FREER GALLERY SMITHSONIAN INSTITUTION WASH1NGTON,D.C. ISO FIELD MUSEUM OF NATURAL HISTORY PUBLICATION 177 ANTHROPOLOGICAL SERIES VOL. XIII, No. 2 CHINESE CLAY FIGURES PART I PROLEGOMENA ON THE HISTORY OF DEFENSIVE ARMOR BY BERTHOLD LAUFER Associate Curator of Asiatic Ethnology 64 Plates and 55 Text-figures The Mrs. T. B. Blackstone Expedition THE !TER ARY CONTENTS PAGE I. HISTORY OF THE RHINOCEROS 73 II. DEFENSIVE ARMOR OF THE ARCHAIC PERIOD . . 174 III. DEFENSIVE ARMOR OF THE HAN PERIOD . .- . . . 201 IV. HISTORY OF CHAIN MAIL AND RING MAIL 237 V. THE PROBLEM OF PLATE ARMOR 258 VI. DEFENSIVE ARMOR OF THE T'ANG PERIOD 292 VII. HORSE ARMOR AND CLAY FIGURES OF HORSES .... 306 CHINESE CLAY FIGURES PART I PROLEGOMENA ON THE HISTORY OF DEFENSIVE ARMOR I. HISTORY OF THE RHINOCEROS. An extensive collection of ancient clay figures gathered in the provinces of Shen-si and Ho-nan during the period from 1908 to 1910 is the basis of the present investigation. As the character of this material gives rise to research of manifold kinds, it has been thought advisable to publish it in two separate parts. Many of the clay statu- ettes which form the nucleus of our study are characterized by the wear of defensive armor, hence this first part is devoted to an inquiry into the history of defensive armor, a task of great interest, and one which here- tofore has not been attempted. It will be recognized that this subject sheds new light on the ancient culture of China and her relations to other culture zones of Asia. The second part of this publication will deal in detail with the history of clay figures, the practice of interring them, the religious significance underlying the various types, and the culture phase of the nation from which they have emanated. Before embarking on our subject proper, a preliminary question must be decided. It is the tradition of the Chou period that the cuirasses l employed at that time were manufactured from the hides of two animals designated by the words se (No. 10,298) and si (No. 42i8). 2 It is imperative to have a clear understanding of what these two animals were in the early antiquity of China. As this problem is still pending, and as a close and coherent investigation of the matter has never been made, I have decided to treat it from the very beginning by means of all accessible methods, with the possible hope of a final solution. The present state of the problem is as follows: EDOUARD Biox, 3 1 "Cuirass" or "cuirbouilly" is the right term for this kind of armor, as these words (like French cuirasse, Italian corazza) go back to Latin coratium ("a breast- plate of leather"), derived from the word corium ("leather"). 2 These figures refer to the numbers of the Chinese characters in the Chinese- English Dictionary of H. A. GILES. 1 Le Tcheou-li, ou Rites des Tcheou, Vol. II, p. 507 (Paris, 1851). 73 74 CHINESE CLAY FIGURES the ingenious translator of the Chou li, has expressed his opinion in these words: "I translate by buffalo the character si, and by rhinoceros the character se. These two characters * denote in the Shi king a rhinoceros or a wild buffalo, without the possibility of distinguishing between them. The skin of the rhinoceros being very thick, it seems difficult to believe that it could have been sliced, and that the pieces were sewed together, in order to make cuirasses. In this case the two characters of the text 2 would designate here two species of buffalo." 3 PALLADIUS, in his Chinese-Russian Dictionary, treats the matter in the opposite way, and renders se by (i) "an animal resembling a wild ox," (2) "Malayan rhi- noceros," and si by "rhinoceros." COUVREUR credits the word se first with the latter meaning, secondly with that of bceuf sauvage.* CnAVANNES 6 has clearly and sensibly expressed the opinion that 1 It should properly read, "words." 1 Referring to the passage of the Chou li where the hide cuirasses are mentioned. 8 In his essay on the Manners of the Ancient Chinese (in LEGGE, Chinese Classics, Vol. IV, Prolegomena, p. 148), BIOT says that "they hunted also herds of deer, of boars, of wild oxen," on which LEGGE annotates, "These wild oxen would seem to be rhinoceroses." But in his original article (Journal asiatique, 1843, p. 321), BIOT has added the following comment: "Le caractere si est traduit ordinairement par rhinp- ce'ros, et c'est, en effet, son sens actuel. Lacharme a traduit, tantdt bos sylvestris, tantdt rhinoceros. II me semble que les grandes chasses devaient etre dirige'es surtout contre des troupeaux de boeufs sauyages ou buffles." The objections raised by Biot in the above passage are not valid; it is certainly possible to slice rhinoceros-hide, and to sew the pieces together. Cuirasses and shields have been made from it, as may be seen from many specimens in the collections of our museums. A shield of rhinoceros- hide is illustrated in Plate XXVII. In accordance with the above definition, BIOT, likewise in his translation of the Annals of the Bamboo Books (Extrait du Journal asiatique 1841 and 1842, pp. 41, 46), rendered se by "rhinoceros" and si by "bceuf- si (rhinoceros)," while LEGGE (Chinese Classics, Vol. Ill, Prolegomena, pp. 149, 153) in both cases has "rhinoceros." It will be seen in the course of this investigation how Biot's error was caused, and that his opinion is untenable. W. R. GINGELL (The Ceremonial Usages of the Chinese, p. 81, London, 1852) treated the two words in a way opposite to that of Biot, translating in the passage of Chou li the term si kia by "rhinoceros-hide armor" and se kia by "wild buffalo's-hide armor." No one of those who from purely philological points of view proposed the rendering "wild buffalo" has ever taken the trouble to raise the question whether anything like wild buffalo exists in China, anciently or in modern times. BUSHELL (The Stone Drums of the Chou Dynasty, Journal China Branch R. As. Soc,, Vol. VIII, 1874, P- I 54) was of the opinion that the ancient Chinese hunted the rhinoceros in the low swamps. 4 The passage in Lun yu (xvi, 7) is translated by COUVREUR (Les quatre livres, p. 250), " Si un tigre ou un boeuf sauvage s'e'chappe de sa cage." Nevertheless in the glossary (p. 664) thejword se is rendered by "rhinoceros." LEGGE (Chinese Classics, Vol. I, p. 307) translates here "rhinoceros," despite Chu Hi's (undoubtedly wrong) interpretation of se being a ye niu ("wild bull "). In his first edition of Lun yii (which is not accessible to me, but this may be gleaned from PLATH, Die Beschaftigungen der alten Chinesen, p. 56), LEGGE translated se by "wild ox." In the text of Mtng-tse (III, 2, ix, 6), LEGGE (Classics, Vol. II, p. 281) and COUVREUR (/. c., p. 452) are in mutual accord in translating the word si by "rhinoceros," and this is likewise the case with reference to the word se in Li ki, II, I, in, 40 (LEGGE in Sacred Books of the East, Vol. XXVII, p. 158; COUVREUR, Li ki, Vol. I, p. 181). In Tsp chuan, vn, 2, LEGGE (Classics, Vol. V, p. 289) renders si se by "rhinoceroses and wild bulls." 8 Les M^moires historiques de Se-ma Ts'ien, Vol. Ill, p. 502. HISTORY or THE RHINOCEROS 75 se niu and s* appear to be two different species of rhinoceros. Also G. DEVERIA* has translated se and si by " rhinoceros." BRETSCHNEIDER, both a naturalist and an eminent sinologue, upheld the opinion that the rhinoceros, and goblets made from rhinoceros-horn, are repeatedly mentioned in the Chinese classics, and that the latter has been reputed from time immemorial for its antipoisonous virtues. He refers the saying that rhinoceros-horn cures all poisons, to the Shtn-nung pfai ts'ao king, attributed by tradition to the mythical Emperor Sh&n-nung, at all events the most ancient Chinese materia medica in existence. 2 In the first edition of his Chinese-English Dictionary, Professor GILES, the eminent sinologue at the University of Cambridge, Eng- land, attributed to both se and si the meaning of "rhinoceros," with- out establishing a distinction between the two. In the second edition, however, we read under se (No. 10,298), "A bovine animal, figured as a buffalo with one horn, known as the se niu. Another name for the si 4128; see 8346 for its confusion with the rhinoceros." Under the last- named heading it is said that the term si niu is "a bovine animal, figured as a buffalo with a single horn;" with the addition that the traditional "rhinoceros" of foreigners seems to be wholly wrong. Further, the reader is requested to correct No. 4128 si, where the meanings "tapir" and "rhinoceros" had been given. In his "Adver- saria Sinica" (p. 394), Mr. GILES has expounded more in detail the reasons which induced him to make these alterations. The arguments advanced by him are briefly three: i. The rhinoceros is known to the Chinese as pi kio, "nose-horn." 2. In two passages of Chao Ju-kua (translation of HIRTH and ROCKHILL, pp. 118, 233), rhinoceroses are spoken of as being shot with arrows, while Giles finds it stated in the T*u shu tsi ch'dng that arrows cannot pierce the hide of the rhinoceros. 3. The si and the se are figured in the latter work as slightly differing 1 Histoire des relations de la Chine avec 1'Annam, p. 88 (Paris, 1880). 1 Chinese Recorder, Vol. VI, 1875, p. 19, and Mediaeval Researches, Vol. I, p. 153. Regarding the materia medica current under the name of Sh6n-nung see BRET- SCHNEIDER (Botanicon Sinicum, pt. i , pp. 27-32). BRETSCHNEIDER, though believing that in India the people from time immemorial attribute the same antipoisonous vir- tues to the rhinoceros-horn as the Chinese do, says he cannot believe that the Chinese have borrowed this practice from the Hindu or vice versa. The Hindu conception is not attested by any passage in Sanskrit literature, but only by Ctesias and Aelian who state that drinking-vessels made from the horn of the unicorn safeguard from poison and various diseases. The belief is likewise absent among the Greeks and Ro- mans, in whose records the number of references to rhinoceros-horn is exceedingly small (H. BLUMXER, Technologic und Terminologie der Gewerbe und Kunste, Vol. II, p. 358). There is no evidence that the Chinese notions are due to any stimulus received from outside; they appear, on the contrary, as legitimate offshoots grown on Taoist soil. The Chinese likewise conceived the idea of carving rhinoceros-horn into cups, girdle-plaques, and fanciful ornaments. We shall come back to these various points in detail. Compare p. 154, note. 7 6 CHINESE CLAY FIGURES FIG. 1. Monoceros of European Armorial Style, introduced into China by the'Jesuitj,Father Ferdinand Verbiest (from T'u shu tsi ch'&ng). HISTORY OF THE RHINOCEROS 77 FIG. 2. Rhinoceros, Design of European Origin, introduced into China by the Jesuit Father Ferdinand Verbiest (from T'u fku tsi ck'tng). 78 CHINESE CLAY FIGURES bovine animals, 1 with a single horn on the head. Says Mr. GILES, "The Erh ya says: the latter is like an ox, and the former like a pig, while the Shan hai king speaks of both as occurring in many parts of China. There is thus hopeless confusion, of which perhaps the explana- tion is that a term which originally meant a bovine animal was later on wrongly applied to the rhinoceros." The first argument advanced by Mr. Giles is not admissible as good evidence in the case. "The rhinoceros is known to the Chinese as pi kio, 'nose-horn,' and is approximately figured in the T'u shu." By referring to the Chinese cyclopaedia we find, however, that this name with the illustration is extracted from the K'un yii t'u shuo. The latter is not the production of a Chinese author, but of the Jesuit FERDINAND VERBIEST, born in 1623, and who arrived in China in 1659 and died in i688. 2 This section of the T'u shu tsi ch'&ng alluded to by Mr. Giles and devoted to "strange animals" contains quite a number of illustra- tions and texts derived from the work of Verbiest; and neither his zoological nomenclature nor his descriptions and illustrations, which are based on European lore, can be laid at the door of the Chinese. The evidence is here produced in Figs, i and 2. In Fig. i, Verbiest pictures a "single-horned animal" (tu kio shou), saying, "India, situated on the continent of Asia, is the habitat of the single-horned animal which is as big as a horse, very light and swift, and yellow in color. On its head it has a horn, four to five feet long, of bright color. It is made into drinking-vessels which are capable of neutralizing poison. As the horn is pointed, the animal can charge a big lion. The lion, while struggling with it, takes refuge behind a tree; and when missing its aim, it butts the tree, while the lion bites it at this moment." In Fig. 2, the pi kio shou referred to by Mr. Giles is pictured. Verbiest com- ments, "The locality Kang-pa-ya 3 in India, situated on the continent of Asia, is the habitat of an animal called 'nose-horn' [rendering of 'rhi- noceros']. Its body is as powerful as that of the elephant, but its feet are somewhat shorter. Its trunk is covered all over with red and yellow spots, and is overlaid with scales. Arrows cannot pierce it. On its nose there is a single horn as strong as steel. It prepares for its battles with the elephant by whetting its horn on the rocks; and hitting 1 This is a debatable point. The two illustrations do not resemble bovine animals, but deer (see Figs. 9 and 10 on pp. 102 and 103). The "bovine animal with one horn" first appears in LIONEL GILES, An Alphabetical Index to the Chinese Encyclopaedia, p. 5 (London, 1911). 2 WYLIE, Notes on Chinese Literature, p. 58; M. COURANT, Catalogue des livres chinois, p. 95; H. CORDIER, L'imprimerie sino-europe'enne en Chine, p. 59; P. PELLIOT Bulletin de I' Ecole frangaise d' Extreme-Orient, Vol. Ill, 1903, pp. 109, 115. * That is, Khambayat or Cambay, in the western part of the province of Gujarat. HISTORY or THE RHINOCEROS 79 the elephant 's paunch , it kills it . " The alleged combats of the rhinoceros with the lion and elephant are classical reminiscences (see p. 84) which are absent from Chinese folk-lore. Verbiest repeats the popular tradi- tions current at his time in Europe, and like Cosmas Indicopleustes, still discriminates between the monoceros or unicornis (tu kio) and the rhinoceros (pi kio), illustrating the former by the unicorn of European heraldry. Consequently the terms employed by Verbiest are literal translations of European nomenclature into Chinese, made by Verbiest for his purpose; and the word pi kio cannot be claimed, as has been done by Mr. Giles, as a genuine term of the Chinese language. It is a foreign term not employed by the Chinese. Indeed, in a long series of Chinese texts dealing with the rhinoceros, and given below, not any use of this name is made. Only a single case is known to me: the Manchu- Chinese dictionary Ts'ing wtn pu hui of 1786 (Ch. 4, p. 23) explains the Manchu word sufen by the said pi kio, adding the definition, "a strange animal bred in Cambaya in India, like an elephant, with short feet, " etc., the same as given by Verbiest. This, accordingly, is a mere repetition of the latter's statement, and is not conclusive. Curiously enough, that expression which Mr. Giles credits as the only authentic word for "rhinoceros" is given a quite different meaning in the Polyglot Dictionary of K'ien-lung (Appendix, Ch. 4, p. 75), where we find the series Chin. pi kio shou, Manchu sufen, Tibetan ba-men, Mongol bamin. The Tibet- an word ba-men, reflected in Marco Polo's beyamini, 1 denotes the gayal wild ox (Bos gavaeus). Whether this equation, as a matter of fact, is correct, is certainly a debatable question; but this point does not concern us here. The point to be brought out is that pi kio in the sense of "rhinoceros" is a term coined by Verbiest, and that it has not yet been pointed out in any Chinese text prior to his time. 2 Simultaneously Mr. Giles's argument directed against Hirth "the T'u shu expressly 1 See the writer's Chinese Pottery, p. 260, note 4. 1 The general Chinese expression for rhinoceros-horn which is even now traded to Canton and there made into carvings is still 5* kio; hence it follows that at the present day the designation of the animal itself, as it has been for several millenniums, is the word si. The English and Chinese Standard Dictionary of the Commercial Press, issued by a commission of Chinese scholars, who must know their language, renders the word "rhinoceros" into se niu and se (Vol. II, p. 1919). COUVREUR (Diet, francais-chinois, 2d ed.) has likewise se niu. DOOLITTLE (Hand-Book of the Chinese Language, Vol. I, p. 411) gives under "rhinoceros" si, se niu, and si niu. SCHLEGEL (Nederlandsch-chineesch Woordenboek, Vol. Ill, p. 622) renders the word by se, si, and si niu. True it is that in recent times the words se and si have been transferred to bovine animals, and the Chinese themselves are well aware of this fact. Thus Li Shi-chen, in his Pen ts'ao kang mu, remarks that the term "hairy rhinoceros "is at present referred to the yak (see p. 1 50). This, however, as will be established by abun- dant evidence, was not the case in former times. In fact, these recent adjustments prove nothing for conditions which obtained in earlier periods. The question as to how the word se became transferred to the buffalo is discussed on p. 161, note 5. 8o CHINESE CLAY FIGURES says that arrows cannot pierce the hide of the rhinoceros" falls to the ground. This is a verdict of Verbiest, and not to be encountered in any Chinese report regarding the rhinoceros. It is, moreover, an argument of no meaning and no value; it is simply a popular notion of fabulous character. The numerous stories formerly current anent the rhinoceros chiefly culminated in three points, its ferocity, the use of its horn as a weapon of attack, and its invulnerability. These notions have been refuted by close observation. We quote an authority, R. LYDEKKER : * " Fortunate- ly, in spite of stories to the contrary, the creature in its wild state appears to be of a mild and harmless disposition, 2 seeking rather to escape from 1 The Game Animals of India, Burma, Malaya, and Tibet, p. 31 (London, 1907). 