877, resolved that this work should be prepared, and entrusted the task to Mr. L. Giles, Assistant in this Department. The methods that have been followed by Mr. Giles in the compilation of this index, together with a history of the Encyclopaedia itself, are fully set forth in his Introduction.
L. D. BARNETT,
Keeper of the Department of Oriental
Printed Books and MSS.
July 11, 1911.
In this latter respect, as well as in size, the 古今圖書集成 Ku Chin T'u Shu Chi Ch'êng marks a notable step in advance. Like its forerunners, this encyclopaedia was designed to provide a comprehensive survey of all that was best in the literature of the past, dealing with every branch of knowledge. It is thus a direct descendant of the Yü Lan. The early history of this great undertaking is obscure. All we know is that the scheme of the work had been conceived and formulated by the Emperor K'ang Hsi 'several decades' before the end of his reign, and that 陳孟雷 Ch'ên Mêng-lei was the man chosen to carry it into execution.  The latter was a scholar of brilliant parts who had been mixed up in the rebellion of 吳三桂 Wu San-kuei and 耿精忠 Kêng Ching-chung in 1674, and subsequently pardoned by the Emperor. For some reason, however, he seems to have incurred the special enmity of Yung Chêng, K'ang Hsi's son and successor; for no sooner was the new sovereign on the throne than he fulminated the following decree of banishment against him (dated January, 1723) :— 'Decree issued to the Grand Secretariat, the Nine Ministries of State  etc.
'Ch'ên Mêng-lei was originally a follower of the rebel Kêng; in spite of which Our Imperial Father, now deceased, refrained in his abounding mercy from putting him to death, but banished him to Manchuria; and on a subsequent tour of inspection, bethinking him that the said Ch'ên possessed some small measure of scholarship, brought him back to the capital and attached him in an official capacity to the household of Ch'êng, Prince of the Blood. As years went on, however, instead of endeavouring to amend his faults, he displayed overweening arrogance and committed numerous acts of lawlessness. In view of the fact that he received grace in former days from Our Imperial Father, We cannot bring ourselves to put him to death; yet, on the other hand, it is certain that he cannot be suffered to remain in Peking. For his late Majesty left behind him strict injunctions as to the consolidation of harmony; and if Ch'ên Mêng-lei continues on the establishment of Prince Ch'êng, he is sure to occasion trouble in the future. Among the Nine Ministers, etc., there are not a few who are personally acquainted with Ch'ên Mêng-lei; if they know of any circumstances which might extenuate his guilt, they must not hesitate to declare the same, and We will then freely pardon him. But if the charges that We bring are true, Ch'ên Mêng-lei and his son must both be banished to some outlying region beyond the frontier. And if there are any adherents of Ch'ên Mêng-lei residing outside the precincts of the capital and stirring up trouble, their names must be reported and their presence notified. There was also one Yang Wên-yen, so-called Minister to the rebel Kêng, who having slipped through the meshes of justice, openly took refuge in Peking, where he published pamphlets and gave rise to seditious talk. He himself has now paid his penalty to the Gods of the Underworld, but if any of his followers should happen to be in the capital, they also must be denounced and driven into exile, and no private considerations should induce you to shelter or conceal them.
'With regard to the work entitled Ku Chin T'u Shu Chi Ch'êng, now lying at the house of Ch'ên Mêng-lei, this was entirely planned under the instructions of the late Emperor my father, who drew up the scheme of the whole, and expended his divine energy thereupon for several decades. He has thus succeeded in linking ancient and modern lore, bringing together classics, history, astronomy and geography, accompanied in every case by illustrations and comments, and including details relating to mountains and rivers, plants and trees, the manifold productions of various arts and industries, and the mysterious inventions of the West. Not one of these things but is treated in full, the result being what may be truly described as a prodigy of literary compilation. As the work of preparing this encyclopaedia is still unfinished, the aforesaid Ministers are hereby directed publicly to appoint one or two scholars of profound learning as editors to complete the task. Seeing that the original draught contains a number of mistakes and other undesirable features, the said editors shall supply the requisite finish, adding or cutting out at their discretion, in order that the work may realise his late Majesty's sublime intention, as a means for delving into antiquity and affording a wide survey of our national literature'. The above Edict, with the wording slightly modified in places, is quoted by 梁章鉅 Liang Chang-chü,  a retired provincial Governor, with these accompanying remarks: 'In my part of the country there is a tradition that the encyclopaedia T'u Shu Chi Ch'êng of the present dynasty was brought to completion under the hands of Ch'ên Hsing-chai,  but the truth of the matter has never yet been investigated. I have respectfully perused the following Edict of the 11th moon  of the 61st year of K'ang Hsi: [Edict] ... It appears, according to this Edict, that the credit of having brought the T'u Shu Chi Ch'êng to a conclusion cannot be exclusively arrogated by Hsing-chai. In reading it, moreover, we gain a general idea of the carelessness and negligence which marred Hsing-chai's talent'. 
