YALE UNIVERSITY LIBRARY 3 9002 04480 3857 Jkp. and only 1 to the Sui period. This book has been printed in a number of editions, differing considerably in the value of their illustra tions, much of the detail of the various ornaments being sup pressed in the later editions. The original edition is probably not obtainable anywhere at present; but copies of a facsimile reproduction of it, — entitled " Chi-ta-chung-siu Suan-ho Po-ku-t'u-lu,"(2) because it appeared during the Chi'-ta period (A.D. 1308-12) as a reprint (chung-siu) of the " Po-ku-t'u-lu," the collection described iii this catalogue being kept in the "Suan-ho" Museum Palace, — though exceedingly rare, are still in existence. I heard of a complete copy at Yang-chou in 1893, but found that the more I offered, the more obstinate the owner became in not wishing to dispose of it. A frag ment of the work, including the chapters on metallic mirrors, is now at the Royal Library of Berlin (Hirth Collection, No. 142), — a folio print, containing in its text portions sixteen columns of seventeen characters on each leaf. A note added to each illustration states whether it corresponds in size to the original, or whether it has been reduced, and the inscriptions are facsimiles of rubbings taken from the originals. In all the later editions the illustrations have undergone considerable 1 The parenthetical references relate to the corresponding numbers in the list of Chinese characters at the end of the paper (p. 253). I 200. FRIEDRICH HIRTH reduction in size, involving the loss of much detail. New editions were published by the firms Pau-ku-t'ang, Tsiang- yang, and Po-ju-ch'ai(8) during the Ming, — the last one being dated 1588, and having as collaborators for illustrations the two best-known artists for linear drawings of the dynasty, Ting Yiin-pong and Wu Ting (see Giles, Introduction to the History of Chinese Pictorial Art, pp. 163 et seq.), — and also during the present dynasty, varying in style of print, but not coming up to the first two editions mentioned. 2. Si-ts'ing-ku-kien,(4) compiled under orders from the Emperor K'ien-lung, dated 1749, by a commission of scholars and art ists, under the superintendence of two imperial princes, — an illustrated catalogue of the bronze works and coins in the collection of the Emperor K'ien-lung. Among the members of the commission in charge of the work, we find the names of the Grand Secretaries Liang Shi-chong (Giles, Chin. Biogr. Diet., No. 1249), Ts'iang P'u (Ibid., No. 337), and Wang Tu-tun (Ibid., No. 2255), and quite an array of talent out of the emperor's surroundings, including well-known artists, such as Tung Pang-ta and Ts'ien Wei'-ch'ong (see my Scraps from a Collector's Note-Book, pp. 36 and 38). Chapters 39 and 40, being the end of the book, which is followed by a supplement on coins, are entirely devoted to metallic mirrors. Of the 93 mirrors reproduced in these illustrations, I is ascribed to indefinite antiquity, 57 to the Han, and 35 to the Tang period, several among which have already been described in the " Po-ku-t'u-lu." Dr. Bushell (Chinese Art, Vol. 1, p. 75) mentions a con tinuation of this work as circulated in manuscript under the title " Si-ts'ing-su-kien."(B) " Si-ts'ing " is the name of one of the Imperial Palace buildings, where the collections described were kept. Another palace containing curiosities was the Ning- shou-kung, to which a descriptive catalogue — also existing in manuscript only, and entitled " Ning-shou-ku-kien "<6) — is de voted. I have not seen either of the two manuscripts men tioned by Dr. Bushell, and am not able to say whether they contain matter bearing on metallic mirrors. 3. T'u-shu-tsi-ch'ong/7' the great cyclopaedia, in over 5,000 Chi nese volumes, commenced under the Emperor K'ang-hi, and com pleted under Yung-chong in 1 726. Chapters 225-228 of its 32d section contain the most exhaustive material we may desire for CHINESE METALLIC MIRRORS 2IO the study of metallic mirrors. In Chapter 225 we find the entire material of the " Po-ku-t'u-lu " reproduced, both text and illustrations. This text is followed by extracts from the native literature, referring to metallic mirrors, arranged in chronological order. We thus have extracts from the " Tung- t'ien-ts'ing-lu,"(8) an archaeological work of the thirteenth cen tury (Wylie, Notes on Chinese Literature, p. 134) ; the " P6n-ts'au-kang-mu,"(9) completed in 1578 (Bretschneider, Botanicon Sinicum, Part 1, pp. 54 et seq.) ; and the " T'ien- kung-k'ai-wu,"(10> published in 1637, from which St. Julien has translated several extracts bearing on technical subjects. These extracts are followed by a number of texts, being poems, legends, rhymes, odes, and other literary effusions bearing on the subject, by authors dating from Wu-wang, the founder of the Chou dynasty (eleventh century B.C.), down to the Ming period. The last three subdivisions (ki-sh'ij tsa-lu, and wai- pien) consist of historical extracts from the whole range of Chinese literature, and from them we may derive many an interesting cultural fact referring to mirrors and mirror lore. Several of the less voluminous cyclopaedias contain accounts regarding the subject, without illustrations: as, for example, the " T'ai-p'ing-yu-lan " (completed in A.D. 983), Chap. 717; the Ming cyclopaedia " T'ien-chung-ki," Chap. 49, pp. 31-42; and of the present dynasty the " Yuan-kien-lei'-han," Chap. 380, pp. 10-34, an (i.e., "Record of Ancient Metallic Mir rors"), by Wang Tu of the Sui dynasty, and the " Pau- king-ki"<20> (i.e., "Record of Precious Mirrors"). II. Foreign. 1. Sir D. Brewster. Account of a Curious Chinese Mirror, which reflects from its polished Face the Figures embossed upon its Back (Philosophical Magazine, Vol. 1, London, 1832, pp. 438-441). 2. St. Julien and Paul Champion. Industries anciennes et modernes de l'empire chinois (Paris, 1869): pp. 63-65, "Miroirs;" and pp. 234-236, "Notice sur les miroirs ma- giques des Chinois et leur fabrication." Reprint of a paper presented by St. Julien to the Academie des sciences, and pub lished in the Comptes rendus, Vol. xxiv (1847), pp. 999 et seq. 3. Museo Espanol de Antiguedades, Vol. iv (Madrid, 1875), PP- 303-320. A paper in Spanish on Chinese Magic Mirrors by Don Florencio Janer. 4. Dr. A. J. C. Geerts in Transactions of the Royal Asiatic So ciety of Japan, 1875-76, p. 39, quoted in the following. 4 CHINESE METALLIC MIRRORS 212 5. Nature, Vol. xvi (London, 1877) : Notes on Japanese Mirrors, by R. W. Atkinson, p. 62; Sam. Highley, p. 132; R. D. Darbyshire, p. 142; S. P. Thompson, p. 163; and J. Par- nell, p. 227. 6. W. E. Ayrton and John Perry, The Magic Mirror of Japan (Proceedings of the Royal Society of London, Vol. xxviii, London, 1879, p. 727). 7. M. Govi, Les miroirs magiques chinois (Annales de Chimie et de Physique, V. Ser., Vol. xx, Paris, 1880, p. 99) ; followed by the same author's Nouvelles experiences sur les miroirs chinois (p. 106), Ayrton and Perry's Sur les miroirs ma giques du Japon (p. no), and A. Bertin and J. Dubosq, Production artificielle des miroirs magiques (p. 143) ; also Bertin, Etude sur les miroirs magiques (in the same Annales, Vol. xxn, Paris, 1881, p. 472). 8. Muraoka. Erklarung der magischen Eigenschaften des japa- nischen Zauberspiegels und seiner Herstellung (Mitteilungen der deutschen Gesellschaft Ostasiens, No. 31, Tokio, 1884). 9. M. Paleologue. L'art chinois (Paris, 1887), pp. 67 et seq. 10. F. Hirth. Chinesische Studien (Leipzig, 1890), pp. 272-274. n. F. Hirth. Uber fremde Einfliisse in der chinesischen Kunst (Leipzig, 1896). Deals chiefly with Hellenistic influences as shown in the so-called " grape-mirrors," of which it contains 15 illustrations reproduced from the " Si-ts'ing-ku-kien." Note the Appendix containing chronological data. 12. F. Jagor. Japanische Zauberspiegel (Verhandlungen der Ber liner Ges. fiir Anthropologic, 1898, pp. 527 et seq.). 13. S. W. Bushell. Chinese Art, Vol. 1 (London, 1904), pp. 94-97, with two photogravures from originals, — " Mirror with Grasco-Bactrian Designs " and " Mirror with Sanscrit Inscription." Early Mention of Mirrors. Mirrors cast of bronze are among the oldest products of Chinese industry. Wang Fu,(21) the art historian of the twelfth century A.D. , who compiled the "Po-ku-t'u-lu," ascribes their invention to the Emperor Huang-ti (2704-2595 B.C.). This, of course, is a statement of very doubtful value, and prob ably means no more than that no definite time can be quoted, within the more trustworthy period of ancient history, for the 5 213 FRIEDRICH HIRTH origin of this object of culture. Since the oldest relics of this kind handed down to posterity date from the Han dynasty, I do not, of course, look upon the specimen depicted in the " Si-ts'ing-ku-kien " (Chap. 39, p. 1, described as a relic of a much earlier period, for which reason the Emperor K'ien-lung provided it with an inscription) as an absolute proof against this assertion. All we know about mirrors previous to this time is based on literary evidence ; but of this we possess quite enough to make it sure that metallic mirrors of some kind were in use early during the Chou dynasty. I do not refer to the many statements to this effect which we find in the several works of the Han dynasty, but shall quote a few instances from the oldest historians. The "Tso-chuan "<22) mentions the gift made by the emperor of " a queen's large girdle with the mirror in it," x which fact is recorded under the year 673 B.C. The word here used for "mirror" is kien,(23) with the radical kin ("gold," "metal") at its side. Another early instance of a mirror mentioned during the Chou dynasty is found in the " Chan-kuo-ts'6,"(24) in connection with a personage by the name of Tsou Ki.