2 Certainly; it is easily kept in confinement and tamed, and has often been trans- ported over vast tracts of water and land. A good example of the overland trans- portation of a tamed rhinoceros or several animals is furnished by Se-ma Ts'ien, in the chapter on the Imperial Sacrifices to Heaven and Earth, when this animal together with an elephant was conducted as far as the foot of Mount T'ai in Shan-tung with a possible view to their being sacrificed; but the Emperor spared their lives, and the animals were allowed to return (see CHAVANNES, Les Me"moires historiques de Se-ma Ts'ien, Vol. Ill, p. 502). The following tributes of living rhinoceroses are on record. In the year 2 A.D. the country Huang-chi (south of Tonking, 30,000 li from the capital of China) sent a living rhinoceros as tribute to the Court of China, as mentioned three times in the Ts'ien Han shu (Ch. 27 B, p. 17 b). These texts have recently been studied by PAUL PELLIOT (T'oung Poo, 1912, pp. 457-460), who has revealed their fundamental importance for the history of Chinese relations with the countries of the Indian Ocean in the first century of our era. On the basis of Pelliot's transla- tions, the country Huang-chi has recently been made the object of an interesting geographical study on the part of A. HERRMANN (Ein alter Seeverkehr zwischen Abessinien und Siid-China bis zum Beginn unserer Zeitrechnung, Zeitschrift der Gesellschaft fur Erdkunde zu Berlin, 1913, pp. 553-561). This author identifies Huang-chi with Abyssinia mainly on the ground that the rhinoceros occurs there. This argument is not cogent, since the home of the animal is in all parts of both In- dias, Borneo, Java, and Sumatra as well. Also for other reasons this identification is unfortunate. The transportation of a live rhinoceros from Abyssinia to China over a maritime route would have been a feat impossible in those days, in view of the im- perfect state of navigation, while it could easily have been accomplished, if Huang-chi, as assumed by me, was located on the Malayan Peninsula ; and as shown by the Chinese records, the live rhinoceroses all hailed from Indo-China or Java. The name Huang- chi, moreover, cannot be derived from Aghazi, as HERRMANN thinks. His decisive argument in support of this theory is, of course, the statement in the Chinese text that Huang-chi is 30,000 li distant from Ch'ang-ngan, the then capital of China. Mr. Herrmann unreservedly accepts this as a fact, and is in this manner carried away to eastern Africa. We have known for a long time (in fact, the Jesuits of the eigh- teenth century knew it) that the Chinese definitions of distances over maritime routes must not be taken at their surface value. Nor have we any reason to be more Chinese in this respect than the Chinese themselves. The following is expressly stated in the Sung shu, the History of the Liu Sung Dynasty (420-478 A.D.; Ch. 91): "The southern and south-western barbarians, generally speaking, live to the south and south-west of Kiao-chi (northern Annam), and also inhabit the islands in the great ocean; the distance is about three to five thousand li for those that are nearer, and twenty to thirty thousand li for those that are farther away. When sailing in a vessel it is difficult to compute the length of the road, and therefore we must recollect that the number of li, given with respect to the barbarians of the outer countries, must not be taken as exact" (see GROENEVELDT, in Miscellaneous Papers relating to Indo-China, Vol. I, p. 127). It is plainly indicated in this passage that the distances HISTORY OF THE RHINOCEROS 81 its enemies by flight than to rout them by attack. When badly wound- ed, or so hustled about by elephants and beaters as to become be- wildered, a rhinoceros will, however, occasionally charge home. In such onslaughts it is the common belief that the animal, like its African cousins, uses its horn as its weapon of offence; but this is an error, the real weapons being the triangular, sharp-pointed low tusks." The same author states in another work 1 on the skin of the animal, "From the immense thickness and apparent toughness of its enormous folds, it was long considered that the hide of the Indian rhinoceros was bullet- proof, and that the only places where the animal was vulnerable were the joints of the armor. ... As a matter of fact, the skin of the living animal is quite soft, and can readily be penetrated in any place by a bullet, or easily pierced by a hunting knife. When dried it becomes, however, exceedingly hard; and it was formerly employed by the Indian princes in the manufacture of shields for their soldiery." given for the routes in the southern ocean are not exact, and that a description of twenty to thirty thousand li is nothing but a convention to denote the very remote barbarians of the south. Compare, on Chinese calculations of sea-routes, particularly G. SCHLEGEL (T'oung Pao, Vol. Ill, 1892, pp. 161-5). In Hou Han shu (Ch. 116, p. 3 a) the location of Huang-chi is positively indicated as being south of Ji-nan (Ton- king), which means that it was situated on the Malayan Peninsula. In 84 A.D. the Man I beyond the boundary of Ji-nan offered to the Court a living rhinoceros and a white pheasant (Hou Han shu, Ch. 1 16, p. 3 b). In 94 A.D. the tribes in the south- west of Sze-ch'uan sent an envoy and interpreter presenting a rhinoceros and a big elephant (ibid., Ch. 116, p. 8 b). At the time of the Emperor Ling (168-188 A.D.) of the Later Han dynasty, Kiu-chenin Tonking despatched a living rhinoceros to the Chinese Court (Huan yu ki, and Ta Ming i t'ung chi, ed. of 1461, Ch. 90, fol. 5, where it is said also that at the time of the Yuan dynasty [1260-1367] Annam presented a rhinoceros). In 539 Fu-nan sent a live rhinoceros (Liang shu, Ch. 54, p. 4). A similar report in regard to the country of Ho-ling (Java) occurs in 819 A.D. at the time of the T'ang dynasty (Kiu T'ang shu, Ch. 197, p. 2 b). Finally the poets Yuan Chen (779-831; GILES, Biographical Dictionary, p. 964) and Po Ku-i have celebrated in verse a tame rhinoceros which had been sent as tribute in the year 796 ; it was housed in the Shang-lin palace, and an official was appointed to care for it; but in the winter of the following year when great cold set in, the poor creature died. In 1009 Kiao-chi (Annam) presented a tame rhinoceros to the Court (Sung shi, Ch. 489), and there are other similar reports by the essayists of the Sung period. TA VERNIER (Travels in India, ed. V. BALL, Vol. I, p. 1 14) saw a rhinoceros eating stalks of millet presented to it by a small boy; encouraged by this sight, the traveller seized some stalks, and the rhinoceros at once approached him, opening its mouth four or five times; he placed some stalks in it, and when the animal had eaten them, it continued to open its mouth to receive some more. Tame rhinoceroses, to which a good deal of freedom was allowed, were formerly not uncommonly kept by the Rajas of India. Surely, not only men, but also animals, are usually better than their reputation among men. One of the most notable facts about the behavior of the rhinoceros in captivity, as al- ready observed by DARWIN (The Variation of Animals and Plants under Domestica- tion, Vol. II, p. 165, Murray's edition, 1905), is that under this condition it breeds in India far more readily than the elephant. The captive elephants, in contrast to the rhinoceros, as pointed out by Darwin and confirmed by others (E. HAHN, Kultur- geschichte der Haustiere, p. 37), but very rarely breed; as a rule, they do not even copulate. There is no doubt that the rhinoceros possesses the qualities fitting it for domestication, and that only the lack of promising advantages has prevented man from embarking on such a plan. 1 The New Natural History, Vol. II, pp. 1055-1056. 82 CHINESE CLAY FIGURES Naturally the skin of the animal is as soft and sensitive as that of any other living creature, and arrows are certainly painful to it. Only when properly prepared and dried does the skin assume that iron-like hardness which has achieved its reputation and probably caused the fable of its being impenetrable in the live beast. The account of the Arab envoy given in 993 to the Chinese Emperor, that "to capture a rhinoceros, a man with a bow and arrow climbs a big tree, where he watches for the animal until he can shoot and kill it," as narrated by Chao Ju-kua, is entirely trustworthy. 1 The fable lies entirely in the "arrows cannot pierce the hide," to which Mr. Giles gives credence. When it is said, "he rips up a man with his horn," Chao Ju-kua simply accepts the belief of all his contemporaries, eastern and western; and the remark certainly proves that he speaks of the rhinoceros, while it is no argument in favor of Mr. Giles's opinion that the animal in question is not the rhinoceros. While the general result at which Mr. Giles has arrived is not novel, being partly anticipated, as we have seen, by Biot, Palladius, and Couvreur, his arguments, as summed up above under No. 3, are original, and deserve serious consideration and discussion. What appears to Mr. Giles as the most weighty evidence in favor of his view are the queer Chinese illustrations of the two animals. Queer they are, but we must make an attempt at understanding and explaining them. For this reason, we shall first enter on a somewhat lengthy digression into the iconography of the rhinoceros; and it will be seen that this, as every- 1 The effect of arrows on the rhinoceros is well illustrated in the following story of CASPAR CORREA, who went to India in 1512, and wrote a detailed chronicle of the Portuguese possessions there. He describes a battle of Bang Cacandar, who availed himself of elephants fighting with swords upon their tusks, and in front of them were arrayed eighty rhinoceroses (gandas) "carrying on their horns three-pronged iron weapons with which they fought very stoutly . . . and the Mogors with their arrows made a great discharge, wounding many of the elephants and the gandas, which as they felt the arrows, turned and fled, breaking up the battles" . . . (quoted by YULE and BURNELL, Hobson-Jobson, p. 363). In India rhinoceroses were hunted with sabre, lance, and arrows. Timur killed on the frontier of Kashmir several rhi- noceroses with sabre and lances, although this animal has such a hard skin that it can be pierced only by extraordinary efforts (PETis DE LA CROIX, Histoire de Timur Bee, Vol. Ill, p. 159, quoted by YULE and BURNELL, Hobson-Jobson, p. 762). In Baber's Memoirs (quoted ibid.) a rhinoceros-hunt is described in these words: "A she rhinoceros, that had whelps, came out, and fled along the plain; many arrows were shot at her, but . . . she gained cover." The hunters of Java hide sickle-shaped knives under the moss on steep mountain-paths; the animal, dragging its paunch almost close to the ground, rips up itself, and is then easily mastered (P. J. VETH, Java, Vol. Ill, p. 289, Haarlem, 1903). HOSE and McDouGALL (The Pagan Tribes of Borneo, Vol. I, p. 145, London, 1912) have this observation to report: "Punans, who hunt without dogs (which in fact they do not possess), will lie in wait for the rhinoceros beside the track, along which he comes to his daily mud-bath, and drive a spear into his flank or shoulder; then, after hastily retiring, they track him through the jungle, until they come upon him again, and find an opportunity of driving in another spear or a poisoned dart through some weak spot of his armor." HISTORY OF THE RHINOCEROS 83 thing else connected with the animal, is an attractive subject of great culture-historical interest. It should be stated at the outset that the Chinese sketches pointed out by Mr. Giles, and other Chinese illustra- tions as well, can never have been intended for any bovines, whatever the alleged bovine character in the animal may be; for there is in this world no bovine animal with a single horn and three toes which, as will be shown, appear in the early Chinese definition, and are plainly out- lined in the sketch of the rhinoceros said in the Erh ya to be of hog-like appearance (Fig. 6). 1 The single horn and the three toes, however, are thoroughly characteristic of the rhinoceros, and of this animal exclusively. But we are first going to study the psychology of the case. On the first day of May of the year 1515 the first live rhinoceros was brought to modern Europe from India by Portuguese, and presented to King Emanuel of Portugal. 2 In commemoration of this event, Albrecht Diirer, who took a deep interest in exotic animals and people, sketched in the same year a likeness of this rhinoceros, published as a wood- engraving, with a somewhat lengthy description in German. Durer's original drawing is still preserved in the British Museum (Plate IX). 8 It is so weak that, as already pointed out by Dr. PARSONS,* the first serious 1 See likewise Fig. 9, p. 102. I The history of this event is narrated in the Decadas de Asia of J. DE BARROS (quoted by YULE and BURNELL, Hobson-Jobson, p. 363): "And in return for many rich presents which this Diogo Fernandez carried to the King, and besides others which the King sent to Affonso Alboquerque, there was an animal, the biggest which Nature has created after the elephant, and the great enemy of the latter . . . which the natives of the land of Cambaya, whence this one came, call Ganda, and the Greeks and Latins Rhinoceros. And Affonso d'Alboquerque sent this to the King Don Man- uel, and it came to this Kingdom, and it was afterwards lost on its way to Rome, when the King sent it as a present to the Pope." I 1 am indebted to Mr. Laurence Binyon of the British Museum for his courtesy in favoring me with a copy of this wood-engraving, from which our reproduction is made. The particulars of the history of this engraving are discussed by C. DODGSON (Catalogue of Early German and Flemish Woodcuts in the British Museum, Vol. I, p. 307, London, 1903). * Die naturliche Historic des Nashorns, welche von Doctor PARSONS in einem Schreiben an MARTIN FOLKES, Rittern und Prasidenten der Koniglich-Englischen Societat abgefasset, mit zuverlaissigen Abbildungen versehen, und aus dem Englischen in das Deutsche ubersezet worden von Doctor GEORG LEONHART HUTH, Nurnberg, bey Stein und Raspe, 1747. The English original of this interesting pamphlet of 16 pages in quarto is not known to me. It is accompanied by three plates engraved on copper representing the first fairly exact figures of the rhinoceros in various views, its horn and other organs of its body. An anonymous copper-engraving was pub- lished in 1748 under the title, "Vera effigies Rhinocerotis qui in Asia, et quidem in terris Mogolis Magni in regione Assam captus et anno 1741 tertio aetatis anno a capitano Douvemont van der Meer ex Bengala in Belgium translatus est." This rhinoceros, a three years old animal, was exhibited in Holland in 1741, and styled on the placards the behemoth of the Bible (JOB, 40) and the unicorn of mediaeval times. It proved an overwhelming sensation. In 1747 it made its appearance at Leipzig where GELLERT set it a literary monument in the poem with the beginning, "In order to behold the rhinoceros, I was told by my friend, I resolved to stroll out." In 8 4 CHINESE CLAY FIGURES student of the anatomy of the rhinoceros, it is impossible to assume that he had ever seen the animal. This fact is quite certain, for it is known that the King of Portugal despatched the animal to the Pope, and that it was drowned off Geneva when the vessel on board which it was being carried was foundered. The only supposition that remains, therefore, is that some one of Lisbon near King Emanuel must have sent on to Diirer a rough outline-sketch of the novel and curious creature, which was im- proved and somewhat adorned by the great artist. But to what sources did he turn for information on the subject? Naturally to that fountain- head from which all knowledge was drawn during that period, the au- thors of classical antiquity. The fact that Durer really followed this procedure is evidenced by the very de- scription of the animal, which he added to his sketch, and in which he reiterates the story of the ancients regarding the eter- nal enmity and struggle of rhinoceros and elephant. 1 The most curious feature about Durer's rhinoceros is Marble Relief of Two-Horned Rhinoceros in Pompeii . (from O. Keller, Antike Tierwelt). that, besides the hOITl On 1748 it reached Augsburg where Johann Ridinger made a drawing and etching of it with the title as stated (L. REINHARDT, Kulturgeschichte der Nutztiere, p. 751, Munchen, 1912). The rhinoceros is a subject which for obvious reasons has seldom tempted an artist. It should be emphasized that no artist has ever made even a tolerably good sketch of it, and that only photography has done it full justice. 