These comments of Liang Chang-chü are obviously dictated by a desire not to offend the Manchu Government, in whose eyes, of course, Ch'ên Mêng-lei was a rebel who had died in disgrace. But to those who can read between the lines of an Imperial Edict it is perfectly plain that the T'u Shu was to all intents and purposes complete at the time of K'ang Hsi's death. Not only was the manuscript ready for the press, but the type had been prepared,  and it is by no means certain that the work of printing had not already begun. It is easy to understand Yung Chêng's feverish anxiety to get rid of Ch'ên Mêng-lei at the earliest possible moment, in order that the name of his enemy should not go down to posterity permanently associated with the most memorable literary undertaking of his father's reign. Although he gained his immediate object, it is important to note that as late as 1845 a tradition still existed as to the work having been actually completed by Ch'ên Mêng-lei. As it now became necessary to have some man of straw who could officially report the termination of the work, a new editor was appointed in the person of 蔣廷錫 Chiang T'ing-hsi, a scholar and statesman of some distinction who in his youth had achieved fame as a poet and a flower-painter. The 'polish' (潤色) which, in the terms of the Edict, he was required to impart, cannot in the nature of things have amounted to much, seeing that the Encyclopaedia consists solely of passages taken verbatim from other writings. It may therefore be assumed that his chief function was to throw dust in the eyes of the public and to mark time until the year 1726, when, a decent interval having elapsed, and the imaginary 'mistakes' having been corrected, the work was finally allowed to appear.
The scale on which this new lei-shu was designed is truly stupendous. The total number of chüan or books, which vary but little in size, is 10,000, as compared with the 450 of the encyclopaedia next to it in size, the Yüan Chien Lei Han. The table of contents alone (printed, it must be said, with a generous disregard of space) occupies no fewer than 40 chüan. There are 9 columns to every page, with 20 characters in each column. Allowing an average of 40 leaves or 80 pages to the chüan, we get a total of 800,000 pages,  7,200,000 columns, and (assuming that each page is filled up) 144,000,000 characters. After deducting nearly a third of this number for blank spaces in the text (the columns being frequently broken for new paragraphs, etc.), there would still remain something like 100,000,000 characters in the encyclopaedia.  In spite of this enormous increase of matter, the number of classes or sections was actually reduced from the 55 of the Yü Lan and the 45 of the Lei Han to 32 only. This was no doubt a wise step, as the multiplication of classes tends, after a certain point, to encumber rather than to lighten the work of reference. These 32 sections constitute the practical framework of the encyclopaedia, although they are further grouped under six main categories, corresponding roughly to (1) Heaven, (2) Earth, (3) Man, (4) Science, (5) Literature, and (6) Government;  so that any one who wishes to obtain some idea of the general contents and arrangement cannot do better than run through the sections, one by one.
I. 乾象 means literally 'Celestial signs', and is translated accordingly by W. F. Mayers  'The Heavenly Bodies'. But, besides the heavenly bodies, it includes terrestrial things which were believed to have their origin in the sky. Hence we find here such entries as Wind, Rain, Dew, and even Fire and Smoke. This is the shortest section of the 32, containing exactly 100 chüan, of which 20 are devoted to the fixed stars.
II. 歲功 deals with the seasons and the various festivals of the year. It includes a subhead 幹支, on the so-called 'stems' and 'branches' used in the cyclical reckoning of time.
III. 曆法 (incorrectly translated 'Chronologie' by Klaproth ) has only six subheads in all, but the first of these, on astronomy and mathematical science in general, extends over as many as 82 chüan. A particularly interesting subhead is concerned with the astronomical and other scientific instruments known to the Chinese, especially those introduced by the Jesuits of the 16th and 17th centuries; and the last enumerates the numerical groups or categories so familiar to students of Chinese, such as the Five Elements, the Six Kinds of Grain, etc.
IV. 庶徵 'Various Manifestations' rather vaguely indicates the contents of this section, and Klaproth and Mayers have been misled into translating the title by 'Divination' and 'Natural Phenomena' respectively. As a matter of fact, the phenomena recorded are all of a strange or unusual character, departing from the ordinary course of nature. They include prodigies of various kinds, eclipses, plagues, floods, droughts, dreams, and so forth. These first four sections form a group conceived as relating to Heaven and its manifestations, in contradistinction to Earth on the one hand, and Man on the other. V. 坤輿 is principally taken up with what we should now class separately as mineralogy and geography. The distinction between the realms of Heaven and Earth, just noted, accounts for 'ice' being a subhead in this section, while 'snow' appears in section I. 輿圖, a chronological survey of the topography of the Empire, is a very lengthy subhead which tends of necessity to overlap, in some degree, the two following sections.
VI. 職方, with a total of 1,544 chüan, is by far the longest of all the sections, no fewer than 56 out of its 223 subheads running to 10 chüan and over. It contains a detailed account of all the prefectures and a few other territorial divisions of China, as they existed at the beginning of the dynasty. In the Index the names of these have been transliterated in roman type, with hyphens, and the abbreviations borrowed from Playfair's Cities and Towns of China.
VII. 山川 consists, as the title suggests, of a string of the most notable mountains and rivers of China, besides a few caves and lakes, and concludes with a long subhead on the Sea. The Yellow River (河) claims 20 chüan, and the Yangtse (江) 12. In consulting this section it should be borne in mind that many, if not most, of these mountains and rivers have a plurality of names, some of them more familiar at the present day than those under which they appear in the Encyclopaedia. The transliteration is in italics.