(25) The anecdote there quoted says, " Tsou Ki was over eight feet tall, and, being of prepossessing appearance, he once had donned his court dress, and, while looking into a mirror {k'ui king(26}), re marked to his wife, ' Do you think I am as handsome a man as Sii-kung [a great beau of his time] ?' She said, ' Certainly, and much better looking, too ;' then he asked his concubine, with the same result; and finally he asked a stranger, who also said Sii-kung could not compare himself to him in point of beauty. He was, however, wise enough not to trust too much to his flatterers, since his wife, he thought, was not an unbiased judge, his concubine was afraid of him, and the stranger expected a favor from him. He then went to court for an audience with King Wei' of Ts'i,(27> to whom he mentioned this little incident with a practical application to his royal position. The king's seraglio, he thought, were not unbiased in their judgment of him, his ministers were afraid of him, and the people at large all expected favors from him. The king took the lesson in good grace," etc. 1 Legge, Ch'un-ts'iu, Vol. i, p. ior. 6 CHINESE METALLIC MIRRORS 214 What concerns us in this anecdote is the unmistakable men tion of a mirror — here called king,i28y with the radical kin ("metal") at its side — in the " Chan-kuo-ts'6 " ("History of the Contending States ") , dating generations before the Han dynasty,1 and referring to Tsou Ki, a well-known historical personage, the philosophic counsellor to King Wei of Ts'i (North Shan-tung) , wlio reigned from 378 to 343 B.C. Tsou Ki must have lived about this time, since he and his relations to King Wei are repeatedly referred to by Ssi-ma Ts'ien.2 Unless the whole anecdote has been interpolated by Liu Hiang, the Han editor of the " Chan-kuo-ts'6,"3 — which, in the face of so much of the statements of the work being accepted as history, is scarcely less reliable than Ssi-ma Ts'ien's, — we have no rea son to doubt that metallic mirrors {king) must have been in use for toilet purposes about the middle of the fourth century B.C. Here we have an instance of the mirror being referred to as actually used. The occurrence of the word (used meta phorically, however) is very much older, thus proving the gen eral use of the article indirectly, since it occurs in the " Shi- king;"4 and the ode in which it occurs is referred by Mau Ch'ang,<31) one of the oldest commentators of the "Book of Odes," to the period 866-854 B.C. I look upon this passage as an even stronger argument in favor of the early use of mir rors (kien) than the unmistakable mention of mirrors in the " Chou-li,"<32) which would indeed bring us back to the twelfth century B.C., if it were certain that this code of government and social life was draughted to its full extent, as it now stands, by Chou-kung, the brother of Wu-wang, the first king of the Chou dynasty (1122-1115 B.C.) ; for we have to take into account the possibility of the " Chou-li " representing a collec tion of government regulations which may have originated with Chou-kung as the first great legislator of the dynasty, but were added to in the course of generations until the text had attained its present wording. 1 I-sh'i,<2S> Chap. 108, p. 12; T'u-shu-tsi-ch'ong, Chap. 227, ki-sh'i, p. 1. 2 Chavannes, Les memoires historiques de Se-ma Ts'ien, Vol. v, pp. 240, 244- 249, 252-254, and 256. a Chavannes, Ibid., Vol. 1, Introduction, p. clii. 4 Part I, Book III, Ode 1, 2, edition of Legce, p. 38; Wo sin fe'i kienim ("My mind is not a mirror "). 7 215 FRIEDRICH HIRTH An important passage, probably still older than that in the " Book of Odes," is contained in the " Shu-king," ' in the so- called "Announcement about Drunkenness," as Legge trans lates the title of the book, in which the Duke of Chou gives his younger brother Fong a lesson on the importance of practising temperance in the use of spirits. In this address he quotes an old proverb (ku-jon yu yen yiiei33)), which may have been on record many generations before his time (the eleventh cen tury B.C.), to this effect: "One should not (have one's own image) reflected in the water, but one should have other folks' (images) reflected; now that Yin (i.e., the Shang dynasty) has seen its fate fulfilled, are they not to us a great mirror (ta-kien) from which (to learn) how to secure the repose of our own time?" The words I have translated by "reflected" and " mirror " (" reflection ") are represented by the mere phonetic kien^^ without the radical kin ("metal"), and I do not see why that old proverb should involve the use of metallic mirrors at all. On the contrary, I believe that the words jon wu yii shui kieni35) ("man should not get reflected in the water") are a remnant of those times, lying far beyond the Chou dynasty, in which no mirrors of any kind were known ; when those who wished to see their own image were obliged to look into a sheet of water, possibly in a flat basin covered with a coat of black varnish. The conclusions we may be allowed to draw from these se lected passages may amount to this : some time during the Shang dynasty (1766-1122 B.C.) there was a proverb involving that one had to look into water in order to see one's image reflected, which means that mirrors were unknown. Metallic mirrors can be shown to have existed in 673 B.C. as girdle-ornaments, and about the middle of the fourth century B.C. to have been unmistakably used for toilet purposes, probably much earlier. Terminology. We find two words applied to the metallic mirrors of an tiquity, kieni23) and king,{28) both usually written with the rad ical kin ("gold" or "metal"), indicating that the two words, 1 Part v, Book x, 12, edition of Legge, p. 409. CHINESE METALLIC MIRRORS 2l6 by force of their meaning, belong to the category of metals. In the classics, kien is sometimes written without this radical, when it would simply mean " to examine," " to reflect," or " the reflec tion; " so Is king, especially in the inscriptions appearing on the mirrors themselves. Altogether, radicals are quite commonly suppressed, not only in the ancient hieroglyphic styles of writ ing, but even in comparatively recent texts, such as the " Chu- fan-chi," by Chau Ju-KUA<36) (thirteenth century A.D.) ; e.g., taa«<37) ("satin"), which is there written without its radical for " silk." There is apparently no difference in the use of the two words, as far as ancient texts are concerned, unless it is this : that kien can be used as a verb in the sense of "to reflect," which is not the case with king. After the introduction of glass mirrors, the term king was applied to them, and from the meaning of " glass mirror" it gradually developed into that of " a transparent lens of glass or of crystal." Yen-kingi38y ("eye-mirror," or "eye-lens") is the modern term for "spec tacles," or a " magnifying-glass." The great encyclopaedia1 quotes from the " Fang-yii-shong-lio "<39) that the country Man- la-kia<40) (Malacca) produces ai-tai-kinglil'> (literally, "cloudy mirrors," or "lenses") : "Old men not able to distinguish fine writing put them on their eyes, when the writing will be clear."(42> This is evidently an early mention of spectacles of some kind, apparently imported from Europe by way of Malacca. A book called " Fang-yii-shong-lan " dates from A.D. 1239 ; but it cannot be the identical one, on account of the name " Malacca," which it appears did not become known in China until two hundred years later. Materials. Polished jade may have been among the earliest materials used as a substitute for the time-honored water-sheet, the mirror of nature. A jade mirror (yu-kiena4'>) is mentioned by Kuan- tzi" (died 645 B.C.) ; and in A.D. 485 a jade mirror (yii- king(i5> ) , with certain bamboo tablets covered by inscriptions in tadpole characters, was found in an ancient tomb near Siang- yang in Hu-pe'i Province, which the great polyhistor Kiang Yen 1 Chap. 228, tsa-lu, p. 4. 9 217 FRIEDRICH HIRTH (A.D. 443-504), in whose biography the account of the dis covery appears, declared to date from the time of Siian-wang (827-782 B.C.).1 The descriptive catalogue of jade-works, the " Ku-yu-t'u-p'u,"C46) published in A.D. 1176, contains accounts of every possible variety of this ancient branch of sculptural art, but none of jade mirrors; and it seems, that, from causes easily understood, jade mirrors never rose to become a regular industry after the introduction of the metallic article. The cyclopaedias also refer to stone mirrors, pure and simple; and an eccentric Buddhist scholar even tried one day to grind and polish a common brick into a mirror, with what success we do not hear.2 Alloys of copper and tin in equal parts yielded the bronze used at one time or another in the manufacture of mirrors during the Chou dynasty. The "Chou-li,"3 in its description of the various offices in that constitutional period, when the entire social and economic life of the Chinese was forced into a detailed system of government interference, speaks also of the officials in charge of bronze-works. Such works, of course, had existed centuries before Chou-kung, the alleged editor of the " Chou-li ;"(32) but I am inclined to look upon that great legislator merely as the creator of the system, while allowing for additions and modifications made by the several govern ments of the dynasty succeeding that of Chou-kung as Wu- wang's prime-minister. The " Chou-li," in its present shape, enumerates the different experts in charge of bronze-works, — beaters, melters, etc. Each of them had his well-defined prov ince; and for the several bronze alloys they turned out, the government had once for all provided a fixed standard. There were six kinds of alloy, defined by the proportion of tin in so many parts of the alloy (consisting of tin and copper), and no other metals are mentioned as entering the alloy. The propor tions were as follows. Nan-shi', Chap. 59, p. 4. 