1 According to the tales of the ancients, the feuds between the two animals were fought for the sake of watering-places and pastures; and the rhinoceros prepared it- self for the combat by sharpening its horn on the rocks in order to better rip the arch- enemy's paunch which it knows to be its softest part (compare DIODOR, i, 36; AELIAN, Nat. animalium, xvii, 44; PAUSANIAS, ix, 21; and PLINY, Nat. hist., vin, 20: alter hie genitus hostis elephanto cornu ad saxa limato praeparat se pugnae, in dimicatione alvum maxime petens, quam scit esse molliorem). The same story is still repeated by JOHAN NEUHOF (Die Gesantschaft der Ost-Indischen Gesellschaft [1655-57], p. 349, Amsterdam, 1669) in his description of the Chinese rhinoceros, which is based on classical, not Chinese reports: " It makes permanent war on the elephant, a_nd when ready to fight, it whets its horn on stones. In the struggle with the elephant it always hits toward its paunch where it is softest, and when it has opened a hole there, it desists, and allows it to bleed to death. It grunts like a hog; its flesh eaten by the Moors is so tough that only teeth of steel could bite it." The Brahmans allowed the flesh of the rhinoceros to be eaten as a medicine (M. CHAKRAVARTI, Animals in the Inscriptions of Piyadasi, Memoirs As. Soc. of Bengal, Vol. I, p. 371, Calcutta, 1906); according to al-Berunl (SACHAU, Alberuni's India, Vol. I, p. 204), they had the privilege of eating its flesh. CTESIAS stated wrongly that the flesh is so bitter that it is not eaten. HISTORY OF THE RHINOCEROS 85 its nose, it is provided with another smaller horn on its neck. This proves that he must have read about a two-horned rhinoceros, for the specimen shipped to Portugal was the single-horned species of India. MARTIAL, in one of his epigrams (Sped. Ep. XXII), has the verse, "namque gravem gemino cornu sic extultt ursum." As long as the fact of a two-horned rhinoceros was not yet scientifically established, and Dr. Parsons was one of the first to point it out, the critics of Martial felt greatly embarrassed over the statement that a rhinoceros with double horn 1 should have lifted a bear, and arbitrarily changed the verse in various ways to get around the double horn. Diirer no doubt had this passage in mind, and accepted it as a fact. Nobody at that time, however, knew the location of the second horn: thus it found its place on the neck. 2 This case is very instructive, for the Chinese 1 The two-horned African rhinoceros is figured on the bronze coins of Emperor Domitian and on Alexandrian coins of the same emperor (IMHOOF-BLUMER and KEL- LER, Tier- und Pflanzenbilder auf Mflnzen und Gemmen, Plate IV, 8), and unmis- takably referred to by PAUSANIAS (/. c.), who describes it as having the one horn on the extremity of its nose, the other, not very large, above the latter. The struggle between bear and rhinoceros is represented on a pottery lamp from Labicum, which is reproduced in Fig. 7 after O. KELLER (Tiere des classischen Altertums, p. 118, Innsbruck, 1887), in order to illustrate the affinity of this creature with the "hog-like " rhinoceros of the Chinese (Fig. 6). Diirer 's picture formerly led astray many a student of classical antiquity by giving the impression that a horn was really growing up from the animal's back. Thus S. BOCHART, in his Hierozoicon (p. 931, Lugduni Batavorum, 1692), a learned treatise on the animals mentioned in the Bible, makes the following observation with reference to the verse of Martial above quoted: " Frustra etiam id observatur, Rhinpcerotem geminum habere cornu. Alterum enim est in dorso, quo ursum extulisse dici non potest. Itaque ad illud cornu non pertinent haec poetae: gemino cornu sic extulit ursum." It was Bochart who proposed several conjectures tending to ameliorate Martial's text. JOHANNES BECKMANN (De historia naturali veterum libellus primus, p. 129, Petropoli et Goettingae, 1766) was the first to point out emphatically the actual truth in the matter, in these words : ' ' Sed non soli philologi, verum etiam physici duo cornua neglectis illis veterum locis [i.e., the passages of Martial and Pausanias] negarunt Rhinoceroti; uti Scheuchzeras, Peyerus. Consultius fuisset nee affirmare nee negare. Hodie enim auctoritatibus gravissi- morum virorum satis probatum est, esse Rhinocerotes etiam bicornes, qui cornu alterum non in fronte, non in dorso, sed etiam in nare habent." In view of our sub- ject, it is of especial interest to us to note that this truth was generally recognized in Europe as late as the latter part of the eighteenth century, while Chinese authors were well informed on the subject from the beginning of our era. 2 It has recently been asserted (compare the notice of S. REINACH, Rebue archio- logique, 1913, p. 105) that the rhinoceros on a marble relief of Pompeii (Fig. 3; repro- duced also by REINACH, Repertoire de reliefs, Vol. Ill, p. 93; and O. KELLER, Die antike Tierwelt, Vol. I, p. 388) is an exact copy of the wood-engraving by Durer and accordingly the work of a forger. This point of view seems to me inadmissible, and I concur with Reinach in the view that a common antique model may have been handed down by the illustrators of the bestiaries. The most striking coincidence between the rhinoceros of Pompeii and that of Durer is the location of the second horn on the neck. This argument, however, is not cogent in establishing a close interdependence of the two; for also in China, on a picture of Yen Li-pen of the T'ang period (Fig. 1 1), the rhinoceros appears with a horn on its neck, and with scales on its body. As the artists all over the world were so much puzzled as to where to place the horn or horns, it is perfectly conceivable that Durer, solely guided by his reading of ancient writers, even without having recourse to an antique pictorial representation, worked out his 86 CHINESE CLAY FIGURES draughtsmen who had set before them the task of portraying a rhinoceros saw themselves in the same predicament as Diirer, in that they were lacking all personal experience of the animal, and for this reason were actuated by the same psychological factors. They, on their part, resorted to the classical definitions of the animal, as laid down in the ancient dictionaries Erh ya and Shuo w$n; they did not intend to picture a rhinoceros true to nature and directly from nature, simply because they were deprived of this opportunity, but they composed and pieced to- gether the creature from certain notions which they formed from bits of information gathered from their literary records. Whatever carica- tures their achievements may be, however, there cannot be the slightest doubt that they intended to represent a rhinoceros, not some other animal. Diirer's work, from a scientific viewpoint, is in details highly inaccurate and untrue; the modern naturalist may even pronounce the verdict that what he represented is far from resembling a rhinoceros at all ; but the bare fact remains and this is the essential point that the artist, as expressly stated in the legend by his own hand, had the intention of representing in this work a rhinoceros. As in most cases, the artist does not reproduce an object as it appears in the world of reality, but conveys to us his own notions of things as they are pro- jected in his mind. Exactly as it happened in China, so Durer's model found many adherents and followers, even among the naturalists who copied him again and again, and who surpassed him in fanciful additions of scales, wrinkles, and other decorations. Even BoNTius, 1 who pre- tends that he saw the animal in exotic forests and stables, and boasts of furnishing a figure of it free from Durer's defects, represents it, instead of with hoofs, with a paw very similar to that of a dog, only that it is somewhat larger. own theory in regard to the second horn. But it is desirable that, as suggested by Reinach, the iconographic question should be studied in detail. Neither should the differences between the two be overlooked. Durer's posterior horn is directly behind the ears; in the Pompeiian picture it is far behind the ears, above the front legs; in the same spot Durer has a small triangular point, the significance of which is not clear. It is certainly astonishing that the artists of Pompeii could commit this error, as the two-horned African rhinoceros was perfectly known in the Roman circus, and is correctly represented on the coins of Domitian mentioned above. ULYSSES ALDRO- VANDUS (Quadrupedum omnium bisulcorum historia, p. 354, Francofurti, 1647) has the figure of a rhinoceros, with an additional horn in the shape of a corkscrew placed on the shoulders. 1 JACOBI BONTII, Historiae naturalis et medicae Indiae Orie5ntalis libri sex, p. 51 (Amsterdam, 1658). The horn is correctly drawn. Bdntius avails himself of the word abada, which was used by old Spanish and Portuguese writers for a rhinoceros, and adopted by some of the older English narrators. The word is probably connected with Malayan badak, "rhinoceros" (see YULE and BURNELL, Hobson-Jobson, p. i). In G. DE MENDOZA (Dell 1 historia del gran regno della China, 1586, p. 437) the word abada is identified with the rhinoceros. HISTORY OF THE RHINOCEROS Archaeologists are agreed that the rhinoceros (Fig. 4) l is represented on the black obelisk of Salmanassar (B.C. 860824) in company with an elephant, human-looking apes, and long-tailed monkeys. This tribute- picture suggests to I. KENNEDY 2 the first certain evidence of Baby- lonian intercourse with India. The animals formed part of the tribute of the Muzri, an Armenian tribe living in the mountains to the north-east of Nineveh. 3 The rhinoceros is called in the inscrip- tion an "ox of the river Sakeya," and Kennedy criticises its repre- sentation as "very ugly and ill- drawn." Indeed, it is no more and no less than a bull, and, as far as natural truth is concerned, much in- ferior to the Chinese sketches. It even has cloven bull-feet, while one of the Chinese drawings has correctly three toes, 4 and the single , , . . , j Rhinoceros from Obelisk of Salmanassar II clumsy horn rises on its forehead (f rom o. Keiier. Antite 1 After O. KELLER, Die antike Tierwelt, Vol. I, p. 386 (Leipzig, 1909). 1 The Early Commerce of Babylon with India (Journal R. As. Soc., 1898, p. 259). 1 According to J. MARQUART (Untersuchungen zur Geschichte von Eran, II, p. 101, Leipzig, 1905), who discusses the same passage in the inscription of Salmanas- sar II, Muzri is the name of a country and mountain-range (Muzur Mountains) west of the Euphrates, and comprising also a part of the mountainous region south of the river. MARQUART translates "cattle of the river Irkea." Others, like Schrader, Hommel, and W. Max Muller (see B. MEISSNER, Assyrische Jagden, p. 20, Leipzig, 1911) identify Muzri with Egypt. KENNEDY does not explain how the rhinoceros could have gotten into that region from India; and it may have been, after all, an African species, although the single horn would rather point to India; the elephant, however, in his opinion, came over the passes of the Hindu Kush. There is, of course, the possibility that the lower Euphrates region may have harbored the rhinoceros, if we can depend upon the report of the Hou Han shu regarding the country of T'iao- chi (HIRTH, China and the Roman Orient, p. 38) ; and I am in full accord with what HIRTH remarks on this point hi the preface (pp. x-xn). However this may be, I agree with KENNEDY, F. HOMMEL (Die Namen der Saugetiere bei den sudsemiti- schen V6lkern, p. 324), MEISSNER, and KELLER that the animal figured on the black obelisk of Salmanassar is intended for a rhinoceros, and not merely for an ox, for there is no ox with single horn as here represented. The Assyrian name for the rhinoceros is kur-ki-za-an-nu = kurkizannu (F. DELITZSCH, Assyrische Tiernamen, p. 56, Leipzig, 1874), wn i c h' according to HOMMEL (/. c., p. 328), is a loan-word received from Ethiopic karkand (compare Arabic karkadan, Persian kerk). The trade-relations of India with Babylon are well established (see particularly G. BUHLER, Indian Studies III, p. 84). 4 The ancients did not notice this fact, nor did the Hindu, who classified the rhi- noceros, owing to a confusion with the elephant, among the five-toed animals (M. CHAKRAVARTI, Animals in the Inscriptions of Piyadasi, Memoirs As. Soc. Bengal, 88 CHINESE CLAY FIGURES between the eyes, as it occurs in the armorial unicorns. It is very instructive to compare this Babylonian representation with those of the Chinese; and whoever will view them together will certainly grant attenuating circumstances to the latter. The Babylonian production is the more surprising, as the supposition is granted that the live animal was sent as tribute; and the "artist," we should think, had occasion to actually see it. The outcome is such a caricature, however, that this point of view seems impossible; the "artist" simply acted on hearsay, or had been instructed to represent a queer foreign animal of the appearance of an ox, but with only a single horn on its forehead. And here we are again landing right at the threshold of the psychology of the Chinese draughts- man who, most assuredly, had never throughout his life viewed any living specimen of a rhinoceros, but merely reconstructed it in a vision of his mind from what he had heard or read. Nevertheless his product is not what it may seem to us on the surface, but it is and remains what it is intended for, the rhinoceros. Another instructive example for the iconography of the rhinoceros is furnished by Cosmas Indicopleustes, the Egyptian monk and traveller of the sixth century A.D. COSMAS * discriminates between the unicorn (monokeros) and the "nose-horn" (rhinokeros) , and has handed down to us sketches of both. In regard to the former, he remarks that he has not seen it, but that he had had occasion to notice four brazen figures of it set up in the four-towered palace of the King of Ethiopia, from which he was able to draw it. His figure 2 looks somewhat like a missing link between a horse and a giraffe, carrying on its head a straight, long horn. "In Ethiopia," Cosmas assures us, "I once saw a living rhi- noceros from a great distance and saw also the skin of a dead one stuffed with chaff, standing in the royal palace, and thus I was able to draw it accurately." The result of this "accurate" drawing is the figure of a maned horse with bushy tail, with two horns planted upright on its nose. 8 Nobody, as far as I know, has as yet inferred from this figure that the Greek word rhinokeros relates to an equine animal and should be translated by "horse." An interesting example of a Persian conception of the rhinoceros is depicted in the Burlington Magazine. 4 This is derived from an Vol. I, p. 371, Calcutta, 1906). In the commentary of Kuo P'o to the dictionary Erh ya (see below, p. 94) and in the Kiao chou ki of the fifth century A.D. it is clearly stated that the rhinoceros has three toes. Compare p. 95, note 6. 1 Ed. MIGNE (Patrologia, Vol. 88), p. 442. 1 Christian Topography, translated by MACCRINDLE, Plate IV, No. 28 (Hakluyt Society, 1897). * Ibid., No. 23. 4 Vol. XXIII, July, 1913, Plate III. HISTORY OF THE RHINOCEROS 89 illustrated "Description of Animals," the Manafi-i-heiwan, translated from Arabic into Persian and completed between 1295 and 1300. Here we have the interesting case that the author of this article, C. ANET, who evidently does not read Persian, mistakes the rhinoceros for "a horned gnu." But the picture is entitled in Persian kerkeden (or kargadan), "the rhinoceros," and it is therefore superfluous to dis- cuss the point that it cannot represent a gnu. 1 Although the creature has the shape of an ox, exactly as on the Assyrian obelisk and in the Chinese woodcut (Fig. 5), with the additional hump of a zebu 2 and black antelope-like stripes on its body, it is unmistakably characterized by a single horn in the form of a crescent. 3 In order to understand how the early Chinese illustrations of the rhinoceros alluded to by Mr. Giles were made, it is imperative to study the ancient definitions of the two words se and si. These definitions are sufficiently clear to place us on the right track in nicely dis- criminating between the two words, which plainly refer to two distinct species of rhinoceros. The weak point in Mr. Giles's definition of "bovine animal" 4 is that it is somewhat generalized, and leaves us entirely in the dark as to the difference between the two words se and si. They are physically differentiated words, and are expressed by different symbols in writing. Se-ma Ts'ien 5 mentions the two species of rhinoceros and elephant as inhabitants of the country of Shu (Sze-ch'uan). 6 The commentator 1 A species of antelope restricted to Africa, which could hardly be expected in Persian art. 1 This hints at the square-mouthed or white rhinoceros of Africa. One of the peculiarities of this species is the prominent, rounded, fleshy hump on the nape of the neck, just forward of the withers (E. HELLER, The White Rhinoceros, p. 20, Washington, 1913). * A representation of the rhinoceros in sculpture is spoken of in a Persian descrip- tion of the province of Pars from the beginning of the twelfth century; in Is^akhr the portrait-statue of King Jamshld was erected in stone, with his left hand grasping the neck of a lion, or else seizing a wild ass by the head, or again he is taking a unicorn (or rhinoceros) by the horn, while in his right hand he holds a hunting-knife, which he has plunged into the belly of the lion or unicorn (G. LE STRANGE, Journal R. As. Soc., 1912, p. 27). In the Annals of the T'ang Dynasty it is on record that in 746 A.D. Persia offered a rhinoceros and an elephant (CHAVANNES, T'oung Poo, 1904, p. 76). 4 What wild bovine animal should be understood has never been indicated. 6 Shi ki, Ch. 117, pp. 3 b, 7 b. 8 Our historians of Japan have been greatly puzzled by the fact that the Japanese Buddhist monk Tiao-jan (Japanese ChOnen), who came to China in 984, stated in his report embodied in Sung shi (Ch. 494, p. 4 b) that there were in his native country water-buffalo, donkeys, sheep, and plenty of thus it has been translated rhinoceroses and elephants (for example, by P. A. TSCHEPE, Japans Beziehungen zu China, p. 89, Yen-chou fu, 1907). O. NACHOD (Geschichte von Japan, Vol. I, p. 22) went so far as to appeal to a misunderstanding on the part of the Japanese informant, which he believes cannot be surprising, as Tiao-jan, though well versed in the written characters of the Chinese, did not understand their spoken language. This argu- go CHINESE CLAY FIGURES states, "The animal se is built like the water-buffalo. The elephant is a large animal with long trunk and tusks ten feet long; it is popularly styled 'river ape' (kiang yuan, No. 13,741). The animal si has a head resembling that of the ape yiian and a single horn on its forehead." 1 mentation is entirely inadmissible. It is certain that neither rhinoceros nor elephant exists in Japan: consequently Tiao-jan, in using the expression si siang (Japanese sai-zo) cannot be understood to convey to it its literal meaning, but he is sure to em- ploy it in a different sense. Chinese expressions (and Japanese are largely based on them) do not always mean what they seem to imply on the surface, but are often literary allusions or reminiscences of a metaphorical significance. The Japanese monk indeed avails himself of a Chinese phrase of classical origin traceable to Meng-tse (LEGGE, Classics, Vol. II, p. 281), and in my opinion, simply means to say that Japan produces "extraordinary wild animals." Yen Shi-ku, defining the word shou ("wild animals") in the Annals of the Han (Ts'ien Han shu, Ch. 28 A, p. 4 b), explains it as embracing such kinds as rhinoceros and elephants, whence it follows that this com- pound si siang is capable of rendering the general notion of wild animals. Si siang has thus become a stereotyped term occurring in many authors, although the literal meaning usually remains, as, for example, in Ts'ien Han shu (Ch. 28 B, p. 17), Erh ya (see p. 94, note 3), Nan shi (Ch. 78, p. 7), T'ang shu (Chs. 43 A, p. i, and 221 A, p. 10 b), and in the History of Shu (Shu kien) written by Kuo Yun-t'aoin 1236 (Ch. 10, p. I, ed. of Shou shan ko ts'ung shu, Vol. 23). HIRTH and ROCKHILL (Chau Ju-kua, p. 1 74) have taken a different view of the matter and suppose that the document utilized in the Sung Annals, and partially copied by Chao Ju-kua (inclusive of the statement that Japan produces si siang), contained a number of clerical errors; they are convinced that Tiao-jan's statement really was to the effect that there are neither rhinoceroses nor elephants in Japan. There is certainly no direct objection to be raised to such a point of view, but I am inclined to believe that with the indication as given there is no necessity of resorting to such a conjecture. 