VIII. 邊裔, literally 'borders' or 'frontiers', is a characteristic Chinese term embracing all foreign nations and countries, from the most insignificant tribes to large and important kingdoms such as Korea and Japan. The arrangement is according to the four points of the compass, east, west, south, and north; but the division naturally lacks the accuracy of modern geographical science, Loo-choo, for instance, being included among the southern and not the eastern countries. A special subhead is devoted to little-known nations of each quarter, and a final one to those which could not be located at all. The Chinese names have been identified so far as possible —a somewhat laborious task, in the course of which the researches of Bretschneider, Watters, Eitel, Hirth, Parker, and others have been utilised. Where an identification is doubtful, the name of its author is added in brackets. If no identification could be made, the name is simply transliterated in italics. As cross-references from the Chinese forms are invariably given, some latitude has been deemed advisable in the use of modern names only roughly indicating the location of ancient states. E.g., Cambodia stands for both Chên-la and Fu-nan, neither of which, of course, can have corresponded precisely with it. In like manner, names of towns (e.g. Samarkand) are taken to represent what was sometimes, like 康居, a very large tract of country. This section ranks third in the number of subheads, but excepting the general subhead, only one of them (Korea) covers more than 10 chüan.
IX. 皇極 'Imperial Perfection' treats of the functions and attributes of the Emperors of China. A subhead occupying 162 chüan, that is, more than half the entire section, is devoted to their historical annals.
X. 宮闈 'Palace Doors' is not concerned with the 'Imperial Buildings', which is the translation given by Mayers, but with the various members of the Imperial Palace, beginning with the Empress Dowager and finishing up with the eunuchs. This is the first section containing biographies, which are arranged chronologically under the dynasties, wherever they occur. first section containing biographies, which are arranged chronologically under the dynasties, wherever they occur.
XI. 官常 is a list of the various grades of officialdom in the Empire, from the great feudatory princes (宗藩) down to the Governors, Magistrates, and official underlings. Its 800 chüan constitute the third longest section of the encyclopaedia, thanks chiefly to the vast number of biographies included. It winds up with three subheads of a more general description, containing the lives of loyal ministers, high-principled officials, and great statesmen. These alone fill close upon a hundred chüan.
XII. 家範 treats of the natural degrees of kinship and affinity, and ends with a subhead on Slaves, male and female. Mayers' rendering, ' Domestic Laws', and Klaproth's 'Instructions domestiques', both fail to convey an accurate idea of the contents of the section.
XIII. 交誼 is perhaps best translated 'Social Intercourse', comprising as it does the manifold relations between man and man, the ceremonies and routine of everyday life. Mayers' 'Private Relationships' is less appropriate to this than to the preceding section. A number of subheads appear here which might equally well have been included in section XV. Such, for example, are 'Praise and Slander', 'Calumny and Abuse', etc.
XIV. 氏族 is a list of nearly 4,000 different surnames, single, double, and polysyllabic, arranged according to their tones and rhyme-values, with biographies appended to over 2,500 of them.  The section, then, practically forms a huge dictionary of national biography.  In the Index these surnames are transliterated in roman type, and on account of the paucity of sounds (which may be illustrated by mentioning that no fewer than 33 different surnames are all pronounced Li) they are further distinguished by the four Pekingese tones placed at the right-hand top corner, the entering tone being denoted by an asterisk. For the sake of convenience a large number of uncommon surnames, with no biographies attached and grouped under comparatively few subheads, have been treated as separate entries.
XV. 人事 'Human Affairs' is a favourite category with the Chinese, occurring in almost all their lei-shu, but it cannot be rendered by any satisfactory English equivalent. Mayers has 'Mankind', but objection may be made to this on the ground that it suggests an unintended antithesis to the next section, 'Womankind'. It here includes parts of the body, the stage's of man's life, and the principal states, actions, and affections common to human beings. It is a comparatively short section with short subheads, none occupying more than six chüan.
XVI. 閨媛 'Beauties of the Inner Apartments' is a section devoted entirely to Woman, and may be considered as complementary to section XIV in that by far the larger part of it consists of biographies. These fall into a number of somewhat loosely determined groups, based on the possession of certain prominent qualities, either innate or acquired. Extending over 376 chüan, it forms a repertory of female biography such as no other nation, even at the present day, can make any pretence of rivalling. The subhead 閨節 'Widows who refuse to marry a second time', with its 210 chüan, is in itself the equivalent of a voluminous work, being only exceeded in length by 'Medicine' in section XVII.
XVII. 藝術 is a very comprehensive section, next in point of size to VI, though the number of subheads (43) is not particularly large. The principal ones are: Agriculture, Medicine (a huge assemblage of treatises and extracts from medical literature), Divination, Astrology, Physiognomy and Palmistry, Geomancy, Prognostication and Magic, Painting and Drawing. Sheaves of biographies, as usual, accompany these. Mayers' 'Arts and Divination' is an improvement on Klaproth's 'Arts magiques'. But the underlying idea of the section is really that of 'skilled occupation or profession', a term which covers not only the above but also some further subheads which at first sight appear arbitrarily thrown in: 'Kite-flying' and other games, for instance, are found side by side with 'Bravos and hired assassins', 'Beggars', and 'Actors'.