2 P'ei'-won-yun-fu, Chap. 83) p. 80. " Transl. Ed. Biot, Le Tcheou-li (Paris, 1851), Vol. II, pp. 491 et seq. 10 CHINESE METALLIC MIRRORS 2l8 Parts Parts Copper. Tin. I. Bells and sacrificial urns 5 1 2. Axes and hatchets 4 1 3. Halberds and lances 3 1 4. Sword-blades 2 I 5. Siau<47) (a knife used for engraving characters on wooden tablets) and arrow-heads 3 2 6. Metallic mirrors (kien-suif4^ i.e., "mirror-igniters") 1 1 It will be seen from this extract, that, according to the " Chou-li," the alloy of mirrors contained a larger proportion of tin than that of any of the other bronze works. This does not mean, of course, that the original standard was not aban doned in later times ; nor is the use of other metals in the alloy excluded in later works, as we learn from some of their inscrip tions : e.g., Han mirrors, described and depicted in the " Kin- shi-so,"1 said to have been cast from an alloy of copper from Tan-yang (the present Ning-kuo-fu), silver, and tin. Mirrors or bronze inlaid with gold, and of solid silver inlaid with iron, are mentioned with iron mirrors in the time of Ts'au-ts'au, who died A.D. 220. Altogether, there is no lack of passages show ing that other metals entered the alloy. The " T'ien-kung k'ai-wu,"<10) quoted in the great cyclopaedia, gives the following account of mirror alloys: " Mirrors are cast in moulds of concrete, the alloy, according to the ' K'au- kung-ki 'c49) (the technical part of the ' Chou-li ') , consisting of copper and tin in equal parts, the so-called kien-suia8) (or ' mirror-igniter ') alloy. The polish of mirror surfaces is done by laying on quicksilver, since the bronze alone will not permit of such fine polish. During the period K'ai-yiian (A.D. 713- 742) of the T'ang dynasty, certain percentages of silver were added to the bronze ; and in such mirrors as contain a few taels of silver, this will cause cinnabar-colored stripes and spots to show on the surface, as the visible indication of the presence of gold or silver in the alloy. The bronze censers of our Suan-to period (A.D. 1426-36) also owe theif origin to the fact that, during a conflagration in some imperial treasury, gold and silver 1 Kin, Vol. 6, fols. 32-34. II 219 FRIEDRICH HIRTH were mixed with bronze works and melted into lumps, which were ordered to be cast into censers. T'ang mirrors and Siian-to censers are therefore found in quantities among our court collections." The legend of the conflagration in the Siian-to period is referred to a Buddhist temple, with its images of gold, silver, and bronze, by Ku T'ai, the author of the " Po-wu-yau-lan "(50) (Chap, i, p. 12) ; but he entirely discredits it in ascribing the origin of the Siian-to bronzes to the emperor himself, who had called for a report from his experts on the best methods of producing a really superior alloy. Ku T'ai was a court official under the Ming dynasty; and his position, in connection with the connoisseurship exhibited throughout his book, gives him a certain authority. Iron mirrors ( t 'ie-kien CB1 ) ) , simply SO' called, are described in the " Po-ku-t'u-lu," one being dated in the Sui dynasty, twenty- one in the T'ang. It seems that they did not survive long after the twelfth century, since they are not represented in the K'ien-lung collection. They may have been consumed by rust in the course of centuries, just as Chinese records speak of iron drums (besides bronze drums), none of which seem to have been preserved. The oldest mention I have been able to find of this variety occurs in the " Si-king-tsa-ki "(52) (Chap. 6, p. 2), where the opening of a tomb under Wu-ti (140-87 B.C.) is referred to as belonging to King Ai of Wei (318-296 B.C.) with " several hundred iron mirrors." Though not directly connected with our subject, it may be of interest to note that glass mirrors, too, had been imported from abroad long before the metallic-mirror industry saw some of its best days during the T'ang dynasty. The " Liang-ssi- kung-ki,"(53) a work referring to the Liang period,1 relates that the ocean ships of Fu-nan (Champa) brought from western India,2 for sale, "mirrors made of green glass" (pi-po-li155'') . The price demanded for such a mirror was a million strings of cash, then about as many dollars, — an incredible price, if we consider the purchasing-power of money in those days. 1 A.D. 502-557, quoted in the T'u-shu-tsi-ch'ong 32, Chap. 227, ki-shi, p. 4. 2 Si-t'ien-chu,'-m by which name sometimes the west of Asia, including Syria, is covered. 12 CHINESE METALLIC MIRRORS 220 The period is quite late enough to explain them as the product of Syria (Ta-ts'in), then probably the only country in the world where they could be produced. Pliny,1 having described the manufacture of glass at Sidon, says of that city that it " had also invented (glass) mirrors" {etiam specula excogitaver at) . This shows that they must have been known in Syria in the first century A.D., though Marquardt2 adds that positive evi dence is not on record before a comparatively late period, pos sibly the sixth century A.D., which would have been quite early enough for the Chinese notice referred to. China had then had old connections with the silk-manufacturing cities on the Phoenician coast, her silk being partly paid for by purple-dyed textures, precious stones, and glass. During the Liang dynasty particularly, the sea-trade with Syria by way of India, Fu-nan (Champa), Annam, and Tongking, was in full swing;3 and the shrewd merchants of Syria probably were not slow in making the most of this new style of mirror with purchasers placing so much value on mirrors generally. Chinese Metallic Mirrors outside of China. The Japanese have from the remotest periods been eager col lectors of Chinese antiquities, and in no country of the world are these works of art more highly appreciated than in Japan. It is almost a matter of course that old Chinese mirrors should have been among them. The island kingdom has entertained lively intercourse with China, both through missions sent to the court and through private travelling-enterprise, ever since the first " tribute-bearers " came from there to China, in A.D. 57. Regular junk traffic, we know, existed about A.D. 1200 between the ports of Japan and Ts'iian-chou-fu in Fu-kien.4 It was the Japanese who went to China in search of articles of vertu, not vice versa: they came as friends and foes, in peace and war; but they hardly ever came without taking back with them as much in the way of literary and art treasures as they 1 Hist, nat., xxxvi, 193. 2 Privatleben der Romer, 2d ed., p. 758. 8 See the Ljang-shu, quoted in my China and the Roman Orient, p. 47. 4 See my Ancient Porcelain (Leipzig, 1888), p. 67. 13 221 FRIEDRICH HIRTH could get. This was particularly the case during the several raids made by Japanese adventurers under the Ming dynasty between the years 141 1 and 1439. Trade with China has, for long periods throughout antiquity and the middle ages, been under the strictest prohibitions, and much of the exchange of produce with foreign countries took place under the pretext of tribute {kung) to the Chinese Court. The tribute-bearers almost invariably received certain gifts in return; and the Chi nese historians, reporting the transaction, often distinctly state that, " such and such counter-gifts being prayed for by the tribute-ambassadors, the issue of them in such and such quan tities was ordered by the Emperor." Independent distant nations, like the Arabs, would not have continued to submit to the many humiliations which Chinese national prejudice im posed on such court visits, had not the " trade " — for such it was, in reality, under the name of "tribute" — been a paying one. With all the drawbacks connected with this court monop oly, the system by which merely the sovereign was supposed to carry on barter-trade through an exchange of substantial cour tesies has had its cultural advantages : for, being nominally destined for a crowned head, none but the best works of indus try and art were sent; and this may account for the best things of a country like China being found abroad, as we find the finest celadon porcelains in the old Khalif countries, and the best old Chinese paintings in Japan. It is from some of these tribute-missions that we may draw conclusions as to an early demand for Chinese metallic mirrors in Japan. We read in "The History of the Three Kingdoms"1 that in A.D. 238 ambassadors arrived from the queen of one of the several little kingdoms then occupying the present Japan, with rich presents for the Court of the Wei dynasty; and in January of the fol lowing year, the Emperor Ming-ti ordered a number of counter- gifts to be handed over to rhe ambassadors, comprising, among precious brocaded stuffs, supplies of painters' silk, gold, pearls, etc., one hundred bronze mirrors. In A.D. 240 a further mis sion came from Japan, when mirrors were given away again; and since these tribute-missions were in the sequel repeated San-kuo-chi, WeT, Chap. 30, p. 27. 