1 This universal notion could have emanated only from the two-horned species with reference to the rear horn, which anatomically is indeed placed over the frontal bone, while the front horn is situated over the conjoined nasal bones (FLOWER and LYDEKKER, Introduction to the Study of Mammals, p. 403). The posterior horn immediately follows the anterior one, and is somewhat beneath the eyes. Curiously enough, this idea of the position of the horn on the forehead was transferred also to the single-horned species, and became a well-established tradition, which one author copied from another. It is found in the classical world as well as among the Arabic authors. CTESIAS (ed. BAEHR, p. 254) seems to be the most ancient writer in whom this tradition has crystallized: he describes the wild white asses of India as "having on the forehead a horn a cubit and a half in length." The fact that he speaks of the rhinoceros, above all, is evidenced by his reference to the horn being made into drinking-cups which were a preventive of poisoning (compare also LASSEN, Indische Altertumskunde, Vol. II, p. 646). The monoceros of India, in the description of PLINY (Nat. hist., vui, 21), had a single black horn projecting from its forehead, two cubits in length (uno cornu nigro media fronte cubitorum duum eminente) . The horn of the rhinoceros sculptured in Assyria, as we have seen, is planted on its fore- head. Of course, when describing a rhinoceros which he saw at the games in the cir- cus, PLINY (vm, 20) states correctly that it has a single horn on its nose (unius in nare cornus) ; so does AELIAN (xvn, 44), and so does likewise Kuo P'o. The Arabic merchant Soleiman, writing in 851 (M. REINAUD, Relation des voyages faits par les Arabes, Vol. I, p. 28), attributes to the rhinoceros of India a single horn in the middle of its forehead, and is duly seconded by his copyist Mas'adi (RusKA, Der Islam, Vol. IV, p. 164). Ibn al-Faqlh, describing the two-horned species of Africa, states that it has on its forehead a horn, by means of which it inflicts mortal wounds; and another minor one is beneath the former and placed between its eyes (E. WIEDE- MANN, Zur Mineralogie im Islam, p. 250). Even al-Berunl (E. SACHAU, Alberuni's India, Vol. I, p. 204), who imparts a sensible account of the Indian rhinoceros, asserts from hearsay that the African species has a conical horn on the skull, and a second and longer horn on the front. Early European observers also believed that the horn of the rhinoceros was growing on its forehead. BARKER, as quoted by YULE HISTORY OF THE RHINOCEROS 91 In the other passage, the definition of Kuo P'o (276-324), the editor of the dictionary Erh ya, is quoted. The following definitions of the words se and si are given in the an- cient dictionary Shuo wBn (about 100 A.D.), and are here reproduced from an edition of this work printed in 1598, which is an exact facsimile reproduction of the Sung edition of the year 986. In all probability, this one faithfully mirrors the text of the original issue. The definition of se consists of only five words: " It is like a wild ox and dark-colored." 1 The character is then explained as a pictorial symbol (compare the re- production of the Chinese text on p. 92). It is doubtless on this enigmatic and incomplete definition that the explanations of PALLADIUS and COUVREUR (above, p. 74) are based. In order to reach a satisfactory result, however, it is always necessary to consult all records relating to a case; and it will always be unsafe to rely upon a single statement, which, after all, may have been curtailed, or in- correctly handed down. Let us note at the outset that the Shuo w$n by no means says that the animal in question is a wild ox, but only that it is like one; a comparison with a wild ox is not yet proof of identity with it. King Ping (9321010), the commentator of Shuo wln y annotates on the above passage as follows, " Its skin is so strong and thick that armor can be made from it," and quotes the Kiao chou ki * to the effect that " the horn is over three feet long and shaped like the handle of a horse-whip." * The fact that this author means to speak of a single horn becomes evident from the statement of Kuo P'o to be cited presently. 4 The and BURNELL (Hobson-Jobson, p. i),wrote in 1592, "Now this Abath [abada, bada = rhinoceros] is a beast that hath one home only in her forehead, and is thought to be the female Unicorne, and is highly esteemed of all the Moores in those parts as a most soveraigne remedie against poyson." 1 K'ang-hi's Dictionary quotes the Shuo win as saying that "the animal se has the shape or body of a wild ox and is dark-colored." * Records of Annam, of the fourth or fifth century, by Liu Hin-k'i (BRETSCHNEI- DER, Bot. Sin., pt. I, p. 159). 1 In a somewhat different way, the Shuo win is cited in Yen kien lei han (Ch. 430, p. 1 6 b), where original text and commentary are blended together: "The animal se resembles a wild ox and has a dark-colored skin which is so strong and thick that it can be worked up into armor. Among the animals on the mountain Po-chung, there is a large number of se." The latter name, according to PALLADIUS, is an ancient designation for a mountain in the west of Shan-si. The fact that the rhinoceros should have occurred there in ancient times is not at all surprising (see the notes below on the distribution of the animal in ancient times). It is noteworthy that we meet here the reading, "it resembles a wild ox," in agreement with the wording of the Erh ya, whence it follows that the se was not straightway looked upon as a wild ox, but as something else; it was merely likened to it a phraseology which is echoed in Baby- lonia and in the classical authors. This simile seems to account for the erroneous at- tempt of later commentators, like Chu Hi, to interpret se as identical with a wild ox. 4 The Kiao chou ki is credited in the Yen kien lei han with the words, "The se has a single horn which is over two feet long and shaped like the handle of a horse- whip." CHINESE CLAY FIGURES J " p HISTORY OF THE RHINOCEROS 93 animal si is defined in the Shuo wn as "an ox occurring beyond the southern frontier. It has a horn on its nose and another one on the crown of its head; it resembles a pig." 1 This definition fits no other animal than the two-horned species of rhinoceros, and has great his- torical value as a piece of evidence in determining the former geograph- ical distribution of the species. The passage shows us that in the first century A.D. it no longer existed in northern China, where its habitat had been prior to that time, and that it was then driven back beyond the southern border, speaking roughly, south of the Yangtse. It was then naturalized in Yun-nan, in the country of the Ai-lao, 2 and in Tonking. 8 To the author of Kiao chou ki we owe the following interesting de- scription of the Annamese rhinoceros: 4 "The rhinoceros (si) has its habitat in the district of Kiu-td (in Tonking). It has hair like swine, three toes, and a head like a horse. It is provided with two horns, the horn on the nose being long, the horn on the forehead short." It is clearly manifest that this description comes from an eye-witness, or one well informed by the native hunters, and that it perfectly fits the two-horned so-called Sumatran rhinoceros (Rhinoceros sumatrensis) , the only living Asiatic species with two horns, and also the most hairy one. 5 Its essential characteristics are well observed and briefly set forth in this definition. The dictionary Erh ya, edited by Kuo P'o (276-3 24) , defines the animal se as resembling the ox, and the animal si as resembling swine. The commentary by Kuo P'o explains that the se has a single horn, is dark in color, and weighs a thousand catties; 6 and "the si resembles in form 1 MARCO POLO (edition of YULE and CORDIER, Vol. II, p. 285) says regarding the rhinoceros of Java that its head resembles that of a boar. * Hou Han shu, Ch. 116, p. 8 b. 1 The question of the former geographical distribution of the rhinoceros in China is studied in detail below, pp. 159-166. 4 Yen kien lei han, Ch. 340, p. I. In Annamese the rhinoceros is called hui (written with the Chinese character for se) and tdy or t& (written with the character for si). 8 Hair grows sparsely all over the head and body, but attains its maximum de- velopment on the ears and the tail, its color varying from brown to black. The long- est known specimen of the front horn is in the British Museum, and has a length of 32 > inches, with a basal girth of ij^i inches; a second specimen in the same collec- tion measures 27^5 inches in length, and 17^ in circumference (R. LYDEKKER, The Game Animals of India, p. 38). The statement of the Kiao chou ki that the horn is two or three feet long is therefore no exaggeration. Concerning the two horns in the si, there is consensus of opinion between that work and the Shuo wen. * This may not be an exaggeration, though merely based on a rough estimate. The average weight of the rhinoceros, for reasons easy to comprehend, has never been ascertained. But if the weight of the skin alone may come to three hundred pounds (E. HELLER, The White Rhinoceros, p. 10), the complete animal may easily total a thousand and more. K'ang-hi and the modern editions of the Erh ya write ' ' thousand 94 CHINESE CLAY FIGURES the water-buffalo, 1 but has the head of a pig, a big paunch, short legs, and three toes on its feet; it is black in color and has three horns, one on the head, another on the forehead, and the third on the nose. The horn on the nose is the one by means of which it feeds [that is, uproots shrubs and trees]; 2 it is small and not long; it likes to eat thorny brambles; there is also a kind with but a single horn." Kuo P'o, accordingly, is fully acquainted with the single-horned rhinoceros (his three-horned species is discussed farther on) , and renders it plain enough that in his opinion neither the se nor the si is a bovine animal, as he treats them in a different section; while in his section on bovines, with twelve illustrations of such, no hint is made at se or si? The last doubt which might still exist as to the acquaintance with the single-horned rhinoceros on the part of Kuo P'o and Hu She*n, the author of Shuo w2n, will be banished by another word, tuan* (or kio tuan), of which Shuo wtn (Ch. n, p. 2) says that it is an animal of the shape of swine, with a horn which is good for making bows, and which is produced in the country Hu-siu. B catties." Yen kien lei Han (I. c.) has the erroneous reading "ten," which is impossible. Also Chang Yu-si, the author of the Pu chu ptn ts'ao of the year 1057, as may be seen from the Cheng lei ptn ts'ao, quotes the Erh ya as saying that "the se resembles an ox and has a single horn." Kuo P'o, accordingly, concurs with Liu Hin-k'i in the view that se is the single-horned rhinoceros. 1 Yen kien lei han (Ch. 430, p. i) offers the variant, "The si resembles swine, but is in shape like an ox;" then the same text as above is given, but the clause in regard to the three horns is wanting. 2 While feeding, the point of the horn of the animal may come in contact with the ground, so that the point is sometimes worn flat on its outer face (E. HELLER, The White Rhinoceros, p. 31). According to Ibn al-Faqlh, the African rhinoceros tears herbage out with the anterior horn, and kills the lion with the posterior one (E. WIEDEMANN, Zur Mineralogie im Islam, p. 250). s The rhinoceros is incidentally mentioned in another passage of Erh ya (Ch. B, fol. 29), where nine mountains with their famed productions are enumerated: "The finest productions of the southern region are the rhinoceros (si) and elephant of Mount Liang" (Liang shan, in Chung chou, Sze-ch'uan; PLAYFAIR, 2d ed., No. 3790, 2; BRETSCHNEIDER, Bot. Sin., pt. 3, p. 575, No. 187). Kuo P'o adds, "The rhinoceros furnishes hide and horn, the elephant ivory and bones." It follows therefrom, as is also confirmed by other sources, that in the third century A. D., the lifetime of Kuo P'o, the rhinoceros still existed in Sze-ch'uan, as seen above; its existence was attested there by Se-ma Ts'ien several centuries earlier. 4 Composed of the classifier kio ('horn') and the phonetic element tuan (No. 12,136). Not in GILES; see PALLADIUS, Vol. I, p. 189. A unicorn is represented on the Han bas-reliefs (CHAVANNES, Mission archlologique, Vol. I, p. 60, Paris, 1913). 6 Nos. 4930 and 4651. Other editions write Hu-lin. A horn bow is not a bow exclusively made from horn, which is technically impossible; but horn is only one of the substances entering into its manufacture. Technically the Chinese bow belongs to the class of composite bows, the production of which is a complicated process and requires a large amount of toil and dexterity. The foundation of the bow is formed of flexible wood connected with a bamboo staff. Along the back a thick layer of carefully soaked and prepared animal sinew is pressed, which, after drying, stiffens into a hard elastic substance. The inner side of the bow is then covered with two long horn sticks joining each other in the centre. The opposite of the horn bow is the wooden (or simple) bow (mu kung), as it is mentioned, for instance, as being used by HISTORY OF THE RHINOCEROS 95 Kuo P'o states in regard to the same animal, "The horn is on the nose and capable of being made into bows. Li Ling presented ten such bows to Su Wu. 1 The animal mentioned in the Life of Se-ma Siang-ju in the Shi ki (Ch. 117) is the k'i-lin 2 kio tuan." The animal with a horn on its nose is the single-horned rhinoceros; and the term tuan or kio tuan is a counterpart of the word monoceros of the ancients, as alluded to by Ctesias, Aristotle, Pliny, Aelian, and others, and which, according to the general consensus of opinion, relates to the one-horned rhinoceros of India. Bows manufactured from the horn are mentioned also in the Annals of the Kin Dynasty. 8 The allusion to armor by King Ping is additional proof for se being a rhinoceros, for, as we shall see, armor was not made in ancient China from the hides of bovine animals. 4 It is beyond any doubt that in those various definitions there is plainly the question of a rhinoceros. We cannot get over the single horn, whether placed on the nose, the head, or the forehead; 6 we can- not get over the fact, either, that a conspicuous distinction between the single-horned (se) and two-horned (si) species is made, a fact which will be discussed in full farther on when we have learned everything that Chinese authors have to report anent the two animals; nor can we get over the three toes which form a prominent characteristic of the rhi- noceros, 6 but assuredly not of any bovine species. In fact, the Chinese definitions, without pretension to scientific accuracy, which could not be the populace of Tonking (Ts'ien Han shu, Ch. 28 B, p. 17), which in connection with it availed itself of flint, bamboo, and sometimes bone arrowheads. 1 See GILES, Biographical Dictionary, pp. 450, 684. 1 Regarding the k'i-lin see below, p. 113. 3 Kin shi, Ch. 120, p. 3. Fossil rhinoceros-horn (from Rhinoceros tichorrhinus) is still employed by the Yakut in the manufacture of bows (B. ADLER, Int. Archiv fur Ethnographic, Vol. XIV, 1901, p. il). 4 Regarding other Chinese notions of monoceroses see p. 1 14. Of later descriptions of the rhinoceros, the one contained in Ying yai sh&ng Ian of 1416 by Ma Kuan is the most interesting. It is the most concise and correct definition ever given of the animal outside of our modern zoology. "The products of Champa are rhinoceros- horn and ivory of which there is a large quantity. The rhinoceros is like the water- buffalo. Animals of full growth weigh eight hundred catties. The body is hairless, black in color, and covered by a thick skin in the manner of a scale armor. The hoofs are provided with three toes. A single horn is placed on the extremity of the nose, the longest reaching almost fifteen inches. It subsists only on brambles, tree leaves and branches, and dried wood." ' As already remarked by CUVIER, the only real animal with a single horn is the rhinoceros. This statement reflects much credit on the observational power of the Chinese, especially as it is not pointed out by any classical author in describing the rhinoceros or unicorn. Al-Berttnl (SACHAU, Alberuni's India, Vol. I, p. 203) is the only early author outside of China to make the same observation. Al-BerunI gives two different and contradictory descriptions of the rhinoceros, apparently emanating from two different sources. First, the animal is sensibly described from personal observation 96 CHINESE CLAY FIGURES expected, are perfectly sound and to the point in stating what a primitive observer could testify in regard to an animal so difficult of access and so difficult to describe. Surely, the Chinese definitions are not worse, and in several points perhaps better, than anything said about the animal in classical antiquity, among the Arabs, or in Europe up to the eighteenth century. And we shall soon recognize that until the very recent dawn of our scientific era the Chinese were the nation of the world which was best informed on the subject. 1 The Chinese likened the rhinoceros to the ox, the water-buffalo, the pig, 2 and its head to that of an ape. as follows: "The gan<}a exists in large numbers in India, more particularly about the Ganges. It is of the build of the buffalo [analogous to the Chinese definition], has a black scaly skin, and dewlaps hanging down under the chin. It has three yellow hoofs on each foot, the biggest one forward, the others on both sides. The tail is not long ; the eyes lie low, farther down the cheek than is the case with all other animals. On the top of the nose there is a single horn which is bent upwards. The Brahmins have the privilege of eating the flesh of the ganda. I have myself witnessed how an elephant coming across a young ganda was attacked by it. The ganda wounded with its horn a forefoot of the elephant, and threw it down on its face." The other account of al-Berunl, which refers to the double-horned African species, is composed of the narrative of a man who had visited Sufala in Africa, and of classical reminiscences freely intermingled with it; to the latter belong the beliefs in the mobility of the horn and in the sharpening of the horn against rocks, and here appears also the wrong notion that it has hoofs. PLINY (Nat. hist., vm, 21, 76) asserts that the single- horned oxen of India have solid hoofs (in India et boves solidis ungulis unicornes), a tradition which savors of the description of a unicorn after a sculpture (on the As- syrian obelisk the animal has bovine hoofs). Even ARISTOTLE (Hist, an., n, 18; ed. of AUBERT and WIMMER, Vol. I, pp. 74, 254), who evidently speaks after Ctesias, characterizes the single-horned "Indian ass" as solid-hoofed (/xcbpuxa). This lacune in the descriptions of the ancients was aptly pointed out by BELIN DE BALLU (La chasse, poeme d'Oppien, p. 174, Strasbourg, 1787), who, in speaking of the familiarity of the ancients with the animal, concludes by saying, "Mais ce qui doit nous 6tonner c'est qu' aucun n'ait parle" d'un caractere particulier de cet animal, dont les pieds sont partage"s en trois parties, revenue chacune d'une sole semblable a celle du bceuf." 