XVIII. 神異 'Spirits and the Supernatural' may conveniently, and with sufficient accuracy, be set down as 'Religion' — without the addition 'and phenomena', which was a feeble attempt on Mayers' part to translate 異. The section opens with a list of popular deities, and the rest is devoted to various departments of Buddhism and Taoism, with numerous biographies, especially of priests and 'immortals'.
XIX. 禽蟲 is an enumeration of species of the Animal Kingdom, distributed under Birds, Beasts, Fishes, and a fourth class comprising Insects and Reptiles. After those considered important enough for a separate subhead are added 'miscellaneous' and 'strange' (in many cases, fabulous) members of each class. In forming the subheads, more regard has been paid to names than to really distinctive characteristics. Thus, under 鳩 and 雞 we find a lot of utterly dissimilar birds brought together, simply because one of the above characters happens to form part of the names by which they are popularly known. There are no very long subheads, 'Horse' being the only one which exceeds 10 chüan.
XX. 草木 'The Vegetable Kingdom' is even more unscientifically subdivided than the preceding section. Even the following rough divisions are somewhat difficult to make out: (1) Cereals and Vegetables; (2) Flowering Plants; (3) Medicinal Plants: (4) Trees and Shrubs. Here again, as in section VII, plurality of nomenclature is often likely to prove a serious stumbling-block. As a rule, however, it is the recognised classical name which is given in each case, local and other designations being prefixed under the heading 釋名. Great pains have been taken to increase the usefulness of this section to botanists by identifying as many of the plants as possible according to the best two authorities, Dr. Augustine Henry and the late Dr. Bretschneider in his invaluable Botanicon Sinicum. Thanks to their investigations, only a relatively small proportion of the 700 entries have had to remain in their original Chinese form. 'Bamboo', it may be mentioned, with 11 chüan, is the longest of the subheads.
XXI. 經籍 'Canonical and other Literature' begins with an unusually long 'general subhead' (50 chüan), after which nearly three-quarters of the remaining space is devoted to the Classics. Then follow History, Geography, Philosophical Literature, and Belles-Lettres in the order named. The title of each work has been rendered into English and placed between inverted commas. It will be understood, of course, that the actual text of these works is not given in the encyclopaedia, but only various information concerning them. There is only one set of biographies in the present section, namely, Classical commentators and scholars. The lives of other authors and men of letters will be found at the beginning of section XXIII.
XXII. 學行 is curiously rendered by Klaproth 'Commentateurs', and by Mayers, with more justification, 'Education and Conduct/ Education, however, which the heading certainly suggests, is hardly touched upon here, being reserved for section XXV. The things dealt with in reality are philosophy, mental and moral, right conduct, and typical virtues, together with a number of men's lives in which they are exemplified. Important among the subheads are: 理數, a highly compressed and well-nigh untranslatable term, here rendered 'the Universe and its phenomena as seen in permutations of numbers'; 任道 Exposition of Tao in the Confucian sense; 孝弟 Filial piety and fraternal love; 篤行 Lofty conduct; and 隱逸 Retirement from the world. Certain virtues and affections here included, such as 廉恥 'Modesty and sense of shame', would not be out of place if transferred to section XV.
XXIII. 文學 'Branches of Literature' may be regarded as complementary to XXI; but, whereas the latter deals with actual works or large classes of works, the present section is concerned only with literary forms, from Imperial Decrees down to various kinds of poetical composition. Klaproth's 'Éloquence' is quite off the lines, and Mayers' 'The Cultivation of Learning' almost equally bad, unless it be referred solely to the general subhead, which is very long and consists chiefly of biographies of eminent literary men. The next longest subhead, dealing with standard poetry, occupies 46 chüan.
XXIV. 字學 'The Study of Characters' is a title that indicates the contents of the section better than in the two previous cases. 音義, literally 'Sound and Sense', is a lengthy subhead containing the complete text of the Shuo Wên and other philological treatises. Various styles of writing and the art of calligraphy are handled next, tones and dialects also receive attention, and the section closes with writing-materials and implements of the study. Biography is represented by the subhead 書家 Calligraphists.
XXV. 選舉 is the system of educating and selecting men for the Government service. In the earlier subheads, education is dealt with from a more practical standpoint than in XXII, and this is followed by information as to the great competitive examinations. A subhead which does not seem to fit in here very well is 歸誠 Returning to Allegiance.
XXVI. 銓衡 'Weighing in the Balance' is a metaphor which Klaproth so far misunderstood as to translate the characters in the literal sense 'Poids et mesures'. It really refers to the promotion and degradation of public officials, and the section may therefore be regarded as supplementing XI on the one hand and XXV on the other, which latter it even overlaps with the subhead 考課 Examinations. It is one of the shortest of the 32 sections, having 120 chüan distributed among only 12 subheads. The longest of these is 官制 Regulations for Officials. The biographies one might look for here must be sought under section XI. XXVII. 食貨 'Foods and other Articles of Commerce' is a section of even more varied contents than its name would imply. With such subheads as Population, Agriculture and Sericulture, Land Regulations, Land-tax and personal service, Tribute, Textiles, Precious Stones, Wine, Tea, and Money, it covers several categories which in other encyclopaedias are usually found separate.