14 CHINESE METALLIC MIRRORS 222 every now and then without the description of tribute-articles and counter-gifts being entered in the histories, we may conclude that great numbers of the bronze mirrors, especially many of the artistic works of the Han dynasty, have gone the same way. It is quite probable that Japan was not the only country to participate in this class of gifts; indeed, we may not be sur prised to find the article being occasionally picked up in any neighborhood the rulers or inhabitants of which have at some time or other had connection with the courts of China. In his excavations in the Caucasus region, the late Professor Virchow discovered, among other ancient relics, a metallic mirror,1 the shape and ornamental arrangement of which (concentric sec tions and central knob with perforation for cord) caused me to suggest that it was not of Scythian, but of Chinese origin.2 I believe so even now; but I am bound to admit the force of the arguments advanced by Dr. P. Reinecke,3 of the Roman and Germanic Museum in Mainz, who draws attention to very important analogies between the construction of the Chi nese metallic mirror, as known to us in existing specimens, and a number of mirrors supposed to be of "Scythian" origin (to use a name which may cover all possible doubtful nationalities filling the gap between ancient Europe and the China of the Chou dynasty). These "Scythian" mirror types — the oldest specimen of which, for reasons given by Dr. Reinecke (p. 144) , should be ascribed to the sixth century B.C. — have no back- ornament, are disk-shaped, and have the perforated central knob, just like the Chinese varieties. It would of course be premature for any one to commit himself to an opinion regard ing these Scythian types — some of which were discovered so far west as the south of European Russia — who has not seen them himself, and, more than this, who has not sufficient expe rience, based on the technical examination of specimens, chem ical analysis, etc.,. both with Scythian and Chinese discoveries. 1 See illustration in Virchow's Report (Zeitschrift fur Ethnologie, 1890, Verhandl., p. 449). 2 Zeitschrift fur Ethnologie, 1891, pp. 808, 809. B Uber einige Beziehungen der Alterthumer China's zu denen des skythisch- sibirischen Volkerkreises (Ibid., 1897, pp. 141-147). 15 223 FRIEDRICH HIRTH The problems before us are: (i) Have the inhabitants of Scythia (this term to be understood in its widest application, and therefore covering the Huns of the Chou dynasty as the next-door neighbors of the Chinese) imitated contemporaneous Chinese patterns, of which we possess no specimens, but which, according to Chinese ancient records, must have existed in China long before the sixth century B.C., the oldest date sug gested for the Scythian types? (2) Could the pre-Christian specimens found in western Scythia possibly have been imported from China, and thus represent those types of the Chou dynasty the existence of which would not be known to us but for certain allusions in the contemporaneous Chinese literature? (3) If we find that these two questions must be answered in the nega tive, who invented the perforated central knob, — the Chinese, or the Scythians ? Shape and Size. Nearly all the Chinese metallic mirrors we know of are disk- shaped; and they are held by a silk cord run through a knob in the centre of the unpolished surface, as shown in an illustra tion of Bushell.1 But few specimens are square. The handle — cast of one piece with the mirror itself, characterizing the Japanese metallic mirror as well as the one of classical an tiquity2 — never appears on the Chinese mirrors depicted in the standard works on the subject; but "hand-mirrors having han dles" are said — in a work which I believe to be made up of archaeological notes left by Li Lung-mien, the painter known in Japan as Ririumin (tenth century), and which I find quoted as "Li-shi-lu" in the "T'u-shu-tsi-ch'ong,"3 — to have been used under the Emperor Wu-ti (died 87 B.C.) by the dancers of pantomimes.(56) Since we know that this emperor was a great curio-hunter, who sent frequent caravans to western Asia to col lect objects of art, there is a possibility of the hand-mirror with handle (called wu-king, i.e., "pantomime mirror," in the pas sage referred to) having been constructed in imitation of Greek 1 Chinese Art, Vol. I, p. 96. 2 Marquardt, Privatleben der Romer, 2d ed., p. 690. * L.c, ki-shi", p. x. 16 CHINESE METALLIC MIRRORS 224 or Roman patterns. The handle-bearing mirror, used for toilet-purposes both in China and Japan, may have originated from it. It was and is certainly so used in China, as well as the standard-shaped knob-mirror. In a modern painting rep resenting a Chinese lady at her toilet, the lady is seated before a round metallic mirror set in a mirror-stand, while she holds behind her head a mirror with a handle, in order to be able to see the back of her coiffure. Very large mirrors, a number of feet long and broad, prob ably rectangular in shape, and reflecting the whole human figure, are also mentioned in the post-Christian literature, some of them being imported from abroad. Several passages are on record which tell of the importation of such large mirrors under the Emperor Wu-ti, who hoarded up many precious things he had sent for from western Asia ; but they are mostly blended with legendary matter, which makes it difficult to treat them seriously. Such large metallic mirrors, aside from the possibility of their having been originally constructed in China itself, could easily be accounted for as importations, or imita tions, of Western specimens, certainly during the first century A.D., when Seneca the Younger mentioned mirrors of sizes equal to those of human bodies.1 Otherwise mirrors were generally so made as to reflect neither more nor less than a human face; and the diameter of the disk chiefly depended on the principle on which the reflect ing surface was constructed, whether plane, concave, or convex. Sh6n Kua (A.D. 1030-93) says, in Chap. 19, p. 3, of his interesting cyclopaedia " M6ng-k'i-pi-t'an,"(57) " In casting mir rors, the ancients would give a large mirror a plane, a small one a convex, surface; for all mirrors will reflect a man's face large if they are concave, and small if they are convex. In a small mirror one cannot see the whole of a man's face, for which reason the surface is made slightly convex; and by reflect ing the human face in reduced size, a mirror may be small and yet take in a man's face complete, though the reflected image will correspond in size to the size of the mirror. This was one of the clever achievements in which the artists of later periods 1 Marquardt, Privatleben der R6mer, 2d ed., p. 690. 17 225 FRIEDRICH HIRTH were not able to turn out work comparable to the ancient mir rors: they could only scrape and polish plane surfaces, thus recalling the music-master K'uang, who would make a mess of his notes (without his pitch-tubes)."1 This passage seems to imply that in the eleventh century A.D., when Shon Kua wrote, the art of making concave and convex mirrors had been lost. Unfortunately he does not state when this was the case; but for that, the manner in which the reflecting surface is worked would in many cases serve as a criterion of age. Neither am I able to say whether the secret of working concave and convex mirrors was re-discovered after the eleventh century; but the fact that the mirrors of the Ming and present dynasty are usually of much greater sizes leads me to think that modern workmen were afraid to approach the more laborious methods of the ancients. We cannot, of course, look upon Shon Kua's statement as a hard and fast rule; for we read in the " Chon-chu-ch'uan,"(58) a work of the Ming dynasty,2 that Sii Hiian,(59) the well-known editor of the ancient " Shuo-won Dictionary " (A.D. 916-991 ) , was in possession of a mirror which would reflect just enough of a face to show merely one of its eyes, and this may have been one of the ancient "igniters;" and according to the " Hua- shu,"c60) a Tauist work of the tenth century,3 a complete outfit contained four mirrors, — one for "seeing big," one for "seeing small," one for "seeing plain," and one that would invert the object to be seen. Uses. For Toilet Purposes. — It need hardly be said that the metallic mirror, being the only apparatus available for the purpose, next to a sheet of water, was an indispensable perqui site of a toilet-room. Many passages to this effect appear even in the pre-Christian literature, the several classical texts using the term quite commonly in figurative language, too (see also " Early Mention of Mirrors," p. 214). From Shon Kua's remarks, it would appear that mirrors of every kind of con- 1 See Mencius, Legge, p. 164. 2 Quoted in Ko-chi'-king-yuan, Chap. 56, p. 6 b. 3 Wylie, p. 127, quoted in Ko-ch'i-king-yiian, Ibid., p. 7. 18 CHINESE METALLIC MIRRORS 226 struction had been in use ; but, from the way in which the phi losopher Han-fei-tzi 1 introduces the mirror into his specula tive discourse, I am led to believe that magnifying-mirrors were much in favor, even for toilet purposes. The philosopher says, "The ancients knew that the eye is too short to see one's self, for which reason they used a mirror to see their face ; and [they knew that] their intelligence was too limited to know themselves, for which reason they used the tau'-61) (Lau-tzi's method of rea soning) to correct themselves. A mirror in which one does not see the flaws (in one's face is like) a tau (or a philosophy) by which one is not enlightened on the wickedness of sin. If the eye has no mirror, it has no means of straightening up the hair on the temples and eyebrows ; if man's self has no tau, he has no means by which to know his errors.". In a passage quoted in the same cyclopaedia from HuAI-NAN-TZl,(62) it is insinuated that " a newly-made mirror is at first dim, but, on having been polished, one can examine by its use ' the finest hairs ' {wei-mau) of the temples and eyebrows." On the other hand, mention is also made, by way of curiosity, of a mirror bearing an inscription of the Han dynasty having fallen into the hands of the poet Su Tung-po (about A.D. 1078), which reflected human figures quite small.2 The ac count adds that this is a peculiarity of ancient mirrors. Ritual. — The "Ku-kin-chu,"(63) a work originally of the fourth century A.D., but as it now stands of doubtful identity with that old text, yet giving us (Chap. 2, p. 5) a clear defini tion of the term yang-sui^6*'' (i.e., " the sun-igniter") , says, " It is made of bronze, and has the shape of a mirror; when held against the sun, it will produce fire, which is obtained through being caught by a heap of dried Artemisia leaves." Artemisia moxa is largely used in medicine as a counter-irritant in cauter izing the skin. " The heat of the sun's rays collected by a mir ror is said to be the proper way of igniting the moxa." 3 The Chinese name is «i,<65) for which ping-t'aim> ("ice terrace") occurs, owing to the leaves being, in default of a mirror, ignited 1 Third century B.C., quoted in T'ien-chung-ki, Chap. 49, p. 33. 