1 The only reproach that can be made to the Chinese authors is that they never point to the peculiar skin-folds of the animal (with the only exception, perhaps, of Fan Ch6n of the Sung period, who describes the rhinoceros of Annam as "clad with a fleshy armor;" see p. 113), and that, despite the live specimens procured for the Imperial Court (p. 80), no attempt has ever been made at a more precise description based on actual observation. But we may address the same charge of omission to the authors of India, the Greek writers on India, and to Pliny and Aelian. PLINY is content with stating that he saw the animal in the Roman circus, but does not de- scribe what he saw, while he is eager to reproduce all the fables regarding the monoce- ros, emanating from India or from former sources relative to India. AELIAN (Nat. an., xvn, 44) thinks it superfluous to describe the form of the rhinoceros, since a great many Greeks and Romans have seen and clearly know it. In matters of descrip- tion the animal presents as difficult a subject as in matters of art. Exact descriptions of it are due only to competent zoologists of recent times. 2 How very natural this comparison is, maybe gleaned from the account contained in Nan Ytie chi (quoted in T'u shu tsi ch'Bng, chapter on rhinoceros), that at the time of the Han a rhinoceros once stampeded from Kiao chi (Annam) into Kao-liang (the ancient name for Kao-chou fu in Kuang-tung Province), and that it was mistaken by the people for a black ox, while those acquainted with the animal asserted that it was a black rhinoceros. The resemblance of the rhinoceros to an ox or buffalo has indeed obtruded itself on the observers of all times; and this notion is so far from being restricted to the Chinese, that it may almost be called universal. As seen above (p. 87), the Assyrians called the animal "ox of the river Sakeya." PLINY (Nat. hist., HISTORY OF THE RHINOCEROS 97 This is all exceedingly good: it is simply the result of that mental process which classifies a novel experience under a well-known category, vin, 21, 72, 76) speaks of the unicorn oxen of India. FESTUS calls the African rhinoceros the Egyptian ox, and PAUSANIAS tells of "Ethiopic bulls styled rhino- ceroses" which he saw himself in Rome (O. KELLER, Die antike Tierwelt, Vol. I, P- 385). The Indian physician Caraka, who lived at the Court of King Kanishka in Kashmir, placed the rhinoceros in the class of buffalo (anupa, Mem. As. Soc. Bengal, Vol. I, 1906, p. 371). The Arabic merchant Soleiman, who wrote in 851, compared the Indian rhinoceros with the buffalo (M. REINAUD, Relation des voyages, Vol. I, p. 29) ; and so did, as seen above, al-Berdnl. Ibn al-Faqlh says regarding the African rhinoceros that it resembles a calf (E. WIEDEMANN, Zur Mineralogie im Islam, p. 250). The Talmud, in three passages, mentions the one-horned ox as an animal sacri- ficed by Adam (L. LEWYSOHN, Die Zoplogie des Talmuds, p. 151, Frankfurt, 1858). The "sea-ox" mentioned by Leo Africanus (HiRTH and ROCKHILL, Chau Ju-kua, p. 145) certainly is the rhinoceros. The Malays designate the two-horned species badak-karbau, "the buffalo-rhinoceros," and the single-horned species badak-gajah, "the elephant-rhinoceros." It is difficult to understand, however, why some of the classical authors allude to the rhinoceros under the designation "the Indian ass" (ARISTOTLE, Hist, an., n, 18, ed. of AUBERT and WIMMER, Vol. I, pp. 74, 254). Aristotle's definition is traceable to CTESIAS (ed. BAEHR, p. 254), who states that there were in India wild white asses celebrated for their swiftness of foot, having on the forehead a horn a cubit and a half in length, and that they are colored white, red, and black ; from the horn were made drinking-cups which were a preventive of poisoning (compare also LASSEN, Indische Altertumskunde, Vol. II, p. 646). The mention of these antipoisonous cups is good evidence for the fact that Ctesias hints at the Indian rhinoceros (HERODOTUS, iv, 191, speaks of horned asses of Libya, but they are not one-horned). Ctesias is an author difficult to judge. His account of India, said to have been written in B.C. 389, it should be borne in mind, was de- rived second-hand, while he resided in Persia as court-physician of King Artaxerxes Mnemon, so that his data may partially be based on Persian accounts of India, and misunderstandings of his informants may have crept in; moreover, his report is handed down in a bad and fragmentary condition, and may have been disfigured by Photias of Byzance of the ninth century, to whom the preservation of his work is due. The definition of Ctesias in the present case cannot be regarded as correct, as we do not find in India, or anywhere else in the East, a comparison of the rhinoceros with an ass, nor any tradition to this effect, a tradition which is not likely ever to have existed. If the ass really was contained in his original text, it must go back, in my estimation, to a misunderstanding on his part of the word imparted to him by the authorities whom he questioned. With the exception of the horn, Ctesias does not seem to have entertained any clear notion of the animal; and his description of the skin as white, red, and black, is baffling. V. BALL (Proceedings Royal Irish Academy, Vol. II, 1885, and in his edition of Tavernier's Travels in India, Vol. I, p. 114) tried to show that the colors seen by Ctesias were artificial pigments applied to the hide, as they are on elephants at the present day; rhinoceroses kept by the Rajas for fighting-purposes were, according to him, commonly painted with diverse bright colors. This forced explanation, shifting quite recent affairs to the days of early antiquity, is hardly plausible. It seems to me that we are bound to assume that the text of this passage is not correctly handed down. The colors white, red, and black would seem rather to have originally adhered to the horn. The Eastern lore of the rhinoceros, as shown by the reports of the Chinese and Arabs, essentially clusters around the horn. MARCO POLO (ed. of YULE and CORDIER, Vol. II, p. 285) says in regard to the Javanese rhinoceros that its head resembles that of a wild boar; and this characteriza- tion is quite to the point, as is that of Kuo P'o when he compares the two-horned si to swine. A glance at Fig. 8, representing the specimen of a Sumatran two-horned rhinoceros in the Field Museum, will convince every one of the appropriateness of this simile. The pig shape of the rhinoceros is apparent also in a Roman representa- tion on a clay lamp from Labicum illustrating the struggle between that animal and a bear (Fig. 7), so that even the most skeptic critic of Chinese animal sketches will be compelled to grant a certain foundation of fact to the hog-like rhinoceros of the Erh ya (Fig. 6). 9 8 CHINESE CLAY FIGURES uld,n&t be anyoetffcf . We should halt a moment to reflectf'by what class of people these observations had been made. Most oJtainly''by tJi* MfSjTm^.. hardy hugrtprs who !*hgs^J the wild b&sfcy We must AoA^r >v ^ , distinguish ^tjetfjeen tne original teller, and the closeted in his study who draughted the definitions for the consumption of the learned. It was not the Chinese philologist who went out into the jungle to study the rhinoceros; he, indeed, never had occasion to see it, but he derived his knowledge from reports made to him by the sportsman. The latter probably was plain and matter-of-fact; the FIG. 7. Struggle of Bear and Rhinoceros, represented on a Clay Lamp from Labicum (after 0. Keller, Tiere des classischen Altertums). FIG. 8. Sumatran Rhinoceros. Sketch from Museum Specimen (compare Elliot. Catalogue of the Collection of Mammals. Zool. Series. Vol. VIII, p. 105). former added a bit of romance and exaggeration. Have we any right to ridicule the Chinese over their embarrassment as to where to locate the horn or the horns, when we observe that this was still a matter of wild speculation amidst Europe in the seventeenth and eighteenth centuries? 1 1 Dr. PARSONS, in the pamphlet quoted, justly remarks, "Nothing could serve as a better proof of how easily men may fall into uncertainty through preconceived conclusions than this very topic of the horn of the rhinoceros." ioo CHINESE CLAY FIGURES Have we any right to look down upon their artists in their naive at- tempts to sketch the rhinoceros in the shape of an ox with a horn on the forehead (Fig. 5), when we observe that the so-called "civilization" of Assyria and the painting of Persia committed the same error, or when we glance at the puerile drawings of Cosmas and recall Durer's work with the horn on the animal's neck? In the above definitions we recognize the elements and tools with which the subsequent Chinese illustrators worked. They set out to il- lustrate, not the rhinoceros, but the descriptions given of it in the ancient dictionaries. They studied, not the animal, but the ready- made definitions of it encountered in book-knowledge. They read, and their reading guided the strokes of their brush. "The se resembles in body a water-buffalo, the si a pig:" consequently such bodies were outlined by the illustrator of Erh ya; and long, curved, and pointed single horns were placed on the heads (Figs. 5 and 6). 1 He apparently shunned the three horns, as the matter was difficult to draw; and no- body knew how to arrange them. He carefully outlined the three toes 1 Our illustrations are derived from a folio edition of the Erh ya printed in 1801 (3 vols.), which is designated as "a reproduction of the illustrated Erh ya of the Sung period" (Ying Sung ch'ao hut t'u Erh yd). The ancient illustrations of the Erh ya by Kuo P'o and Kiang Kuan are lost (see BRETSCHNEIDER, Bot. Sin., pt. i, p. 34), and were renewed in the age of the Sung, presumably without any tradition connect- ing the latter with the former. This fact may account for the purely reconstructive work of some illustrations, and we may well assume that the earlier sketches were far better. Many other illustrations of the Erh ya have been brought about in the same manner as those of the rhinoceros. Compare, for instance, the picture of the fabulous horse po (No. 9393) surrounded by flamed fluttering bands and about to lacerate a tiger seized by its carnivora-like, sharp claws; while a panther is swiftly making for safety to escape a similar fate. Of course, the craftsman has never observed this scene, but faithfully depicts the definition of the book, "The animal po is like a horse with powerful teeth, devouring tigers and panthers." This notion, as indicated by Kuo P'o, goes back to the Shan hat king, which says, "There is a wild animal styled po, like a white horse with black tail and powerful teeth, emitting sounds like a drum and devouring tigers and panthers." (Here we have a parallel to, and pre- sumably an echo of, the flesh-eating horses of Diomed and the man-devouring Bucephalus of the Alexander legend; see J. v. NEGELEIN, Das Pferd im arischen Al- tertum, pp. 43, 75, Konigsberg, 1903.) Otherwise the horses pictured in the Erh ya, aside from their technical drawbacks, are quite realistic; and so are the oxen and other animals which came under the every-day observation of the Chinese. It is still a mystery, and a problem worth while investigating, why the Chinese were rather good at drawing some animals and completely failed in others. It may be pointed out that the tapir of the Erh ya, aside from the exaggerated trunk and wrong tail, is rather correctly outlined with its white saddle, and corresponds to a well-known species (Tapirus indicus). In view of the retrospective and reconstructive sketches of this work, we have the same state of affairs as in the illustrations accompanying the Shan hat king, and as formerly shown by me in Jade, in the San li t'u, and to a certain extent in the Ku yu t'u p'u. The illustrators of the ancient Rituals did not directly picture the actual, ancient ceremonial objects, most of which were lost past hope in their time, but reconstructed them from the descriptions supplied by the commentators of the ancient texts, and for better or worse, based their illus- trations on these artificial reconstructions, which to a large extent are erroneous or imaginary. HISTORY OF THE RHINOCEROS 101 in the animal si; and this feature, combined with the single horns, is sufficient flavor of the rhinoceros to guard from any rash conclusion even one who has not considered the psychological foundation of these sketches. From the fact that the animal se is drawn in the shape of an ox, Mr. Giles infers that the word se does not denote the rhinoceros, but "a bovine animal." Then, how about the word si? The animal si (Fig. 6) is undeniably represented in the Erh ya t'u with the body of a hog, why not, to be consistent, also translate the word si by "swine"? If a child who was invited to make a sketch of a whale should delineate it in the shape of a fish, should we conclude for this reason that the whale is a fish? To make use of an illustration for a far-reaching philological and zoological conclusion, it is indispensable to ascertain the real value of such an illustration, and to make a somewhat critical study of its origin and basis. Mr. Giles is right in stating that there are illustra- tions of the animal se that are purely those of an ox. The ill-reputed San li t'u, for instance, stooped to this wisdom when the difficult task arose of illustrating in the shape of a rhinoceros the target used by the lords and ministers in the practice of archery, and spoken of in the Chou li and / /*. But what wonder ! Those illustrators who employed the pure-ox design simply stood on the platform of the sober and incom- plete definition of the Shuo wen, "The animal se is like a wild ox." Nothing could be more convenient to the unthinking and mechanical craftsman; this plain recipe freed him from the responsibility for the horn. Anybody could outline an ox with two regular horns; and by inscribing it se, the satisfaction at this achievement was naturally the greater. It is incorrect, however, to say that the animal se, as outlined in T'u shu tsi ch'eng (Fig. 9), is the picture of an ox. In its general features it resembles a kind of deer, as does likewise the animal si (Fig. 10). A lengthy discussion of the "deer-like" rhinoceros follows below (p. 109). Again, in Fig. 9, the draughtsman has taken particular pains to set off distinctly *hree toes in the left front foot; and where is the bovine animal with three toes? And where is the bovine animal with a single horn, and with this peculiar shape of horn? As to Fig. 10, it presents itself as an illustration of the legend that, while the rhinoceros is gazing at the moon, the peculiar designs within its horn are formed (p. 147). This notion exclusively refers to rhinoceros-horn, so that the animal here intended can be no other than the rhinoceros. 1 1 The two illustrations of T'u shu tsi ch'ing are derived, with a few slight altera- tions, from San ts'ai t'u hui (section on Animals, Ch. 3, p. 7; Ch. 4, p. 12), where, curi- 102 CHINESE CLAY FIGURES 23. FIG. 9. The Animal se (from T'u shit tsi ch'eng). HISTORY OF THE RHINOCEROS 103 PIG. 10. The Animal si gazing at the Moon (from 7"* MM tti ek'tog). 104 CHINESE CLAY FIGURES The three-horned rhinoceros described by Kuo P'o is perhaps not so fabulous as it may appear at first sight; for it is known to naturalists that the animal has also the tendency of developing three horns. E. HELLER 1 states in regard to the black rhinoceros covering the whole of Africa with the exception of the Congo Basin that, although the species is almost invariably two-horned, occasional variations of one and three- horned specimens are met with. In the light of this observation, PLINY'S (Nat. hist., VIII, 21) notice of oxen of India, some with one horn, and others with three (Indices boves unicornes tricornesque) , is apt to lose much of the legendary character with which it was formerly charged. As far as I know, a three-horned specimen has not yet been pointed out among the species of the Indo-Malayan region; notwith- standing, the possibility remains that such may have occurred in times of antiquity. However this may be, whether we assume that the notion of a three-horned species was founded on a natural observation or not, the fact of the coincidence between Kuo P'o and Pliny remains, and hints at the existence of a tradition anent a three-horned variety in the beginning of our era. 2 At any rate, whether real or imaginary, the latter is but a variation of the two-horned species; and by omitting Kuo P'o's illusory "horn on the head," we arrive at a fairly accurate description of it, and then Kuo P'o exactly agrees with Hii Shen's definition of the word si. And there can be no doubt of the point that ously enough, they are separated and dispersed in two different chapters. In the latter work, the horn of the se is decorated with different designs, which are white on black, while they are black on white in T'u shu. The si of San ts*ai is adorned with flamed and fluttering bands, and the crescent of the moon is absent. 1 The White Rhinoceros, p. 35 (Washington, 1913). Again on p. 17: "The num- ber of dermal horns on the snout is of less importance. These have been found to show some individual variation in the African species varying from one to three in number in the same species. The front horn, however, is nearly always the better developed and is never wanting." 