XXVIII. 禮儀 'Ceremonial Usages' opens with a general subhead on Ceremonies and Music (habitually classed together by the Chinese), although Music is also allotted a section to itself. The chief ceremonies dealt with here are those connected with the attainment of manhood, the celebration of marriage, funerals and mourning, and sacrifices to popular deities who have already appeared in section XVIII. The section concludes with a number of articles of clothing. As in XXVII, there are a good many subheads of considerable length, notably Mourning and Funeral Rites (68 ckuan), Canonisation and Ancestral Worship (25 each), Sacrifices to Heaven and Earth (23), Court Congratulation (18), and Marriage (17).
XXIX. 樂律 'Music' consists mainly in an enumeration of different musical instruments, among which the Lute (琴瑟) holds the premier place. There is a long subhead on Music in general, and another on the Pitchpipes and musical scale.
XXX. 戎政 'Military Administration' includes a gigantic subhead of 132 chüan (longer than several whole sections) on Military History, and another of 52 chüan on Military Organisation. Others that may be mentioned are Imperial Hunting Expeditions and Naval Warfare. The complete text of several noted military treatises is given under 'The Art of War'. The biographies of famous generals will not be found here, as one might expect, but under section XL.
XXXI. 祥刑 'Punishments a Blessing' is a curious heading which could not be properly understood unless recognised as a classical allusion.  'Law and punishment' is perhaps a better description of the contents than Mayers' 'Administration of Justice'. These include, among other subheads, The Penal Code in 80, Thieves and Robbers in 34, and Pardons in 16 chüan. Most of the others are devoted to different forms and instruments of punishment.
XXXII. 考工 was translated by Mayers 'Handicraft', which is certainly an improvement on Klaproth's 'Ouvrages publics'. But, as usual with these headings, it is impossible to find any single term that will fit the contents exactly. There are 154 subheads —more than in any other section except those enumerating zoological, botanical, or proper names —but none of them is of any great length. The longest, it is rather surprising to find, is that on 儀仗 'Insignia carried in processions', including all sorts of banners and emblems (15 chüan), while 'Cities' and 'Palaces' only claim 14 each, 'Vehicles' coming next with 12. These examples may serve to give an idea of the scope of the section, which deals first with industries and handicrafts, and then with the products of these industries, ranging from bridges and parks to umbrellas and lanterns, and including a particularly large and varied assortment of vases and jars. It does not clash with section XXVII, as that treats of raw materials, while this is concerned only with the manufactured article. The 部 pu or subheads, it will be noted, vary so much in length, that the true magnitude of the Encyclopaedia can hardly be gauged from an Index like the present, which consists for the most part in a mere enumeration of these. As the Chinese editor puts it, 'One pu may comprise tens or even hundreds of chüan; one chüan may also contain ten or more pu'. A few additional statistics, therefore, may fittingly be given here. Out of the total number of subheads, there are no fewer than 208 which comprise 10 chüan and upwards; of these, 124 have between 10 and 19 chüan, 38 have between 20 and 29, 23 have between 30 and 49, 16 have between 50 and 99, 5 have between 100 and 199, one has 210, and one the colossal total of 540 chüan, or something like 43,200 pages. This last subhead, 醫 Medicine, is again minutely subdivided, 5 of the sub-subheads containing between 10 and 19 chüan, 2 between 20 and 29, and 2 between 40 and 50. Passing to the other extreme, we find a number of subheads which are extremely short, sometimes occupying only a column or two. Such occur especially among the uncommon surnames and foreign countries about which next to nothing is known. So many of the subheads being of such prodigious length, and yet hardly susceptible, like Medicine, of further subdivision according to subject, some other expedient was obviously necessary in order that this enormous bulk of matter might be rendered more approachable for purposes of reference. What the editors did was to introduce under each subhead a systematic classification of a novel kind. In previous compilations little or no distinction had been made between the various sources laid under contribution, and the extracts appearing under each subhead had been jumbled together without any attempt at marking their relative value or authenticity. This defect was now remedied by the expedient of arranging the material in certain fixed groups, occurring in the following invariable order: —
(1) 匯考 Hui k'ao. The principal series of extracts from standard works of every description, arranged chronologically so far as the nature of the subject permits. Included in this division are also (a) 圖 T'u, pictures, maps and diagrams, about which a word will be said presently; and (b) 表 Piao, chronological and other tables.
(2) 總論 Tsung lun. Disquisitions of a general character, or constituting an all-round treatment of the subject. These are also drawn from the most authoritative sources, and form the principal group in cases where a chronological survey is not admissible. As a rule, however, they represent opinions rather than facts, and may. therefore be considered as of less importance than the Hui k'ao.
(3) 列傳 Lieh chuan. Biographies. These are indicated by the letter 'b' in the present Index. 