1 Ibid., p. 41. 3 Porter Smith, Contributions towards the Materia Medica, etc., of China, p. 25. 19 227 FRIEDRICH HIRTH by means of a piece of ice, cut into a roundish form, which col lected the sunbeams.1 According to the institutes of the Chou dynasty, by which the doings of the people, from the emperor down to the lowest subject, were forced into certain regulations, fire used for any purpose, including kitchen fires, lamp and candle lights, was produced from heaven under certain ceremonies. There were two kinds of fire, — the people's fire {min-huo) , and the state fire {kung-huom)) , — each being under the superintendence of a special government officer, who had to see that at certain seasons the so-called " new fire " was obtained under the proper ceremonies. Whoever used fires in his household had to keep a flame alive day and night, all through the season ; and to allow it to go out was punishable. The " new fire," obtained by rubbing with a gimlet certain kinds of wood, varying in the four seasons, was kept in a temple, where the people, on having extinguished their " old fires," got delivery of it for household use till the season changed again. The change was later on confined to an annual term in the spring, when the ceremony took place on the fifteenth day of the second moon (some time in the beginning of April) ; and since their forefathers of the Chou dynasty had on that day to content themselves with cold meats, the Chinese, up to the present day, speak of this festival as the "Cold Meats Day" {han-shi^). The "state fire," used for sacrificial purposes, was obtained by collecting the rays of the sun in a concave mirror, the casting of which had also to be done under certain ceremonies exactly at midnight on the day of the solstice. This is clearly one of the earliest uses to which the mirror was put, and the art of manufacturing concave ones was doubt less well known some time during the Chou period. Another kind of mirror, which need not necessarily be one of the same construction, was used to " receive the brilliant water from the moon," apparently by exposing it during clear moonlight nights, when it would be covered by the night dew, the water thus obtained being used in the sacrifices.2 It is probably in allu- 1 Bretschneider, Botanicon Sinicum, Part ni, p. 147. 2 See Biot, Le Tcheou-li, Vol. 11, pp. 194 and 381. CHINESE METALLIC MIRRORS 228 sion to its ancient use in drawing water from the moon, that certain associations between the mirror and the moon have arisen. Thus we read that in the K'ai-yiian period (A.D. 713- 742) a monthly feast was given to the ladies of the city of Ch'ang-an, — possibly as representing the female principle to which the moon is subject, — when metallic mirrors would be rapped on while being held against the moon. Another anec dote speaks of a mirror having the figure of a hare on the back. If held against the moon at full-moon time, the hare would dis appear as if by magic.1 To increase Lamp, Candle, or Sun Light. — In some of their early superstitions, the Chinese have anticipated the lead ing ideas of great inventions, the final discovery and develop ment of which fell to the lot of Western thinkers. It looks as if the inventors of such legends were guided by an instinctive intuition as to what might be within the range of possibility; and they do not stand alone in this respect, unless ancient le gends, like the one about the mirror placed on top of the Pharos tower of Alexandria and " contrived in such a manner that the inhabitants might observe the motions of their enemies at the distance of fifty days' sailing," are more than mere myths.2 The Chinese have in their ancient and mediaeval books similar accounts, clearly the result of a lively imagination, but yet in volving matter-of-fact principles well known to later genera tions. What else is it but a practical demonstration of the theory of sound-reflection, coupled with that of light-reflection, when we read, among a mass of fabulous accounts in the "Shi-i-ki,"c69) — a work of the fourth century A.D.,3 — that at the time of Ling-wang (sixth century B.C.) a concave metallic mirror, three feet in diameter, was credited with lighting up a room at night, and that, " if you spoke into the mirror, the sound would be reflected by way of answer"? And is it not an involuntary anticipation of many a catoptric apparatus now 1 P'ei'-won-yiin-fu, Chap. 89, p. 7. 1 See Benjamin of Tudela, translated by B. Gerrans, p. 157; and for a Chinese repetition of the legend, my paper Die Lander des Islam nach chines- ischen Quellen, pp. 52 et seq. 3 Chap. 3, p. 7. 21 229 FRIEDRICH hirth described in the Chinese " List of Lighthouses," when we hear of ancient mirrors able to throw light to a distance of 200 li?1 The account of an Indian mirror in the "Liang-ssi-kung-ki,"(53) which refers to the period A.D. 502-557, is much more modest when it ascribes to a Western king of Varanashi (Benares) the possession of two big mirrors, one of which would shine 30, the other 10 li. At a conflagration of the palace, the bigger mirror came forth shining as bright as ever ; whereas the smaller one was slightly injured, and had lost its brightness, though retaining its wonderful power of rendering poison harmless. To ward off Evil Influences. — The principal supersti tion connected with the use of mirrors is probably the belief that they will ward off evil influences; and many of the popular beliefs entertained at the present day with regard to mirrors of any kind, now chiefly the glass mirror, whether imported or of native manufacture, can be traced back many centuries before us, when none but metallic mirrors were to be had. The mir ror has at all times been regarded as a charm against the attacks of those unseen evil spirits that beset the life of the Chinaman unless he do something to ward them off. We do not know how and when these superstitions arose; some of them may have been born with the Chinese nation itself. History abounds with instances proving their early existence ; but it appears that nothing has contributed so much towards their development as the abuses made by the later Tauist schools of that mystic doc trine of the great philosopher Lau-tzi. An early adherent of this school, the alchemist Ko Hung (fourth century A.D.), speaks, in his " Pau-p'o-tzi," 2(70) of Tauist scholars being in the habit of hanging on their backs a bright metallic mirror, nine inches in diameter, so that sprites or evil spirits would not dare to come near them ; for the mirror had the power to reflect the bodies of all invisible birds and beasts of evil influence which infested the air, and thus caused them to assume form. Another old passage is on record in the "Tung-ming-ki,"C71) a chronological account of wonderful events concerning the Court of the Emperor Wu-ti, supposed 1 Ko-ch'i-king-yuan, Chap. 56, p. 8. 2 Ibid., Chap. 56, p. 8 b. CHINESE METALLIC MIRRORS 230 to date from the Han dynasty, according to which1 a certain foreign country presented the Emperor in the Yiian-fong period ( 1 10-104 B.C.) with a mirror four feet wide, which would " reflect and make visible the otherwise imperceptible hidden forms of sprites," — di'i-meif™ the very word used in the " Tso-chuan " 2 for the " sprites and evil things " which the Em peror Shun wanted certain tribes of evil-doers to face in the uninhabited distant regions to which he had banished them. It quite goes with the spirit of the time that such a mirror comes from abroad so soon after the expeditions of the great dis coverer Chang K'ien, when the several countries of the Tarim Basin were being organized into Chinese colonies; for the much-dreaded desert of Eastern Turkistan, with all its dismal loneliness, was alive with unseen creatures known as ch'i-goh- lins"2' (imaginary beings like mu-mei, or " tree-goblins, "(73) in one of Li Lung-mien's mythological pictures) . The " Tung-t'ien-ts'ing-lu " weight, 351 grams. Orna mented with a dragon, head downwards, playing in the clouds, the latter represented by ling-chi ornaments 1 raised in relief above a flat bottom, on which the water-pattern (shui-zvdni91'>) has been faintly marked. The dragon (lungm'>) is represented in its modern shape in high relief, and possesses all the attributes we are accustomed to see nowadays on the representations of the dragon: such as the five claws of its feet, clearly distinguishable; the body of an alligator, or some such reptile, covered with scales; protruding eyes; horns; antennae; and beard. An inscription in seal characters appears on a framed tablet with an orna mental roof, and standing on a base resembling a padma-like flower. The entire back is surrounded by a raised plain flat margin with sharp edges. It appears that a companion to this mirror was contained in the col lection of the Emperor K'ien-lung, inasmuch as the " Si-ts'ing-ku- kien " 2 w contains an illustration which leaves little doubt as to its being of the same class, in all its details, as the one now belonging to the Musee Guimet. It is described under the name Han Yiin-lung;-kienm) (i.e., "The Cloud-and-Dragon Mirror of the Han Dynasty").3 The description of " The Cloud-and-Dragon Mirror " in the " Si-ts'ing-ku- kien," says, "Diameter, 3.4 Chinese inches; weight, 10 ounces. On the back is represented the shape of a dragon frisking in the Pond of Heaven (trien-chiiei)), the drawing being interspersed with patches of clouds (to-yuni9^) . Plain margin; plain knob (pif9e) literally, ' nose '). Of the inscription, five characters are readable."