2 The case could certainly be argued also from a purely philological point of view. Kuo P'o's creation might be explained as an ill-advised combination of the single- horned and two-horned species, or even regarded as a subsequent interpolation in his text, due to a scribe who meant to be sure of his definition being as complete as possible. Pliny's tricornis might be rationally interpreted as the result of an arithmet- ical process, providing the rhinoceros as a species of ox with two bovine horns, and adding the nose-horn as the third. In this manner Damlrl's three-horned rhinoceros must have arisen (RusKA, Der Islam, Vol. IV, 1913, p. 164), for it has one horn between the eyes and two above the ears. The natural explanation based on zoologi- cal observation appeals to me to a much higher degree, for we must not be forgetful of the fact that it is impossible for the human mind to invent spontaneously such an observation; a feature of this kind, in order to be observed by man, must have some- how pre-existed in nature. It means nothing, of course, to say that the three horns are a fable; if fable it is, then how did the fable come into existence? It is not the question of a mythological conception, or of a mythical monster, but plainly of a really existing animal described in sober words. I feel confident that the three-horned variation in a living or extinct species will be found some day also in Eastern Asia. HISTORY OF THE RHINOCEROS 105 what Kuo P'o intends to describe is the two-horned species of rhinoceros, not any other animal: his statement in regard to "the horn on the nose" excludes any other idea, and the bovine animal with such a horn remains as yet to be discovered. Li Shi-che'n of the sixteenth century, as will be seen below (p. 150), rejects the definition of Kuo P'o as erro- neous; that is to say, he did not know of any three-horned variety, and recognized in it the two-horned species. An illustration of this three- horned creature may be viewed in the Wa-Kan San-sai-zu-e , the Japa- nese edition of the Chinese cyclopaedia 5an ts'ai t'u hut. 1 The defini- tion runs thus: "The rhinoceros has the hair of swine and three toes on each foot; it has the head of a horse and three horns, on the nose, the forehead, and on the skull, respectively." The three toes and three horns are exactly drawn in accordance with this prescription; curiously enough, however, the head is not that of a horse, but of a bull. The old tradition of the draughtsmen is retained in spite of the definition. Kuo P'o, in all probability, is not the first or the only author to speak of a three-horned variety. A work Kiao Kuang chi, 2 Account of Kiao chou (northern part of what is now Annam) and Kuang-tung, reports, "In the territory of the Barbarians of the South-west occurs a strange rhinoceros with three horns emitting light at night like big torches at a distance of a thousand paces. When it sheds its horns, it hides them in a remote and dense jungle to prevent men from seeing them. The sovereigns hold this strange product in high esteem, and make it into hair-pins. These are capable of checking evil and rebel- lion." Here we have the testimony of an eye-witness or one reproducing a hearsay account; and, quite correctly, he points out this variety as a freak of nature. The exact date of the work in question is unfortunately not known to me; but as the quotation is placed between one from Kuang-chi by Ku Yi-kung, who according to BRETSCHNEIDER* belonged to the Liang dynasty (502 556), and one from Kuang chou ki, a work of the Tsin period (265419), the inference may be justifiable that Kiao Kuang chi likewise is a production of the Leu-ch'ao period. However remote from truth all these Chinese illustrations may be, most of them are fairly correct as to the outlines of the horn, naturally because 1 The illustration is easily accessible in L. SERRURIER, Encyclop&iie jappnaise, le chapitre des quadrupedes, Plate VIII (Leiden, 1875). This cut is not contained in a recent edition of this Japanese work (Tokyo, 1906), but is replaced by a rhinoceros with two horns, the one on the forehead, the other on top of the skull. These attempts clearly prove that Japanese as well as Chinese illustrators did not draw the animal from life, but from the definitions of the books. In the Chinese San ts'ai t'u hut (Ch. 4, p. 32) only a three-horned animal (san kio shou) is depicted. 1 Quoted in the chapter on Rhinoceros in T'u shu tsi ch'tng. 1 Bot. Sin., pt. I, p. 164. io6 CHINESE CLAY FIGURES the horn as an article of trade was always known, but not the animal itself. 1 The r61e played by the rhinoceros in Chinese art is limited. As shown by the symbol illustrated in the Po ku t'u lu (Fig. 18), it was pictured in early antiquity; and other representations of that period mentioned in Chinese records are discussed on p. 160. The animal lacks those aesthetic qualities of form which tempt the brush of the painter; and this may be the reason why despite the living rhinoceroses sent up as tribute to the capital (see p. 80) it has never been immortalized on any Chinese scroll known to us. 2 There is, however, one case on record. Chang Shi-nan, who wrote the book Yu hitan ki wtn early in the thir- teenth century, 3 narrates that he once saw in Sze-ch'uan (Shu) the painting of an unknown artist showing the outlines of a rhinoceros with a horn on its nose. 4 The inhabitants of Sze-ch'uan, accordingly, were familiar with the animal, and for this reason represented it correctly. On some Buddhist pictures it may owe its existence to a mere lucky chance; that is, to the fact that it was so copied from an Indian- Buddhist model. On Yen Li-pn's picture showing Samantabhadra's elephant, 8 the rhinoceros is unmistakably contrasted with the elephant as the smaller animal with scaly body, and head surmounted by a single horn. Another illustration of the same subject is reproduced in Fig. n from Ch'tng shi mo yuan (Ch. 6 B, p. 16) published in the Wan- li period, after 1605. Possibly it occurs also on the later typical paint- ings of Buddha's Nirvana in the group of wailing animals. 6 On the sculptures of Angkor- Vat the rhinoceros is represented as the vehicle of the god Karttikeya. 7 The Mongol emperors made practical use of the typical, conventional designs of the rhinoceros on the standards of the army: there was a standard with the picture of the animal se, "resembling an ox, with a single horn, and of dark color," and another with a picture of the 1 A modern Chinese school-book published at Shanghai in 1901, and illustrated by Wu Tse-ch'Sng of Su-chou, illustrates the word si with the cut of a rhinoceros of European origin, and the word se with a jovial ox of his own invention; while the text accompanying it, imbued with the spirit of the Shuo wSn and Erh ya, speaks of one horn on the nose and three toes. 1 It is likewise absent from classical Greek art. The marble relief of Pompeii, the lamp from Labicum, and the coins of Domitian referred to, are the only known ex- amples of its representation in late Roman art. 1 WYLIE, Notes, p. 165. 4 The text is reprinted in T'u shu tsi ch'ing, chapter on rhinoceros, hui k'ao, p. 5. 8 Reproduced in the writer's Jade, p. 342. See for example A. GRUNWEDEL, Buddhistische Kunst in Indien, p. 1 14, or Bud- dhist Art in India, p. 124 (in the right lower corner). 7 According to M. G. CoEofes, Les bas-reliefs d'Angkor-Vat, p. 12 (Paris, 1911). HISTORY OF THE RHINOCEROS 107 FIG. 11. 'Brushing the Elephant." Rhinoceros with Scaly Armor in Front. Wood-engraving from Ch'tng-shi mo yiian. io8 CHINESE CLAY FIGURES rhinoceros si niu, which is not described. They had also standards with designs of a three-horned animal (san kio shou) and the unicorn (kio tuan), which was outlined "like a sheep, with a small tail and a single horn on its crest." * In plastic art, 2 the rhinoceros has been carved from jade either as the handle of a paper-weight or as the knob of a seal. 3 An example of either kind is illustrated in Ku yti t'u p'u (Ch. 74, p. i, reproduced in FIG. 12. Ancient Paper- Weight of Jade surmounted by Figure of Rhinoceros (from Ku y& t'u p'n). Fig. 12; and Ch. 37, p. n). The traditional reconstructions of the animal are here faithfully preserved; the three toes (the third, of course, is not visible) and the shape of the horn, though it is wrongly placed, come somewhat near the truth. The manufacturers of ink-cakes availed themselves of the same design for printing on the surface of their products. The Ch'Gng shi mo yiian (Ch. 13, p. 30) illustrates "a spiritual rhinoceros" (ling si) with body of an ox, hump of a zebu, cloven feet, snout of a pig, and horn on the front. 1 Yiian shi, Ch. 79, p. 10 (K'ien-lung edition). 2 BUSHELL (Chinese Art, Vol. I, p. 91) figures a bronze vessel of the type styled hi ts'un, and describes it as being "shaped in the form of a rhinoceros standing with ears erect and a collar round the neck. But this explanation conflicts with Chinese tradition, according to which the animal hi is a sacrificial ox; and an ox is apparently represented in this bronze. Neither is there a single or double horn, which would be necessary to establish such a case. 8 Seals surmounted by the full figure of a rhinoceros seem to make their first appearance in the Han period (see Hou Han shu, Ch. 40, p. 5). HISTORY OF THE RHINOCEROS 109 The most curious item in the history of the iconography of the rhinoceros is the illustration of the animal in the Chdng lei p$n ts'ao published in 1208 by the physician T'ang Sh&i-wei 1 (reproduced in Fig. 13). Here we see the animal represented as a hairy and spotted deer, its head being surmounted by a single curved horn, peacefully chewing a bunch of leaves with a most innocent expression on its face. The legend is si kio ("rhinoceros-horn"), all illustrations of animals FIG. 13. Deer with Single Horn, labelled Rhinoceros-Horn, being an Echo of the Indian Legend of Ekagririga (from Chtng lei pin ts'ao. edition of 1523). in this work being named for the product yielded by them; and the il- lustration is immediately followed by the description of the two animals se and si, so that there can be no doubt that this figure, in the mind of the author, is intended for the rhinoceros. It will certainly not induce us to propose for the word si the new translation "cervine an- imal;" but a rhinoceros of cervine character has really existed in the imagination of the ancient world. The idea started from India, has taken a footing in the classical authors, and long survived even down to our middle ages. It is a fascinating story, deserving full discussion, the more so as it has never been clearly and correctly set forth. Two classical texts may first be quoted which fit well as an explanation to our Chinese woodcut. PLINY (Nat. hist., VIII, 21) tells regarding the Orsaean Indians that "they hunt the indomitable, fierce monoceros (unicorn) which has the head of a stag, the feet of the elephant, the 1 Regarding this work and its history see Toung Poo, 1913, p. 351. In the edi- tion of 1523 from which our illustration is taken it is in Ch. 17, fol. 20 b. no CHINESE CLAY FIGURES tail of a boar, while the rest of the body is like that of the horse; it emits a deep roar, and has on the middle of its forehead a single black horn two cubits in length. This beast, it is asserted, cannot be captured alive." 1 In the Cyranides, a curious Greek work written between 227 and 400 A.D., 2 it is said, "The rhinoceros is a quadruped resembling the stag, having a very large horn on its nose. It can be captured only by means of the perfume and the beauty of well dressed women; it is indeed much inclined toward love." 8 The importance of this passage, first of all, rests on the fact that the single-horned cervine animal is here clearly identified with the rhinoceros, an identification not yet made by Pliny, who speaks of rhinoceros and monoceros as two distinct species; and we remember that Cosmas Indicopleustes makes the same distinc- tion in regard to India. In his introduction, F. DE MELY* observes that the Cyranides is the first work to reveal to us the starting-point of the legend of the chase of the unicorn which is nothing but the rhino- ceros. This, however, is very inexact. The first Occidental source relating this legend is the Physiologus which is older than the Cyranides. The Physiologus 6 tells of the monoceros that it is a small animal re- sembling a buck, but very cunning; the hunter cannot approach it, as it possesses great strength; the horn grows in the centre of its head; it can be captured only by a pure virgin who suckles it; then she seizes it, and carries it into the palace of the king; or according to another version, the unicorn falls asleep while in the lap of the virgin, whereupon the hunters gradually approach and fetter it. The monoceros is located by PLINY in India; and the western legend of the unicorn ensnared by a virgin was first traced by S. BEAL B to the ancient Indian legend of Ekagringa, the hermit Single Horn. H. LUDERS, T who has traced with great ingenuity the development of the legend in the sources of Indian 1 Orsaei Indi . . . venantur asperrimam autem feram monocerotem, reliquo corpore equo similem, capite cervo, pedibus elephanto, cauda apro, mugitu gravi, uno cornu nigro media f rente cubitorum duum eminente. hanc feram vivam negant capi. (Ed. of C. MAYHOFF, Vol. II, p. 104.) 1 F. DE MfiLY, Les lapidaires grecs, p. LXXI; DE MELY is the first editor and translator of this work. 1 L. c., p. 90. 4 L. c., p. LXV. S F. LAUCHERT, Geschichte des Physiologus, pp. 22, 254 (Strassburg, 1889); F. HOMMEL, Die aethiopische Vbersetzung des Physiologus, p. 68 (Leipzig, 1877); E. PETERS, Der griechische Physiologus und seine orientalischen Ubersetzungen, p. 34 (Berlin, 1898); K. AHRENS, Das "Buch der Naturgegenstande," p. 43 (Kiel, 1892). 'The Romantic Legend of C^kyamuni Buddha, p. 125; see also his Buddhist Records of the Western World, Vol. I, p. 113. 7 Die Sage von Rsyasringa (Nachrichten d. k. Ges. d. Wiss. zu Gottingen, 1897, pp. 1-49), p. 29; an additional study from his pen on the same subject ibid., 1901, pp. 1-29. HISTORY OF THE RHINOCEROS in literature, justly points out that all our mediaeval versions of the story, 1 as a last resort, go back to the Greek Physiologus, and that the last clause of the Greek text contains a visible trace of the old Indian legend of the king's daughter who carries away the penitent into the palace of her father. Luders rises also against the view of Lauchert, who inter- prets the story in Physiologus from a misunderstood passage of AELIAN (XVI, 20) ; and I am in full accord with the criticism of Luders, to which the argument should be added that this alleged influence of Aelian on the Physiologus is out of the question, as Aelian is in time posterior to the latter. 2 F. W. K. MILLER studied the same question in connection with a Japanese No play, the plot of which is the legend of Ekagringa. 8 Muller likewise thinks Lauchert's explanation to be hardly plausible, and admits, with excellent arguments, the dependence of the Physiologus story on the tradition of India. There is but one point in which my opinion differs from the one expressed by Muller. Muller, at the close of his highly interesting study , advances the theory that the real unicorn, as already recognized by Marco Polo, may always have been the 1 Of the mediaeval versions, that of JOHN TZETZES, the Byzantine poet and gram- marian, who flourished during the twelfth century, in his Chiliades (v, 398), deserves special mention : " The monoceros carries a horn on the middle of its forehead. This animal is passionately fond of perfumes. It is hunted in this manner. A young man disguised as a woman exhaling the odor of the most exquisite perfumes takes his position in the places frequented by this quadruped. The hunters lie in ambush at a short distance. The odor of the perfumes soon attracts the monoceros toward the young man; it caresses him, and he covers its eyes with perfumed woman's gloves. The hunters hasten to the spot, seize the animal which does not offer resistance, cut off its horn, which is an excellent antidote to poison, and send it back, without in- flicting on it further harm." 1 Claudius Aelianus flourished under Septimius Severus, and probably outlived Elagabalus (218-222 A.D.). His writings come down from the beginning of the third century (BAUMGARTEN, POLAND, and WAGNER, Die hellenistisch-romische Kultur, p. 615, Leipzig, 1913), while the Physiologus was written in Alexandria as early as the second century (ibid., p. 622). Little is known about Aelian's life; only Philostratus and Suidas have some brief notes regarding him. He availed himself of the writings of Athenaeus, who wrote at the time of Elagabalus, or in the first years of Alexander Severus](222-235) ; Philostratus mentions his death in his Lives of Sophists composed between 222 and 244. As regards the Physiologus, it is necessary to discriminate between the final Greek recension clothed in a Christian-theological garb, as we have it now, and the primeval source or sources of animal stories without the allegories, from which the former was extracted. LAUCHERT (/. c., p. 42) certainly is quite right in rejecting the hypothesis of an " Urphysiologus" in the sense that it was a literary production serving as model to our Physiologus; but a primeval Physiologus must be presupposed for about the beginning of the first century, in the sense that it simply was an assemblage of verbal stories current in Alexandria, and some of which were imported from India (compare T'oung Poo, 1913, pp. 361-4). 1 Ikkaku sennin, eine mittelalterliche japanische Oper (Bastian Festschrift, PP- 5*3-538. Berlin, 1896). Luders, whose work appeared in 1897, did not take note of Muller's investigation ; it seems that the treatises of both scholars originated about the same time, and independently of each other. Compare also J. TAKAKUSU, The Story of the Rsi Ekasrnga (Hansei Zasshi, Vol. XIII, 1898, pp. 10-18); and K. WADAGAKI, Monoceros, The Rishi (ibid., pp. 19-24). ii2 CHINESE CLAY FIGURES rhinoceros. Also O. KELLER 1 has arrived at the same result, and reduced all ancient traditions and representations of the unicorn to the Indian rhinoceros. This opinion seems to me fundamentally wrong. Not one of the numerous variants of the ancient Indian tradition re- garding the Hermit Single-Hofn alludes in this connection to the rhinoceros; he is miraculously born from a gazelle, and has received his horn from the latter. 2 Single-Horn is not even his original name, but this one was Antelope-Horn (Rishya-cringa) ; and according to LUDERS,* the name Single-Horn has arisen from the latter, owing to popular etymological re-interpretation caused by the tradition, already appearing in the Mahabharata that the penitent had a single horn on his head. In other texts, the Padmapurana, Skandapurana, and Kanjur, he is even equipped with two horns, while the versions of the Ramayana and the Pali Jataka make no statement with regard to the horn. The Greek Physiologus, in the story alluded to, avails itself of the word monokeros ("unicorn"), which literally corresponds in meaning to Sanskrit Eka- gringa, and describes the creature as a small animal resembling a buck, without any qualities inherent in the rhinoceros; and this is plainly corroborated by the illustration accompanying the Physiologus, in 1 Die antike Tierwelt, Vol. I, pp. 415-420; this is presumably the weakest chapter of an otherwise intelligent and excellent book. I do not understand how Keller arrives at the opinion that the ancients in general treat monoceros, unicornis, and rhinoceros as identical notions, and in most cases conceive them as the African rhinoceros. The historical connection of the unicorn legend with Ekagringa has escaped Keller en- tirely. 