(4) 藝文 I wên. Elegant compositions, especially pieces in the poetical prose known as 詞. The shorter specimens are reproduced entire, from the longer ones only the cream is extracted. Generally speaking, in the more ancient pieces, down to the T'ang dynasty inclusive, fullness is aimed at, while from the Sung onwards there is more abbreviation. (5) 選句 Hsüan chü. Very short excerpts of peculiar literary excellence, which have an epigrammatic turn or otherwise deserve to be singled out and handed down as household words.
(6) 紀事 Chi shih. Minor historical and other passages not deemed of sufficient importance to find a place in the Hui k'ao. A multifarious collection of notes, often of an anecdotal character.
(7) 雜錄 Tsa lu. Miscellaneous extracts from the Classics and from general literature which (a) lack authenticity and therefore cannot be included in the Hui k'ao; (b) treat the subject in a cavilling and one-sided spirit, and are therefore not suitable for the Tsung lun; (c) are not sufficiently polished to be admitted among the I wên.
(8) 外編 Wai pien. Passages of fantastic, extravagant or allegorical import derived from Buddhist and Taoist literature, or other schools of heterodox philosophy.
These divisions, of course, do not all occur in every subhead, but in the majority of cases one will find at least Nos. 1 or 2, 4, 6 and 7. Thanks largely to this method of classification, the T'u Shu will be found at least as easy to consult as the T'ai P'ing Yü Lan or the Yü Hai. Another improvement contributing to the same result is the detailed synopsis of contents prefixed to every chüan, in which all the works or authors quoted from are set forth, with further indication of chapters and illustrations.
No descriptive account of the T'u Shu Chi Ch'êng would be quite complete without some mention of the illustrations, which, as may be gathered from the title of the encyclopaedia, were intended to form one of its distinctive features. The word 圖 includes not only pictures properly so called, but also maps and diagrams, all of which are profusely scattered throughout the work. Thus, section I contains star-maps and astronomical diagrams, III drawings of scientific instruments, VI geographical maps, VII pictures of nearly all the mountains described, and VIII sketches of the natives of various foreign countries. (It may be remarked that little-known or fabulous countries are those which are most liberally illustrated.) The 明倫 category affords no scope for pictorial representation, but when we come to section XVII, we find illustrations to some of the games, and diagrams under the subhead 星命 Astrology; in XVIII, miscellaneous deities are portrayed, and in XIX and XX nearly every animal and plant is accompanied by a picture, which often proves of no little help in the task of identification. Caution must be exercised, however, as these plates are sometimes not only crude in execution, but grotesquely at variance with the text. Under the heading Bear, for instance, not one of the three plates can be said to resemble the animal in question. In XXIV, inkslabs are depicted, in XXVIII, articles of clothing, in XXIX, an array of musical instruments. Various engines and implements of war are represented in XXX, and miscellaneous manufactured articles—vases, banners, insignia, etc.—in XXXII. The above is very far from being an exhaustive list of the illustrations in the T'u Shu, which if massed together would fill quite a collection of stout volumes. Just as there are no original articles written specially for this compilation, so the plates have been taken from a number of earlier works, among others the well-known illustrated encyclopaedia 三才圖會 San Ts'ai T'u Hui.
With regard now to the printing of the T'u Shu, perhaps the greatest typographical feat that the world has seen, movable types, instead of the usual wooden blocks, are known to have been employed. But one 吳長元 Wu Ch'ang-yüan states further, in a small guide-book to Peking which was published towards the end of 1788, that the types were cast in copper. These are his words: —
'The Wu-ying Palace, where sets of movable type are kept, stands on the east side of Pei-ch'ang Street outside the Hsi-hua Gate. My investigations show that these movable types were formerly cast in copper and produced in the reign of K'ang Hsi for the purpose of printing the T'u Shu Chi Ch'êng ... . After an interval of many years, these copper types were found to be defective by more than half the original number, and in 1773 they were replaced by wooden type for printing the books in the Imperial Library, after a pattern suitable for engraving. This set of type received the name of 'Gem-collection', a poem being composed by the Emperor on the occasion'. 
The poem here mentioned took the form of a short eulogy of the undertaking, and it was reprinted in each subsequent work produced by the same means. In a note appended to one of the lines it is distinctly stated that 'when the T'u Shu Chi Ch'êng was compiled in the reign of K'ang Hsi, the characters were engraved in copper (刻銅字)'.  The examination of a few pages of the British Museum copy can leave very little doubt as to which of these two conflicting assertions is correct. Mr. Alfred Pollard, Assistant Keeper in the Department of Printed Books, who has been good enough to give me his opinion on the matter, points out that the differences readily discernible between examples of the same character practically exclude the possibility of our having here to do with a fount, that is to say, the types must have been cut, not cast in matrices. And M. Émile Blochet, who has kindly examined the copy in the Bibliothèque Nationale at Paris, speaks in very much the same sense.