*97' The five The ling-chi is a kind of fungus looked upon as an emblem of longevity, possibly Agaricus noctilucens, Lev. (see my Scraps from a Collector's Note-Book, New York, 1905, p. 78). 2 Chap. 40, p. 35. 3 Every work of art described in any of the standard works containing illus trations is given a name by the art historians. This name is usually headed by the character denoting the dynasty during which, in the opinion of the native authorities, the work of art referred to has originated. We learn thus from the title given to the companion mirror in the K'ien-lung collection that the editors of the Si-ts'ing-ku-kien consider it a work of the Han dynasty, whose rule extended from 206 B.C. to A.D. 220. The name itself usually contains an allusion to the chief ornament or some characteristic inscription found on its surface, and always describes the class of work to which, according to Chinese terminology, it belongs, — whether tripod (ting), water-basin (si), mirror (kien), etc. 34 Boas Anniversary Volume. Plate XV. 2 3 CHINESE METALLIC MIRRORS, MUSe'e GUIMET, PARIS. CHINESE METALLIC MIRRORS 242 characters which the authors of the " Si-ts'ing-ku-kien " say were dis cernible in the specimen belonging to the K'ien-lung Museum are stated by them to have been . . . ki . . . chong-yiie j'i taz«(98) (Le., "... happiness . . . manufactured in the days of the first moon "). In the specimen now belonging to the Musee Guimet the inscription seems to have been somewhat better preserved. I can make out the last four characters to be identical with those on the specimen of the K'ien- lung Museum, but the reading ki ("happiness") does not seem to be justified by what appears on the cast now before me. I am rather under the impression that the characters preceding chong are intended to mean ir-shi-ir-nienW (i.e., "in the twenty-second year"). I must admit that the shape here given to nien ("year") is very unusual, and that its palseographic explanation is not based on parallels easily procurable. What I make out on the cast is a hieroglyphic which I cannot explain with confidence. The nearest equivalent in form which I have been able to discover will be found in the well-known " Corpus Inscriptionum," published in 1804 at Yang-chou by the former Viceroy of Canton, Yuan Yuan, the " Chung-ting-i-k'i-k'uan-shi," 1(100) where, in an inscription of the year 24 B.C., the character nien in the date appears in a somewhat similar shape. I regret not being able to decipher the first two charac ters, in spite of every possible attempt to discover parallels among the inscriptions of the period. If others should be more lucky, the discovery of a date, if a date is at all contained in the inscription, will be of great importance. It must strike every one familiar with the development of the ornamental representation of the dragon that in this case all the modern attributes (clearly-defined claws, rolling eyes, antenna?, etc.) are given to this animal, which, in stone sculptures originating as late as the second century A.D., are marked in a much more primitive man ner. I refer to Plates xxxi and xxxm in Chavannes' work,2 which may be said to represent the style of dragon followed towards the close of the Han dynasty. However, the native critics of the " Si-ts'ing-ku- kien," whose judgment certainly deserves consideration, ascribe the companion mirror referred to, not to the T'ang, where it would not call for doubt of any kind, but to the Han dynasty. The padma-ornament at the base of the frame containing the inscription is clearly of Buddhist origin, and seems to exclude the former Han dynasty as the time of origin. Moreover, the manner in which the dragon is drawn, especially the twisting of its tail around its right hind-leg, appears quite similar 1 Chap. 9, p. 25. 3 La sculpture sur pierre en Chine. 35 243 FRIEDRICH HIRTH on three dragon mirrors of the T'ang dynasty,1 though these T'ang dragons show only three or four claws, whereas the alleged Han dragon shows five. Magic Mirrors. Plate xv, Fig. 2: diameter, 9.75 cm.; weight, 125 grams. Fig. 3: diameter, 9 cm. ; weight, 99 grams ; rather thicker than the average ; the cast shows a slight break on the back. Fig. 4: diameter, 8.75 cm.; weight, 119 grams; the cast appears broken on the back. Fig. 5: diam eter, 7.75 cm.; weight, 89 grams. The class of mirror to which the four specimens shown in Plate xv, Figs. 2-5, belong, is described in the " Si-ts'ing-ku-kien." 2 The orna mental part is plain rather than artistic, the cause for which I believe lies in the use to which these mirrors were put. They were all, I believe, meant to serve as " magic mirrors," the well-known t' ou-kuang- kienf10^ first described by Shon Kua in his " Mong-k'i-pi-t'an." 3 (BT) Since the first half of the nineteenth century, magic mirrors have com menced to attract the attention of European savants, and the literature now on record * throws considerable light on the subject. I refer in the first instance to the justly celebrated memoir on " Miroirs magiques," laid before the Academie des sciences by the late Professor Stanislas Julien.5 For a review of the various experiments since made and placed on record, I would refer to the paper on the subject read before the Berlin Anthropological Society in November, 1898, by the late Dr. F. Jagor.6 There can be scarcely any doubt that the Chinese were the inventors of these magic mirrors. They have probably been known as curiosities ever since the Han dynasty, when Shon Kua, in his cyclopaedia, drew renewed attention to them, without being able to explain their manu facture. The poet Ma Chi'-ki of the Kin dynasty/102' whose name seems to have been misunderstood by Julien,7 and who wrote the poem " T'ou-kuang-king-sh'i "<103> (" Ode on a Magic Mirror "), is distinctly stated not to have been able to explain the principle underlying the phe nomenon. (104) Ma Chi'-ki, whose biography is preserved in the " History of the Kin Dynasty,"8 died in A.D. 1232 at the age of fifty-nine. Wu 1 Po-ku-t'u-lu, Chap. 30, pp. 4 and 5. * Chap. 39, pp. ri-24. ' Chap. 19, p. 4. 4 See pp. 211, 212, Nos. 1-8 and 12. * Published in Comptes rendus, Vol. xxiv (1847), pp. 999 et seq. 6 See p. 212, No. 12. 7 Le poete Kin-ma les a celebres en vers, p. 1000. 8 Kin-shi, Chap. 126, pp. 10 et seq. 36 CHINESE METALLIC MIRRORS 244 Tz'i-hang, (105) who is stated to have furnished the first explanation of the phenomenon, is described as having lived during the Yiian dynasty (1260-1368). The discovery of Wu Tzi'-hang's theory must for this reason belong to the thirteenth or fourteenth century. I have not been able to find out who Wu Tz'i-hang was, nor does it appear that any independent works now exist under his name in the Imperial Library of Peking. The work from which Julien's translation is derived is the " Yu-t'ung-su-lu,"(106) — a collection of miscellanies published in sixty- five chapters by Ho M6ng-ch'un,1<107) who graduated in 1493 and died in 1527, 2 leaving behind him a few works, including a commentary on the " Family Sayings of Confucius." 3 <108) The passage quoted from the " Yii-t'ung-sii-lu " in the " Ko-ch'i-king- yiian " * leaves no doubt whatever as to the meaning of the term t'ou- kuangSws) In modern Chinese, t'ou-kuang may be looked upon as the technical term for the " transparency of glass or crystal." In the Chi nese version of the " List of Lights, Buoys and Beacons, published by order of the Inspector General of Customs," the term " dioptric lens " is rendered by t'ou-king; C110) e.g., in the description of Gutzlaff Light No. 47 of the list, published in January, 1905. Williams5 renders t' ou-kuang by " a kind of skylight." In the older literature, however, the meaning of the term is the one described by Shon Kua, in whose account it is looked upon as a aro.£ Xeyo^evov by an author of the Southern Sung dynasty (thirteenth century), Chou Mi,(111) in his work " K'ui- sin-tsa-shi." 6 (112) Chou Mi says that " the theory of the magic mirrors has something that cannot be understood," and that " among the records of former generations it is only the ' Pi-t'an ' of Shon Ts'un-chung (i.e., Shon Kua) which touches the subject." 7<113) We may infer from this remark, that, although magic mirrors were known in ancient times, the literature on record at the time when Chou Mi wrote (thirteenth cen tury) contains no descriptive notice of them except the account of Shon Kua as translated by Julien. This explains somewhat the absence in the " Po-ku-t'u-lu " of every allusion to certain classes of mirrors possessing the power of reflecting the shadows of ornaments; and we may take it for granted, that, when this work was published (A.D. 1107), the magic 1 See the Great Catalogue of the Imperial Library, Chap. 127, p. 16. 2 Ming-sh'i, Chap. 191, pp. 13 and 19. 