2 The iconography of Ekac.ringa in Indian art has been traced by LUDERS and MtiLLER. It is notable that any suggestion of a rhinoceros is absent. As proved by the masks of the hermit used in the dramatic plays of Japan and Tibet (Plate X), he was conceived as a human being with a single, short, forked horn, or with a very long, curved horn. The illustration of the Japanese mask is derived from the work Nogaku dai-jiten (Dictionary of No Plays) by Masada ShOjirO and Amaya Kangichi (Tokyo, 1908; compare Bulletin de I'Ecole jranqaise d* Extreme-Orient, Vol. IX, 1909, p. 607). The Tibetan mask, much worn off by long use, was obtained by me from a monastery of Bagme, in the western part of the province of Sze-ch'uan. It is very striking that the rhinoceros hardly plays any rdle in the culture-life, folk- lore, or mythology of India. The allusions to it in literary records are exceedingly sparse. The word khadga appears but a few times in Vedic literature, a rhinoceros- hide being mentioned in one passage as the covering of a chariot (MACDONELL and KEITH, Vedic Index, Vol. I, p. 213, London, 1912). The animal is mentioned in the inscriptions of King Acoka (third century B.C.); and the consumption of its flesh,' blood, and urine plays a certain r61e in Indian pharmacology (see CHAKRAVARTI, Mem. As. Soc. Beng., Vol. I, p. 370, Calcutta, 1906; and HOOPER, /. As. Soc. Beng., Vol. VI, 1910, p. 518). It is very curious that no Indian record regarding rhinoceros- horn cups and their antipoisonous virtues has as yet been pointed out; our information on this point rests on Ctesias, Aelian (see below, p. 115), some Arabic authors, and more recent observers like Linschoten and GARCIA AB HORTO (Aromatum et simph'ci- um aliquot medicamentorum apud Indos nascentium historia, p. 66, Antverpiae, I 567), who says, " Illud tamen scio Bengala incolas eius cornu adversus venena usur- pare, unicornu esse existimantes, tametsi non sit, ut ii referunt qui se probe scire autu- mant." It remains to be pointed out also that the literatures of India contain no accounts of unicorns. 1 L. c., p. 28. HISTORY OF THE RHINOCEROS 113 which the animal is outlined as a long-tailed antelope with a large single horn curved like that of a gazelle. 1 PLINY, as we saw, credits the monoceros of India with the head of a stag and a single horn on its fore- head (that is, the gazelle-horned Ekagringa), but does not identify it with the rhinoceros, which was well known to him from the circus. For the first time, as far as the West is concerned, the identification of the single-horned cervine animal with the rhinoceros is made in the Cy- ramdes. 2 In the East, the first intimation of it leaks out in our Chinese illustration from Chtng lei pen ts'ao, which depicts the rhinoceros in the form of a deer with one horn on its forehead, and which, without any doubt, is an offshoot of the Indian conception of Ekacrihga. Now, we encounter the curious fact that at a much older date also the Chinese mention a single-horned deer under the name p'ao (No. 9104), described in the Erh ya as an animal "with the tail of an ox and one horn." PAL- LADius 3 straightway translated the word by "rhinoceros," but this venture is not justified by Chinese tradition; the Chinese, in this case, make no reference whatever to the rhinoceros. On the contrary, Kuo P'o, the editor and interpreter of Erh ya, states that the animal p*ao is identical with the deer called chang (No. 407) ; and Yen Shi-ku (579-645), as quoted in K'ang-hi's Dictionary, maintains that it re- sembles in shape the deer chang. The very definition shows that the animal p'ao is a near cousin of the k'i-lin* which has likewise "the tail 1 Figured by STRZYGOWSKI, Der Bilderkreis des griechischen Physiologus, Plate XII (Byzantinische Zeitschrift, Erganzungsheft I, 1899), and KELLER (/. c., p. 419). Regarding the illuminated editions of the Physiologus see also O. M. DALTON, Byzan- tine Art, p. 482 (Oxford, 1911). * Neither LIJDERS nor MULLER has consulted these two important passages of Pliny and the Cyranides. * Chinese-Russian Dictionary, Vol. I, p. 58. 4 At times a temptation was felt to identify the animal lin with the rhinoceros. Shen Kua, the versatile author of the Ming k'i pi fan of the twelfth century, narrates that in the period Chi-ho (1054-56) the country Kiao-chi (Annam) offered a lin like an ox, having the entire body covered with large scales and a single horn on its head. There is no question that this animal was a rhinoceros; this follows also from the further observation of the author that it did not resemble the lin, as described in ancient records, and that there were people designating it as a mountain-rhinoceros (shan si, a variety recognized also by Li Shi-chen). But as Sh6n Kua could not trace any report in which scales are attributed to the rhinoceros (for explanation see p. 149), he formed the erroneous theory that the animal in question was identical with the T'ien-lu cast in bronze by the Emperor Ling in 186 A.D., a specimen of which he had beheld at Nan-yang in Teng chou in Ho-nan. In a similar manner, Fan Ch6n of the Sung period, in his work Tung chai ki ski (Ch. I, p. 8; in Shou shan ko ts'ung shu, Vol. 84), tells the story of two K'i-lin sent as tribute from Kiao-chi in the period Kia- yu (1056-63), which he had occasion to see in the imperial palace. He describes them as having the shape of water-buffalo clad with a fleshy armor, and equipped with a single horn on the extremity of the nose; they subsisted on grass, fruit, and melon, and every time before feeding had to be beaten on their horns with a stick. This writer likewise concludes with a discussion, in which serious doubts of the identifica- tion of these animals with the lin are expressed. ii4 CHINESE CLAY FIGURES of an ox and a single horn." l Indeed in the Erh yu f w, both creatures are figured almost alike, and agree in their essential characteristics. It is obvious that, as iconographic types, these creatures are not derived from any rhinoceros, but point in the direction of the fabulous one- horned monsters (known in archaeology as "Oriental animals") de- veloped in the art of Mesopotamia. 2 In regard to the type of k*i-lin, this has been aptly pointed out by A. GRUNWEDEL;* and as the same West-Asiatic forms found their way into the art of India, we here have the basis for the origin of the single-horned gazelle (deer or antelope) transferred to, or personified in, the person of Ekac.ririga. In Baby- lonia, these types of unicorn are very ancient, going back to the third millennium B.C., 4 and could not have been developed there from a rhinoceros. The conclusion therefore presents itself that the notion of a unicorn cervine animal which was developed in Western Asia from remote times spread together with artistic motives into India and China, 5 while the identification of this fabulous creature with the 1 Regarding the k'i-lin see Yen Shi-ku (in Ts'ien Han shu, Ch. 6, p. 5 b) ; MAYERS (Chinese Reader's Manual, p. 127); F. W. K. MILLER (in Feestbundel aan P. J. Veth, p. 222, Leiden, 1894); DE GROOT (The Religious System of China, Vol. II, pp. 822- 4); and H. DOR (Recherches sur les superstitions en Chine, pt. I, Vol. II, pp. 446-8). I dp not subscribe to everything that the last two authors say about the subject. The Chinese illustrations are reproduced in C. GOULD (Mythical Monsters, pp. 350, 353, 354, London, 1886). 8 A distinction must be made between iconographic or archaeological type or artistic representation, and traditions or speculations regarding such a type. The lin, as early mentioned in Shi king and Li ki, may very well be an indigenous Chinese thought. Nevertheless its subsequent portrayal in art rests on a borrowed type, which has again fertilized native ideas as to form and behavior of the creature. An interesting example of the fact that iconography and literary tradition may move along lines widely different and emanating from diverse sources is afforded by the unicorn of Europe. The unicorn tradition of the Physiologus is traceable to India; the iconography of the creature, however, has no connection with Indian art, but leans in the beginning toward the ancient West-Asiatic types. Throughout the middle ages, there is not a trace of the rhinoceros in the representations of the unicorn (compare Marco Polo's astonishment when he saw the ugly beast on Java, "not in the least like that which our stories tell of as being caught in the lap of a virgin, in fact, altogether different from what we fancied"); now it is an antelope, now an ox, now a narwhal, now a hybrid formation composed of various creatures. My opinion in this respect deviates from the one expressed by STRZYGOWSKI (/. c.) that there may be interaction between the animal types of the earliest Buddhist art in India and those of the Physiologus. It is not there the question of interaction, but of affinity, solely caused by West-Asiatic productions which both have in common as their source. 1 Bemerkungen uber das Kilin (Feestbundel aan P. J. Veth, pp. 223-5, Leiden, 1894), and Buddhist Art in India, p. 19. 4 E. SCHRADER, Die Vorstellung vom monokeros und ihr Ursprung (Abhandlungen der preussischen Akademie, 1892, pp. 573-581). 6 In order to dispel the doubts of those who may not feel inclined in this case to link China with the West, another striking analogy may be indicated, which will show that Chinese ideas regarding unicorns coincide with those entertained in the West, and which crop up in the classical authors. In the Erh ya is denned an animal called chui (written with the classifier 'horse' and the phonetic complement sui, No. 10,388), ' 'like a horse with a single horn ; those without horn are spotted." Kuo P'o comments, HISTORY OF THE RHINOCEROS 115 rhinoceros owing to the single horn is the product of a much later period; this is not the starting-point, but the final result of the matter. It is, of course, necessary to assume that this result was brought about in India itself; 1 otherwise it would be unintelligible why it appears on the surface in the Cyranides and in China. 2 In my opinion, we are even "In the eighth year of the period Yuan-k'ang (298 A.D.) it was in the territory of Kiu-chen (in Tonking) that hunters captured a wild animal of the size of a horse with one horn, the horn being soft as the core of the young antlers of the deer (lu Jung). This is identical with the animal chui. At present men sometimes meet it in the dense mountainous jungles, and there are among them also those without horn." Kiu-ch6n is situated in,Tonking; and on p. 81 mention has been made of the tribute of a live rhi- noceros sent from there to the Emperor Ling (168-188 A.D.); indeed, that region was always famed for this animal, which is apparently intended in the text of Kuo P'o. The same conception of the rhinoceros as a horse or horse-like animal with a single horn is met likewise in the West. The ancients enumerate altogether five animals as having single horns, the Indian ass first traceable to Ctesias, the single-horned ox, the monoceros, the single-horned horse, and the oryx of Africa. STRABO (xv, 56) quotes from Megasthenes' remarks upon Indian animals that there are horses in India with one horn. AELIAN (Nat. anim., in, 41) says, "India, it is reported, pro- duces horses with a single horn, likewise single-horned asses. Cups are made from these horns; and if a mortal poison is poured into them, it will do no harm to him who drinks it, for the horn of both animals seems to be an antidote against poison." In another chapter (xvi, 20) AELIAN describes the unicorn of the Indians, "called by them kartazonos [a word apparently connected with Assyrian kurkizannu, mentioned above, p. 87], said to equal in size a full-grown horse." HORACE (Serm., i, 5, 58-60) speaks of a wild horse having a single horn in the midst of its forehead. As a matter of fact, the rhinoceros has no similarity to a horse; and it is difficult to see how the simile could ever arise. The bare fact remains, however, that it did; but it is incon- ceivable that this notion, not founded on a natural observation, could spontaneously spring up in the West and East alike. There is no other way out of this puzzle than to presume that India, to which the account of Megasthenes reproduced by Strabo and Aelian refers, is responsible for this idea, and disseminated it to the West and to China. 1 It may be pointed out in this connection, though it is not wholly conclusive for the present case, that the Sanskrit word vdrdhranasa means a rhinoceros and an old white goat-buck. 1 We meet also in ancient China a unicorn conceived of as a wild goat. This is the animal termed chat (No. 245) and hiai (No. 4423) chai. The fundamental passage relating to it is in the Annals of the Later Han Dynasty (Hou Han shu, Ch. 40, p. 3), where a judicial cap in the shape of this animal, and worn by the censors, is mentioned. The definition given of the animal in the text of the Annals is, "A divine goat (shin yang) which is able to discriminate between right and wrong, and which the king of Ch'u used to capture." Huai-nan-tse is quoted in K'ang-hi (under hiai) as saying that King Wen of Ch'u was fond of wearing hiai caps; the un-Chinese word hiai chat, therefore, will probably be a word of the language of Ch'u (T. DE LACOUPERIE, Les langues de la Chine avant les Chinois, p. 17, Paris, 1888), as above all proved by the vacillating modes of writing (FoRKE, Lun-heng, pt. II, p. 321). The comment added to the text of Hou Han shu is extracted from / wu chi, which may be read in SCHLE- GEL'S Uranographie chinoise, p. 587 (it is, of course, impossible, as proposed by Schle- gel, to identify the animal with the Tibetan chiru; see below, p. 120). It is not stated in Hou Han shu nor in / wu chi (nor in K'ang-hi) that "it eats fire in its ravenous fury, even to its own destruction " (GILES). This is a subsequent addition which arose un- der the influence of Buddhist art. F. W. K. MILLER (Fecstbundel aan P. J. Veth, p. 222, Leiden, 1894) has recognized correctly that this explanation is derived from the iconography of the animal, which is represented as being surrounded by flames. Muller, however, omits to state that this is a secondary development, which has nothing to do with the previous pre-Buddhistic conception of the creature on Chinese soil, when it was not equipped with flames, nor set in relation with a lion. The n6 CHINESE CLAY FIGURES forced to admit that the counterpart to the illustration of the Chtng lei pn ts'ao has already pre-existed in India, and was transmitted from there to China; for neither the author of that work, nor any other Chinese source, as far as I know, furnishes any explanation for this picture. An unexpected confirmation of this opinion comes to us from another quarter, Tibet. In the Tibetan language we meet the word bse-ru which at present denotes two animals, first, the rhinoceros, and second, a kind of antelope. The former is the original and older significance, the latter is secondary. The second element of the compound, ru, means "horn," and may be dropped; the proper word is bse (pronounced se). The stem is se, the prefixed labial 6- not being part of the word-stem, and like most prefixes in Tibetan nouns, representing the survival of an ancient numerative. This is corroborated by the corresponding Lepcha word sa and the Chinese word se, all three referring to the rhinoceros. This linguistic coincidence leads to the conclusion that the Chinese and Tibetans as stocks of the large Indo-Chinese family of peoples were acquainted with the rhinoceros in prehistoric times, for otherwise they could not have the word for it in common; and this conclusion will be fully upheld by our historical inquiry into the subject. This fact of comparative philology is also apt to refute the supposition of Mr. Giles that "a term which originally meant a bovine animal was later on wrongly applied to the rhinoceros." As proved by comparison with the Tibetan and Lepcha words, the Chinese term originally must have designated the rhinoceros. 1 Above all it is incumbent upon me to demonstrate that the Tibetan word bse really designates the rhinoceros, and that the Tibetans were familiar with this animal. The ancient translation "lion-unicorn" adopted by Muller is not to the point, as far as the time of Chinese antiquity is concerned. The kiai chai is not explained as a lion (nor could this be expected, as the lion was unknown in ancient China), but as a divine wild goat (sh&n yang). The fact that the conception of the animal existed among the Chinese in times prior to the contact with India is clearly proved by the occurrence of the word in Huai-nan-tse, in Tso chuan (Suan Wang I7th year: LEGGE, Chinese Classics, Vol. V, p. 332), Se-ma Ts'ien's Shi ki (Ch. 117), Lun htng, Hou Han shu, Erh ya, and Shuo wdn. Only in such late compilations as the Japanese version of the San ts'ai t'u hui do we meet the statement that the animal resembles a lion, merely because it is sketched like a lion crowned with a single horn (see L. SERRURIER, Encycl. japonaise, le chapitre des quadruples, Plate III; or E. KAEMPFER, The History of Japan, Vol. I, p. 195, Glasgow, 1906). The connection of this creature with the rhinoceros, and its transformation into a goat, will be discussed below (p. 171). 1 The hypothesis of such "confusions, " which are usually assumed to suit one's own convenience, is untenable also for other reasons obvious to every ethnologist: people in the primitive stages of culture, being nearer to nature than we, are surely the keenest observers of animal life and habits, and will most assuredly never con- found a bovine animal with a rhinoceros; they may, by way of explanation, compare the one with the other, but from comparison to confusion is a wide step. HISTORY OF THE RHINOCEROS 117 Sanskrit-Tibetan dictionary Mahavyutpatti 1 renders the Tibetan word bse by the Sanskrit word ganda which refers to the rhinoceros. 2 Wherever this word appears in the works of Sanskrit Buddhist litera- ture, it is faithfully reproduced in the Tibetan translations by the word bse. An interesting example of its application appears in a Tibetan work from the first part of the ninth century. 3 It is well known that in India the Pratyeka-Buddha was styled Single-Horn Hermit and com- pared with the solitary rhinoceros; 4 and this simile is explained in that Tibetan book in the words that the Pratyeka-Buddha, who in the course of a hundred eons (kalpa), through the accumulation of merit, is no longer like ordinary beings, resembles the rhinoceros in his habit of living in the same solitary abode. It is interesting to note that in this early Tibetan text the word bse-ru is used for the designation of the rhinoceros. This comparison has passed into Tibetan poetry, and is frequently employed by the mystic and poet Milaraspa, who speaks of himself as being "lonely like a rhinoceros." 