The actual number of types thus cut by hand is stated by Julien to have been 250,000. Dr. Macgowan,  on the other hand, puts it down as 230,000. The authority for neither statement is known. But Dr. Macgowan adds that 'the entire font  ... was melted after only about thirty impressions were struck off', which appears to contradict the testimony both of Wu Ch'ang-yüan and of the anonymous commentator. According to Mayers, the current tradition at Peking in his time was that 100 sets were printed, and this certainly seems more likely to be the correct figure. Two sizes of type were employed: the larger, of which the bulk of the work consists, measuring just over one square centimètre, and the smaller, used for notes, etc., about 5 mm. square. The dimensions of the 本 pên or volumes, each of which contains two chilan, are 27.5 x 17.5 cm. or thereabouts; and the whole collection is enclosed in 523 函 han or cloth cases, between 8 and 10 pên to each. Finally, to quote Mr. Mayers: 'A limited number of sets, printed on fine white paper, were distributed as gifts among the Imperial princes and the highest functionaries of government, in accordance with the munificent system followed in all similar cases; and a further number were struck off on a more ordinary quality of paper'. The introductory matter in the first volume comprises:—
(1) A preface, reproduced in semi-cursive script, from the pen of the Emperor 世宗憲 Shih Tsung Hsien, dated the 27th day of the 9th moon of the fourth year of 雍正 Yung Chêng (October, 1726). [8 folios.]
(2) The report (表文) presented by Chiang T'ing-hsi and his colleagues on the completion of their task. This is dated the 27th day of the 12th moon of the third year of Yung Chêng (January, 1726), and contains a description, couched in ornate and difficult language, of the general scheme of the work. [75 folios.]
(3) A less florid and more business-like account of the various divisions (凡例), such as is commonly prefixed to any Chinese work of importance. First, the 6 main categories, next the subdivisions of the 部, and finally the 32 sections are discussed one by one. [17 folios.]
(4) A 總目 or general synopsis, setting forth the 32 sections and the number of chüan in each.  [3 folios.]
Then follows the 目錄 or detailed table of contents, which not only enumerates all the 10,000 chüan with the subheads falling under each, but also indicates the various classes (Hui k'ao, etc.) contained in each subhead.
So far as I am aware, the only complete and homogeneous copy of the original edition to be found outside China is that possessed by the British Museum. This is on yellow paper, and has been bound up, in European style, into 745 thick volumes, now standing in glass-protected cases on the floor of the King's Library. The Königliche Bibliothek at Berlin possesses about four-fifths of the whole, in the same edition. The Bibliothèque Nationale in Paris has lately acquired a white-paper copy of the same, very nearly complete, a few gaps being filled up by corresponding volumes of the second edition. The existence of this second edition was unknown to Mayers, and it seems to have been first mentioned in print by M. Paul Pelliot in 1902.  Complete copies of it, presented by the Chinese Government, are in the possession of (1) the China Society in London; (2) Columbia University, New York; (3) the Institut für Kultur- und Universalgeschichte, attached to the University of Leipsic. It was printed, probably from wooden blocks,  in the province of Shantung, and published by the Tsung-li Yamên in an edition of 250 copies.  It is almost an exact facsimile of the original, but the dimensions of the leaf are larger by a few millimètres, just enough to distinguish it by. This may explain how it so long escaped the notice of foreign bibliographers, who doubtless confused it with the first edition.  The only additional matter is a list of errata (考證) in 24 pên, beginning with the 目錄 and ending with the 雜飲器部 (XXXII, 198).
The second reprint was undertaken by a Chinese syndicate, and carried out by the firm of Major Brothers in Shanghai, leaden movable type being employed, and the process of printing lasting three years, from 1885 to 1888. By greatly reducing the size of the type, though not so much as to be trying to ordinary eyesight, it was found possible to issue the whole in 1,620 pên, measuring about 19.5 x 13 cm. Copies of this handy little reprint are in the possession of several sinologues and public institutions both in Europe and America.
The present Index, as already stated, consists primarily in a translation and alphabetisation of the titles of the 6,109 subheads, together with the Chinese of the same, and reference to the 典 section and 卷 chüan or book. The numbering of the chüan starts afresh in each of the 32 sections, while the subheads are not numbered. Reference to the chüan, however, thanks to the list of contents standing at the head of each and the device of catchwords on the edge of the leaf, enables one to discover any required subhead almost instantaneously. I have aimed in every case at giving a clue to the real contents of the subhead rather than a literal rendering of its title. This was necessary, though greatly increasing the labour of preparation, because the Chinese headings as they stand are often obscure, if not unintelligible, to any except a practised scholar. 
In addition to these main entries, a very large number of cross-references have been inserted. These may be classified as follows:—
(1) Cross-references from the transliterated form (Wade's orthography) of the Chinese heading, the characters being also given. This has been done in the case of all names of foreign countries, books, animals, and plants, except the very commonest; also when the Chinese form is the more familiar, or there is no good English equivalent. For example, 'I Ching. 易經. See "Book of Changes"'; 'Hsün-chio. 勛爵. See Rank conferred for merit'.
(2) Other cross-references of form, or of synonyms. E.g. 'Broach. See Baroach;' 'Water-lily. See Lotus'.
(3) As a considerable number of entries consist of double-barrelled phrases, such as 'Kindness and malice', 'Uncles and nephews', cross-references are given from the second member.