8 The Great Catalogue, Chap. 95, p. 1. * Chap. 56, p. 6. ° Syllab. Diet., p. 877. 6 Compare the Great Catalogue, Chap. 141, p. 34, against Wylie, Notes on Chinese Literature, p. 159, who places the work in the fourteenth century. ' Sii-tsi, Chap. 2, p. 28 ; cf. also P'ei'-won-yiin-fu, Chap. 22 A, p. 90. 37 245 FRIEDRICH HIRTH mirrors preserved among the treasures of the Emperor Hui-tsung's Museum were not known to possess any but the qualities of ordinary mirrors, because they were never suspected in this respect. A still more palpable reason for Wang Fu's silence on the subject may be that the collection he describes contained no specimens of this class. Since his time, it appears, new discoveries have been made; and among these the mirrors referred to as being described in the " Si-ts'ing-ku-kien," which closely resemble certain specimens in the Guimet collection as regards both ornamentation and inscriptions, are uniformly described as t'ou- kuangfW9) i.e., magic mirrors. The analogies which lead me to believe that the four mirrors of the Musee Guimet shown in Figs. 2-5 of Plate xv were originally intended to be magic mirrors, are based on ornamentation as well as inscription. The characteristic feature of their ornament is an octangular star occu pying the centre of the disk. In the specimen seen in Fig. 3 the knob is surrounded by twelve globules, as in the illustrations on pp. 13, 19, 21, and 23, Chap. 39, of the " Si-ts'ing-ku-kien," called "the twelve nip ples" (shi-ir ju{lli)) in the Chinese description, where the octangular star is explained as the " blossom of the water-chestnut " (ling-huaills'>). Chinese art critics are very cautious, refraining from throwing out hints as to what an ornament may possibly mean unless able to speak with a certain amount of confidence. An examination of the most ancient mirrors of the Han dynasty, as reproduced in the " Po-ku-t'u-lu," 1 how ever, seems to suggest the idea of " the twelve nipples " being intended to indicate the "twelve horary characters" (shi-ir ch'6n(-lle)) marking the divisions of the day. The " blossom of the water-chestnut " is well known in poetry; but, more than the flower itself, its representation on metallic mirrors has excited the imagination of Chinese bards, in witness whereof I would refer the sinological reader to a number of passages quoted in the " Pien-tzi'-lei-pien," 2(18) The most curious to us is the quotation from a poet, Sie F6ng,<117> who, according to his biography,3 flourished during the ninth century, and who says in his poetical account of a metallic mirror found by a ploughman in the fields, " The surface of the mirror having been polished for fully a month, one could in the [reflected] sunshine gradually discover the spread-out blossom of a water- chestnut." i (118) 1 Chap. 28, pp. 8-10 et passim. 2 Chap. 185, pp. 60 et seq., sub -verbo ling-hua.m" 3 T'ang-shu, Chap. 203, p. 17. * T'u-shu-tsi-ch'ong, Sect. 32, Chap. 227, i-won 11, p. 6. 38 CHINESE METALLIC MIRRORS 246 I am very much inclined to look upon this passage as an early men tion of a magic mirror; for, since the image became visible only after long-continued careful polishing (mo-jung, "made to shine by grind ing "), the gradual appearance of an image " in the sun " cannot refer to the ornament on the back of the mirror, but to its counterpart on its smooth surface. I would here suggest that a hint might be taken from the procedure resorted to in this one instance during the ninth century, and an attempt be made to polish the several " water-chestnut-blossom mirrors"1 (ling-hua-kingills) ) now in the Musee Guimet as a test of their power to reflect ornaments. The specimens in the K'ien-lung col lection, described in the " Si-ts'ing-ku-kien" as t'ou-kuang110^ (i.e., pos sessing the power of magic mirrors), cannot possibly have preserved the original smoothness of their front surface since the time of the Han dynasty, and that they are called t'ou-kuang can only be due to their having been repolished. If this be done with the necessary care, the experiment may succeed. Should it fail, however, this need not become an argument against the conclusions we may derive from the native literature as to the original meaning of the mirrors showing this kind of ornament: for it appears that counterfeits- — mirrors resembling the true " water-chestnut-blossom, or magic mirror," in all respects, without possessing the power of reflecting the ornament — exist in great numbers ; at least, so I am led to believe from the description of a speci men styled chau-sin-kingil20'> (i.e., The Mirror reflecting the Heart") in the " Kin-shi'-so," on fol. 54 B of the Mirror volume. The inscription on this mirror contains sixteen characters clearly distin guishable. (121) It seems a characteristic feature of many of the inscrip tions found on these magic mirrors, that, even apart from any appar ent mutilation to which they may have been exposed through the effects of time, it is difficult to make sense of what we read on the back. In the above-mentioned legend, for instance, the character zr(122) seems to be a mere ornamental expletive, inserted for the purpose of puzzling the reader. I suspect that in this legend, as well as in others, certain char acters were purposely not reflected through the mirror; possibly certain characters were but incompletely reflected, or slightly modified, in the shadow; all this being calculated to increase the surprise of the reader, who may have tried in vain to make sense of the inscription visible on the back, and who finds that its reading is easier than he thought, after he has seen its shadow reflected on a wall. The brothers Fong, editors of the " Kin-shi'-so," accompany the legend — which they reproduce with the above character as it appears on the back of the mirrors — with the 1 The name by which the poets in the Pien-tzi-lei'-pien call them. 39 247 FRIEDRICH HIRTH following remark: "Of this class, counterfeits are very frequent [the legends of which] cannot be made into sentences ; but the legend of this mirror is still perspicuous, for which reason we have here recorded it."(12S) The character z><122) ("and," the copulative particle), in similar inscriptions reproduced in the " Si-ts'ing-ku-kien," x is explained by t'iena2i'> ("heaven, day"). It seems quite possible that the difficulties now standing in the way of deciphering the inscriptions on the mirrors shown on Plate xv, Figs. 2-5, will disappear, if we succeed in obtaining an image of the shadow they produce on a white wall. The magic mirrors brought to Europe from Japan usually reflect a facsimile of the back ornament, it is true; but the existence of mirrors like the one in the possession of Dr. Milchner of Berlin, and exhibited at the April meeting of the Berlin Anthropological Society in 1898. 2 shows that this need not necessarily be the case. The mirror referred to shows on the back the Chinese characters given in No. 125 (p. 255), whereas the shadow thrown on the wall reveals the image of a human figure surrounded by an aureola from which emanates a network of lumi nous rays. The inscription was rightly explained by Professor F. W. K. Muller of the Berlin Ethnographical Museum as the Chinese tran scription of the Buddhist incantation " Namo Amitabha Buddha;" and the human figure represented by the shadow was proved by the same gentleman to be identical in shape with certain representations of this very deity found among the pictorial treasures of his museum.3 It must strike every observer that some of the mirrors coming, ac cording to the arguments I have tried to derive from Chinese literature, under the category of "magic mirrors," are rather thicker than the average; and this seems to favor my suspicion as to the reflected image being somewhat different from the one seen on the back, chiefly as regards the inscription. What the general sense of these inscriptions may be, we are perhaps allowed to guess from the legends of a series of mirrors discussed in the " Kin-shi'-so " 4 under the title Ji-kuang-king of the Han Dynasty /126> which, owing to the peculiarity of their back orna ment as well as to. their inscriptions, I strongly suspect to be " magic mirrors." The inscription on one of them says, for instance, kien ji-tchi kuang Tien-hia ta-ming^12^ (i.e., "When the light of day is visible, the world is full of brightness "). Chap. 39, pp. 13, 15, 17, and 19. 2 For illustrations see Verhandlungen, 1898, pp. 199 and 200. 3 See Verhandlungen, 1898, pp. 549 et seq. ' See fols. 45-48. 40 Boas Anniversary Volume. Plate XVI. CHINESE METALLIC MIRRORS, MUSEE GUIMET PARIS. CHINESE METALLIC MIRRORS 248 The Ts'ing Kai Mirror. Plate xvi, Fig. 1 : diameter, 9.75 cm. ; weight, 348 grams. The knob in the centre of the back is uncommonly large, and forms a perforated button fully 1.5 cm. in diameter and 0.75 cm. in height. The heads of two monsters are glaring at each other, possessing all the attributes of dragon-heads (rolling eyes, horns, tongue, antennae, with the peculiar shape of a camel's head), but being connected with the bodies of rep tiles covered by pustules instead of scales. The abdomen of the mon ster on the right hand stands out in exaggerated relief, no claws being distinguishable; whereas the animal on the left hand appears in dislo cated portions, which it requires some imagination to unite. At the bottom of the disk, opposite the two dragon-heads, we find a small naked human figure, and the very clumsy shape of some quadruped, probably meant to be a deer, or a ki-lin. The margin of the disk shows a series of patterns such as may be found among the mirrors of the Han dynasty. The characters Ts'ing Kai(12S> appear close to the neck of the monster on the left hand. This we must assume to have been the name of the first owner of the mirror. The same name occurs on the back of a mirror exhibiting similar ornaments (two dragon-heads, with the disjointed portions of the body of a doubtful reptile) and repro duced in the " Po-ku-t'u-lu." J The inscription of this mirror leaves no doubt as to the two characters Ts'ing Kai being the name of a person, since it begins with the words Ts'ing Kai tso king**29'1 (i.e., "A Mirror made by Ts'ing Kai"), this being the usual formula denoting the maker or first owner of mirrors in quite a number of ancient legends still on record. The Ts'ing Kai mirror of the " Po-ku-t'u-lu " contains an inscription of thirty- four characters/130' from which we are led to conclude that it was made at a time when " the four barbarians (the foreign nations on the boundary of China) had just been brought to sub jection, and that in consequence thereof the government as well as the people enjoyed peace; that, the Hu-lu<131) (the Tartars in the north, especially the Hiung-nu) being extinguished, the nation could enjoy the benefit of good harvests," etc. The allusion to the Hiung-nu having been " extinguished " may refer either to the decapitation of Chi-chi Shan-yii in 36 B.C., and the complete submission of his rival, the Hu-han-ye Shan-yii, to Chinese rule, or the victory of the Chinese Gen eral Tou Hien over the so-called Northern Shan-yii in A.D. 91. 2 Various other occasions are, however, equally possible. I have not 1 Chap. 28, p. 32. 2 See DE Mailla, Histoire de la Chine, Vol. in, pp. 174 and 395. 41 249 FRIEDRICH HIRTH succeeded in discovering in Chinese history any particular year in which a series of decisive victories over any of the Tartar nations in the north or west of China is followed by an exceptionally good year as regards the crops of grain. If such a year were on record, it would probably not have escaped the attention of the Chinese art critics, who would have otherwise made use of this argument in fixing the period in which the mirrors containing the above or a similar inscription must have origi nated. All that the brothers Fong suggest in the " Kin-shi'-so "x — where a large mirror is depicted containing an inscription similar to the one on the Ts'ing Kai mirror of the " Po-ku-t'u-lu," together with ornaments in the shape of two chariots drawn by four and six horses, and two winged Buddhas sitting on the characteristic padma-throne (lien- t'aill32)) — refers to the legend as one frequently found on works of the Han dynasty. The ornamentation reminds the native critic of the stone sculptures in the Wu-shi'-shan tombs, which have been shown by M. Chavannes to belong to the second century A.D. ; and the appearance of what must be looked upon as a Buddhist image proves that the mirror is not likely to have been made previous to the year A.D. 70, • when Buddhism was probably first introduced into China, and probably a good deal later. The " Kin-shi'-so " 2 contains three illustrations of mirrors with the owner's mark " Ts'ing Kai," two of them bearing inscriptions bespeaking good luck (the aversion of evil influences, profit in trade, a good career, and long life) to the owner and his offspring/133' the third one being identical in all respects with the specimen now in the Musee Guimet. I have never come across any mention of a personage named Ts'ing Kai, and we shall probably not succeed in finding out who he was; but from the fact of a mirror with an inscription attributed to the Han dynasty, though probably the latter part of the Eastern Han (say, any time between A.D. 70 and 221), being described in the " Po-ku-t'u-lu " as belonging to an owner of the same name, we may form an idea as to the period to which it belongs. The " Po-ku-t'u-lu " 3 contains an illustration of a mirror very similar to the Ts'ing Kai mirror of the Musee Guimet. The two monsters are similarly placed, the one on the right hand showing likewise an abdomen of exaggerated dimensions. The two characters Ts'ing Kai112^ appear in the same place behind the neck of the other monster; but the naked human figure is replaced by a quadruped which it seems difficult to identify ; and the other small figure, adjoining the tail of the 1 Vol. vi, fol. 21. 2 Vol. vi, fol. 26. 3 Chap. 29, p. 25. 42 CHINESE METALLIC MIRRORS 250 big-bellied monster, seems to be meant as a bird. The title given to this mirror by Wang Fu, the author of the " Po-ku-t'u-lu," is Han T'o- lung kienf13^ i.e., " Mirror [showing as ornaments the] T'o [water- lizard, triton, iguana] and the Lung [dragon], of the Han Dynasty." It appears that by t'o-lung, not two animals, but one animal is meant, — the " lizard-dragon," whatever that may be, — since we observe on fol. 20 of the Mirror volume of the " Kin-shi'-so " the representation of three reptiles which the editors explain as " three t'o-lung." ,(185) The Nippled Mirror. Plate xvi, Fig. 2: diameter, 15.66 cm. Under the heading Mei-ju-mon/136'' i.e., the class of nippled (mir rors), the " Po-ku-t'u-lu " x describes six mirrors, all belonging to the Han dynasty, two of which offer decided analogies with the specimen now in the Musee Guimet, more especially the one illustrated on p. 19. The mirror referred to is called Han pai-ju-kienil37) ("the mirror with a hundred nipples " ) . The ornament lies embedded in a star with six teen points. A smaller sixteen-pointed star surrounds the centre. The interval between these two star-margins is occupied by a broad zone bordered by two concentric rings. The surface of the broad zone is divided into four equal parts by four rosettes with mammiform centres, each of the four divisions showing sixteen nipples (;a<188)), the greater number of which are connected by raised lines. The central ornament contains eight nipples besides some fancy ornaments surrounding the knob. A comparison of the illustration in the " Po-ku-t'u-lu " with the above mirror will show that the two represent the same class of work. The size appears to be identical, to start with, since the Pai-ju-kien of the " Po-ku-t'u-lu" (p. 22) is described as measuring 5.8 Chinese inches in diameter. Since I have reason to believe that the inch of the Sung dynasty, in which the measurements of the " Po-ku-t'u-lu " are given, is the equivalent of about 2.7 cm., the Pai-ju-kien must have measured just about 15.66 cm., like the one in the Musee Guimet. The weight of the one described in the " Po-ku-t'u-lu " is 10.3 Hang, or so many ounces, which need not, of course, agree with the equivalent of the modern ounces or taels. The ornament will be found to offer some slight variations. The four rosettes are replaced by four ornaments of a sim ilar type (a central nipple surrounded by eight round leaves), and the number of nipples occupying the space between the four rosettes is 1 Chap. 29, pp. 19-29. 43 251 FRIEDRICH HIRTH reduced from sixteen to ten. The knob assumes the shape of a big mamma supported by eight smaller nipples, tapering conspicuously above all the remaining ornaments, and is surrounded by four smaller nipples. In his explanation of the ornament, Wang Fu refers to two other classes of bronze work in which the " nipple " ornament may also be looked upon as the chief characteristic; viz., the various ancient bells of the chung (" nipple "), seems to indicate that their being compared to the nipple of a breast is an after-thought of later interpreters. I am not sufficiently versed in the theories underlying the manufacture of bells to speak with authority on this subject, but it seems that the nipples on ancient bells had something to do with tuning, or regulating, the sound of bells. This I am led to infer from a remark made in Wang Fu's description of two nippled bells,3 where one Li Chau of the Sung dynasty(147) is said " to have held that nipples [on bells] were used to regulate the surplus sound, since sounding [a bell] without regulation results in jingling."<148) I do not pretend to produce the exact sense of this passage, since I possess no technical knowledge of bell-founding. If regulating sound, or tuning, be the real purpose for which the nipple ornament has been invented in the case of bells, the question arises, What does it mean in connection with the sacrificial vessel above mentioned, called Pai-ju-i/141^ and the various metallic mirrors distin guished by this " symbol of nutrition "? If any of the originals adorned with " nipples " should have a clear, bell-like ring when struck with a clapper, I would suggest that it might have served as a musical instru ment, apart from the category indicated by its ornamental shape. Note. — Since compiling these notes, I have received Professor Ed. Chavannes' paper " Le cycle ture des douze animaux," published in the "T'oung Pao," Serie 11, Vol. vii (Leiden, 1906), pp. 51-122, which contains numerous notices of metallic mirrors with special reference to the " Twelve Animals," also a number of illustrations, among which Fig. 4, being a reproduction of one of the finest specimens now in the Musee Cernuschi of Paris, is of particular interest. 1 Biot, Vol. 11, p. 499. 2 " L'entre-bord des ceintures est appele garniture de boutons, Me'i. Les boutons sont appeles les brilliants, king." The commentary adds, " Les boutons, Me'i, sont les mamelons de la cloche." " Po-ku-t'u-lu, Chap. 23, p. 26. 45 253 FRIEDRICH HIRTH tl * Chinese Characters. \fc fa m +& fe' £ $$£. ?m -/£ ff sm k >%>'$& 9$ ? 1$ @ *£ $tf '"* -ff" «*£,/£ »fl S « tf#& "& it ?.§ ^|£ ™A *#U| •%££ #- ii lip •%£ - |tt MJ» if ®# * # *s »m HA n° ® w f | ii If « it a * #p jg» f *& B 11 «a «# fK ## Bjf «jl^ 46 CHINESE METALLIC MIRRORS 254 ®S «* s. at s