(4) In a few cases, where a noteworthy part of the subhead does not actually appear in the title. E.g., under 'Gongs', one is referred to 'Musical Instruments miscellaneous, XXIX, 136'. Such cross-references may be multiplied with advantage by any one who has occasion to use the work frequently.
(5) Full cross-references are supplied to the subdivisions of the lengthy 部 Medicine. E.g. 'Cold, catching. 傷寒門 See Medical science, XVII, 355-378'.
- A glance at the Table of Contents will show that most of the above sections have been embodied in the more comprehensive scheme of the T'u Shu Chi Ch'êng.
- Of these, hardly more than one-fifth are now extant.
- Completed in 1407, in 22,877 books, with table of contents in 60 books. It was never printed, and the last remaining transcript was burned with the Hanlin College in 1900. Four odd volumes are now in the Chinese library at the British Museum.
- In the T'u Shu the corresponding sections are called 典, and the subheads 部.
- In the Industries anciennes et modernes de l'Empire Chinois, by Stanislas Julien and Paul Champion (Paris, 1869, p. 159), the following statement is made, without indication of the authority: 'Sous le règne de l'empereur Khang-hi, qui monta sur le trône en 1662, des missionnaires européens, qui jouissaient d'un grand crédit auprès de ce monarque, le décidèrent à faire graver deux cent cinquante mille types mobiles en cuivre, qui servirent à imprimer une collection d'ouvrages anciens'.
- For the Chinese text, see Appendix I.
- These were: 都察院 The Court of Censors, 通政司 The Office of Transmission, 大理寺 The Grand Court of Revision, and the Six Boards of 吏, Civil Office, 戶 Revenue, 禮 Ceremonies, 兵 War, 刑 Punishments, and 工 Works.
- In a collection of desultory notes entitled 歸田瑣記, published in 1845 (chüan 4, fol. 9 r.).
- 'Hsing-chai' was Ch'ên Mêng-lei's hao or fancy name.
- In reality, as we have seen, it was the 12th moon.
- 吾鄉相傳，國朝圖書集成一書，成於陳省齋之手，實未核也， 恭讀康熙六十一年十一月諭。。據此，則圖書集成之成帙，非省齋所能專其功，而省齋之負才跅弛，讀此亦可見其概矣。
- See infra, page xviii, note 2: 而設康熙中.
- The pagination is not continuous, but commences afresh with every new subdivision in the chüan. As Professor Hirth, however, assigns a total of 852,408 pages to the work, it would appear that they have been actually counted. (See Zeitschrift für Bücherfreunde, Bd. I, p. 278.)
- For purposes of comparison we may take the 11th edition of the Encyclopaedia Britannica, which contains on a rough estimate some 40,000,000 words. As 150 English words, or thereabouts, are required to translate 100 characters of the Chinese book-language, we may say that the T'u Shu contains between three and four times as much matter as the largest English encyclopaedia.
- For a synopsis in tabular form, see Table of Contents.
- In the China Review, vol. VI, p. 221. I must acknowledge great indebtedness to his article throughout.
- In the Journal Asiatique for 1826, tom. IX, p. 56. This is the first notice of the T'u Shu Chi Ch'êng by a foreigner.
- The number of common surnames is much smaller in China than in England, although the population is at least eight times as great. The consequence is that a far larger number of people in China bear the same surname. The six which appear to be the most widely distributed, with the number of chüan here allotted to each, are: 王 Wang (28); 張 Chang (20); 李 Li (18); 劉 (17); 陳 Ch'ên; (14); 朱 Chu(11).
- The biographies given under special headings, however, will be found much fuller as a rule than those under the surname. The life of 韓信 Han Hsin, for instance, takes up 34 pages under the heading 'Generals' (XI, 445), and only 3 pages in XIV, 159.
- It was first of all intended to include the other groups as well; but it soon appeared that the amount of labour entailed would be quite incommensurate with the advantage to be gained.
- 辰垣識略, ch. 3 fol. 24 v.
- We learn incidentally from the same source that the incomplete set of type which still remained in the time of Ch'ien Lung was melted down for purposes of coinage on the suggestion of the officials in charge, who feared the possible results of an inquiry.
- In the Journal of the North-China Branch of the Royal Asiatic Society, no. II (1859), p. 174.
- It is noticeable that, whereas Dr. Macgowan speaks of a 'font', and Klaproth (loc. cit.) of 'caractères fondus en cuivre', Julien uses the word 'graver'. See above, note 5.
- Reproduced in Appendix II.
- See Bulletin de l'École Française d'Extrème-Orient, tom. II, p. 133.
- This is the opinion of Dr. Shimmi, Japanese Professor in Leipsic.
- Information given to Professor Hirth, of Columbia University, by the Viceroy Tuan Fang. This fixes the date of the reprint as subsequent to 1862, when the Tsung-li Yamên was first established.
- The second edition appears to have been printed on white paper only. All the above copies, at any rate, are on white paper, and also two odd volumes, belonging to the same edition, in the British Museum and Cambridge University Library. The only white-paper copy of the first edition known to me is the incomplete one at Paris.
- E.g. 聖壽,風節,三恪,留守.