6 This meaning of bse is confirmed by two Chinese lexicographical sources, the Hua i yi yti, which in its Tibetan-Chinese vocabulary 6 renders bse-ru by Chinese si niu; and the Polyglot Dictionary of the Emperor K'ien-lung (Ch. 31, p. 4 a), where bse is explained by Chinese si ("rhinoceros"). The national Tibetan word bse, akin to Lepcha sa and Chinese se, naturally bears out the fact that the ancient Tibetans were familiar with the 1 Tanjur (Palace edition), Sutra, Vol. 123, fol. 265 a. This work was written in the first part of the ninth century. * Al-Berunl (SACHAU, Alberuni's India, Vol. I, p. 203) knew this word, and cor- rectly described under it the rhinoceros of India (p. 95). It is likewise mentioned by GARCIA AB HORTO (/. c.) and other early European travellers enumerated by YULE and BURNELL (Hobson-Jobson, p. 363). The rhinoceros brought to Portugal in 1515 (mentioned above, p. 83) was labelled "rhinocero, called in Indian gomda." 1 Entitled Sgra sbyor bam-po gnis-pa (Tanjur, Satra, Vol. 124, fol. 14 a, 4), cor- rectly dated by G. HUTH (Sitzungsberichte der preussischen Akademie, 1895, p. 277) in the first part of the ninth century. Compare also the application of the word in Taranatha (SCHIEFNER'S translation, p. 245): the sorcerer Ri-ri-pa summoned the fierce beasts of the forest, the rhinoceros and others, and mbunted on their backs. 4 EITEL, Hand-book of Chinese Buddhism (pp. 76, 123, 197); F. W. K. MILLER, Ikkaku sennin (1. c., p. 530); and H. KERN, Manual of Indian Buddhism (pp. 61 and 62, note i). * G. SANDBERG (Tibet and the Tibetans, p. 297), who is ignorant of the fact that bse or bse-ru means "rhinoceros," and who merely carries the modern popular meaning of the word, "antelope," into the sphere of literature, makes Milaraspa say that he is "lonely as a seru" (antelope). The antelope, however, is not a lonely, but a highly social animal living in herds. Nowhere in Buddhist literature has bse-ru the signifi- cance of ' 'antelope, but only that of ' 'rhinoceros. ' ' The Tibetan poet, who in every line is imbued with the language and spirit of India, most obviously intends with this simile a literary allusion to the Buddhist comparison of the Pratyeka-Buddha with the rhinoceros. * Copied by me from the manuscript deposited by HIRTH in the Royal Library of Berlin. Regarding the work see HIRTH (/. China Branch R. As. Soc., Vol. XXII, 1888, pp. 207 et seq.), and Bull. Ecole franf aise, IQI2, p. 199. u8 CHINESE CLAY FIGURES animal. We know that the primeval habitat of the Tibetan stock was located along the upper course of the Huang-ho (where Ptolemy knows them as Bautai, derived from the native name Bod, "Tibetans;" the Yellow River is styled by him Bautisos) , as well as along the upper Yang- tse. There they lived in close proximity to the ancient Chinese; and in that locality, as will be established from Chinese records, the rhi- noceros was their contemporary. Large parts of the present Chinese provinces of Kan-su and Sze-ch'uan are still settled by Tibetan tribes; and we shall see that the rhinoceros occurred there in the times of antiquity, and long survived, even down to the middle ages. The Pai- lan a tribe belonging to the Tibetan group of the K'iang, and border- ing in the north-east on the Tu-yu-hun in 561 A.D. sent an embassy to China to present a cuirass of rhinoceros-hide (5* kid) and iron armor. 1 Whether they had made this cuirass themselves, or had received it from an outside source (this fact is not indicated), this tribute, at any rate, shows that they were acquainted with this material and its manu- factures. 2 The Ptn ts'ao yen i of 1116 extols the horns of the Tibetan breed of rhinoceros for the fine quality of the natural designs displayed in them (see p. 148). Li Shi-ch6n,in his Pen ts'ao kang mu (see p. 149), expressly names as habitats of the rhinoceros the regions of the Si Fan and Nan Fan; that is, the western and southern Tibetans, the former scattered over Sze-ch'uan and Yun-nan with their borderlands, the latter peopling the valley of the Tsang-po (Brahmaputra) and the Himalayan tracts adjoining India. Indeed, down to the middle of the nineteenth century, or even later, the rhinoceros was to be met with along the foot of the Himalaya as far west as Rohilkund and Nepal; and it survived longer still in the Terai of Sikkim. 4 J. CH. WHITE 4 notes the 1 Chou shu, Ch. 49, p. 5 b. 1 In the year 824 the Tibetans offered to the Chinese Court silver-cast figures of a rhinoceros and a stag (T'ang shu, Ch. 216 B, p. 6 b). BUSHELL (The Early History of Tibet, p. 88) translates the word si in this passage by "yak," but this point of view is not admissible. True it is that some modern Chinese writers on Tibet call the yak si niu, but this usage of the word is not earlier than the eighteenth century. The T'ang Annals, however, persistently designate the Tibetan yak by the word li niu (No. 6938) ; and in the very passage alluded to, the gift of the rhinoceros and stag silver figures is immediately followed by the words, "and they brought as tribute a yak" (kung li niu), which BUSHELL correctly interprets likewise as yak. The words si and li niu in the same sentence cannot possibly refer to the same animal; and it becomes evident from a consideration of all Chinese sources concerned that down to the end of the Ming dynasty the Chinese word si with reference to Tibet and Tibetan tribes invariably denotes the rhinoceros, and nothing else. Rhinoceros-horn was formerly included among the tribute gifts which the Dalai Lamas of Tibet were obliged to send to China; it took its place between coral, genuine pearls, precious stones, amber, etc. (Wei Tsang t'u chi, 1792, Ch. A, p. 17). R. LYDEKKER, The Game Animals of India, p. 30. 4 Sikhim and Bhutan, p. 322 (London, 1909). HISTORY OF THE RHINOCEROS 119 rhinoceros in a few of the lower valleys of Bhutan, though not common. In Tibet proper, the animal does not occur at present, but fossil remains of it were discovered at high elevations by Sir R. Strachey near the source of the Tsang-po. 1 The early Tibetan translators, when they correctly rendered the Sanskrit word ganda by bse, must have entertained an exact notion or reminiscence of the rhinoceros; but the animal, as every- where, became rapidly exterminated in those territories where Tibetans had occasion to behold and to hunt it, while the inhabitants of Central Tibet seldom or never had this opportunity. For this reason, also in Tibet, the rhinoceros underwent the process of fabulous "unicorniza- tion." Reports of a Tibetan unicorn greatly stirred the imagination of European explorers, and gave rise to wild speculations. Captain S. TURNER, 2 I believe, was the first to circulate such a report, being in- formed by the Raja of Bhutan that he was in possession of a unicorn, a sort of horse, with a horn growing from the middle of its forehead; it was kept at some distance from Tassisudon, the capital, and the people paid it religious respect, but Turner had no occasion to see it. The Lazarist fathers Hue and Gabet, who reached Lhasa in 1846, are said to have even claimed the discovery in Tibet of the unicorn of Scripture. Major Latter, in the first part of the nineteenth century, was very sanguine of being able to find a veritable unicorn in the interior of Tibet: he was advised by a native that he had often seen these an- imals, which "were fierce and exceedingly wild and seldom taken alive, but frequently shot;" and that they are commonly met with on the borders of the great desert, about a mile from Lhasa. From a drawing which accompanied Major Latter's communication, the presumed unicorn was something like a horse, but with cloven hoofs, a long, curved horn growing out of the forehead, and a boar-shaped tail. Un- der the heading "Unicorns in Asia," 8 a writer revived the opinion of the existence of veritable unicorns, such as were reported to Major Latter: the animal in question was of the deer kind, having a single horn at the top of the head; it was known by the name of the Seru. 4 Then 1 A. R. WALLACE (The Geographical Distribution of Animals, Vol. II, p. 214; also Vol. I, p. 122) refers to this in the words that more than twenty species of extinct rhinoceroses are known, and that one has even been found at an altitude of 16,000 feet in Tibet. Mr. L. A. WADDELL (Lhasa and its Mysteries, p. 315) has this sugges- tive remark: "The dense rank growth of wildflowers and weeds along the borders of the fields was such as to make this part of the Tsang-po oasis a quite suitable habitat for the rhinoceros, and to bring the discovery of the fossil remains of that animal by Sir R. Strachey near the source of this river into harmony with present-day facts. ' * An Account of an Embassy to the Court of the Teshoo Lama, p. 157 (London, 1800). 1 Asiatic Journal, Vol. II, 1830. * Compare W. HAUGHTON, On the Unicorn of the Ancients (Annals and Magazine of Nat. Hist., Vol. IX, 1862, pp. 368, 369). 120 CHINESE CLAY FIGURES the famous J. D. HOOKER x took the matter in hand, and published a sketch of the Chiru Antelope with the addition "unicorn of Tibet," a name which he thought was suggested by the animal when viewed in profile. It is identified as Antilope or Paniholops Hodgsoni, having been described by HODGSON. 2 It remains a mysterious creature, and little is known about it. 3 P. LANDON* denies that this antelope, as pointed out by Hooker, occurs near the Cholamu Lake at the present day. L. A. WADDELL 5 reports under Chiru, "None were seen and the people did not appear to know of any." In Anglo-Indian nomenclature we now find two words in use, chiru and seru, the latter also Anglicized as serow, on which YULE, in his " Hobson-Jobson, " unfortunately has not commented. Serow has be- come a household stock-word of the Anglo-Indian sportsman to denote a large variety of different Indian, Burmese, and Tibetan antelopes. 6 G. SANDBERG' recognizes in it the Tibetan word bse-ru, and identifies the latter with the species Nemorhaedus bubalinus. JASCHKE S says under bse or bse-ru, "Unicorn, 'tchiru,' an antelope, probably the same as gtsod," with reference to Hooker. CHANDRA DAS, S who has fully 1 Himalayan Journals, 2d ed., p. 401 (London, 1893). 2 Journal As. Soc. Bengal, 1846, p. 338. 8 N. KUEHNER, Description of Tibet, in Russian (Vol. I, pt. 2, p. 157; and notes P- 77). 4 Lhasa, Vol. I, p. 393. 8 Lhasa and its Mysteries, p. 483. 6 R. LYDEKKER, The Game Animals of India, pp. 139 el seq. M. DAUVERGNE (Bull. Musee d'hist. nat. de Parts, Vol. IV, 1898, p. 219) describes the animal as follows: "Serow; Ramu de Kashmir, ou chevre-antilope, Nemorhaedus bubalinus Hodgs. Habite les rochers escarpe's et broussailleux des montagnes, a une hauteur de 3,000 metres, dans 1'Himalaya et Kashmir. Tres difficile a chasser, il tient te"te aux chiens, qu'il fait rouler dans les precipices. C'est ge'ne'ralement 1'hiver qu'on le chasse, car alors il se de"tache sur la neige, grace a la teinte noire de sa robe, et comme il est tres lourd, il s'effondre et se fait prendre par les chiens." 7 Tibet and the Tibetans, p. 297. On p. 298 he points out that the word chiru should be written gcig ru ("one horn"). This derivation is impossible, as "one horn" can be in Tibetan only ru (or rva) gcig, or ru &ig. The name Ekagringa is rendered in- to Tibetan Rva gcig-pa. (Compare also Hor c'os byun, ed. HUTH, p. 16, 1. 14.) Chiru is simply a local or dialectic variation of se-ru. Strange words exert a singular fascina- tion upon the human mind. The Anglo-Indian chiru has had several good fortunes. Thanks to the imaginative powers of G. SCHLEGEL (Uranographie chinoise, p. 587), it has found cheerful hospitality in Chinese astronomy, the Chinese animal hiai being wrongly identified with it. A few years ago the chiru was deemed worthy of the honor of being admitted into the sanctum of classical philology. O. KELLER (Die antike Tierwelt, Vol. I, p. 293) identifies the Indian Oryx mentioned by Aelian, and the Oryx on the Hydaspes mentioned by Timotheus, with the Tibetan chiru, a venture which has no foundation; in fact, the oryx of Aelian is located in India, and corresponds to the Indian black-buck. 8 Tibetan-English Dictionary, p. 593. Skr. khatfga rendered by JASCHKE "a cer- tain animal" is the rhinoceros. 9 Tibetan-English Dictionary, p. 1319. HISTORY OF THE RHINOCEROS 121 recognized the original meaning of bse-ru as "rhinoceros," proceeds to state that in Tibet the word is applied to the clumsy-looking deer known to sportsmen as the "serow." Both lexicographers, in this respect, rely on the statements of the European sportsmen, but leave us in the dark as to the opinion of the Tibetans on the point. The question arises, Do those European speculations on a Tibetan unicorn identified with an antelope styled se-ru have any foundation in a Tibetan tradi- tion? The French Missionaries, in their Tibetan Dictionary (p. 1056), give a slight intimation of the existence of such a tradition by remarking that the animal bse-ru is believed in Tibet to belong to the genus of goats (ex genere caprarum), but that nobody has ever seen it; the latter clause doubtless means that nobody has encountered this wild goat in the shape of a unicorn which it is fabled to be. I. J. SCHMIDT l had a certain presentiment of the matter when he annotated a passage in his translation of the Geser Saga, that the Tibetan and Mongol name of the unicorn is seru, that the existence of this animal in the wild moun- tains of Tibet is asserted in Tibetan books, but that the description given of it does not at all fit the rhinoceros. The unicorn which stopped Chinggis Khan on his expedition to Tibet and induced him to return, 2 judging from the description given by the Tibetan historian, * is identical with the Chinese k'i-lin, as already recognized by G. ScHLEGEL. 4 Another association of the unicorn with Tibet appears on the tribute painting ascribed to Li Kung-lin (Li Lung-mien), where BONIN B has pointed it out among the envoys from the Kingdom of Women. In the Polyglot Dictionary of the Emperor K'ien-lung 8 we find the Tibetan 1 Die Thaten Bogda Gesser Chan's, p. 56 (St. Petersburg, 1839). Compare also P- 125. 1 G. HUTH, Geschichte des Buddhismus in der Mongolei, Vol. II, p. 25. 1 "An animal of green color with the body of a stag, the tail of a horse, and a single horn on its head." 4 T'oung Poo, Vol. VI, 1896, p. 433. According to Chinese tradition, however (see the texts of Kui sin tsa chi and Ch'o keng lu, in T'u shu tsi ch'tng, Chapter kio tuan, ki ski, p. i b), the marvellous animal opposing the conqueror belonged to the class of unicorns (kio tuan), and is described^ as a hundred feet high, with a single horn like that of the rhinoceros, and able to speak a human language. 6 Le rpyaume des neiges, pp. 40, 299 (Paris, 1911). M. Benin's description of this painting is based on a copy of it in the Mus6e Guimet, which is certainly not the original from the hand of Li Kung-lin; it is a much later and somewhat weak copy, as stated also by TCHANG Yi-Tcnou and HACKIN (La peinture chinoise au Musee Gui- met, p. 59). On Plate V of the latter publication, the portion of the picture illustrat- ing the envoys of the Kingdom of Women is reproduced; the unicorn is a wretched production. Mr. Freer of Detroit owns two copies of the same painting, both far superior to the one in the Muse"e Guimet. One or these offers 3uch high qualities as come very near to an original. The other is a copy of the Yuan period, executed in 1364. Appendix, Ch. 4, p. 53. 122 CHINESE CLAY FIGURES FIG. 14. Se-ru as Emblem of Long Life (from Tibetan Wood-engraving). word bse-ru rendered by Chinese shen yang ("divine goat"); 1 and this is thus far the only literary indication which I am able to trace in regard to a Tibetan unicorn of goat-like character. 2 Such a bse-ru is represented on a Tibetan woodcut as an emblem of long life (bse-ru ts'e rin; Fig. 14). The picture, of which it forms a 1 The Manchu has the artificial formation Sengkitu, and three other words besides, sacintu, tontu, and tubitu (see SACHAROV, Manchu-Russian Dictionary, P- 734) f r the designation of this unicorn. It will be remembered that the term shin yang occurs in Hou Han shu in defining the unicorn hiai chat (p. 115, note 2). 1 The Mongols have adopted seru as a loan-word from Tibetan in the sense of "rhinoceros," as stated by KOVALEVSKI and GOLSTUNSKI in their Mongol dictionaries; but they take the word also in the sense of a "deer," as shown by the Mongol transla- tion of the Tibetan medical work translated into Russian by A. PQZDNAYEV (Vol. I, p. 288). The Mongol equivalent of Tibetan bse-ru and Chinese si kio is here bodi guriigasiin ("the animal of the bodhi," Sanskrit bodhimriga); that is, the gazelle. Besides, the Mongols have a seemingly indigenous word for "rhinoceros," kiris, keris, or kers-un dbar. HISTORY or THE RHINOCEROS 123 part, is known as "the six subjects of long life" (ts'e rin drug skor). These are, the Buddha Amitayus (the Buddha of Endless Life) , the long-lived wishing- tree (dpag bsam Sin ts'e rin) figured as a peach-tree in Chinese style, the long-lived rocks (brag ts'e rin), the Chinese God of Longevity Shou-sing (in Tibetan Mi ts'e rin) seated on a mat and holding a rosary, a pair of cranes (krun krun ts'e rin) pecking at some peaches (k'am-bu) that are planted in a jar, and a pair of bse-ru. Though apparently inspired by the deer, which is the emblem of the Chinese God of Longevity, their outlines considerably differ from the latter, and approach the Tibetan notion of the appearance of a'bse-ru; l but, curiously enough, they are without any horns. There can be little doubt, ac- cordingly, that in recent times, when the rhinoceros had almost vanished from the memory of the Tibetan people, the word bse-ru was transferred to a species of deer or antelope; and, as the ancient tradition of the bse-ru being a single-horned animal had persisted through the centuries, the single horn, in popular imagination, was fixed on the antelope. When we inquire why it was just the antelope, and not any other animal on which the idea of a unicorn was projected, the story of Ekagringa pre- sents itself again as the happiest solution. We know that this legend, in a Tibetan translation, has been incorporated in the Kanjur; and A. SCHIEFNER 2 has translated it from this version. It is likewise extant in Kshemendra's Avadanakalpalata, of which a literal versified rendering, and an abridged prose edition made for children by order of the Fifth Dalai Lama, exist in the Tibetan language. This plain version has ren- dered the story immensely popular among Tibetans; and, as pointed out, it is current also in a dramatized form. The Tibetan mask of Ekasringa (Plate X) is equipped with an unmistakable antelope-horn. 3 The psychological process is therefore quite clear. The rhinoceros was grad- 1 My explanation is based on the interpretation of this woodcut given me by an intelligent Lama. A. GRfJNWEDEL, in his Russian Description of the Lamaist Collec- tion of Prince Uchtomski (Bill. Buddhica, No. 6, p. 26), has figured a similar woodcut, but without explanation. The God of Longevity bears the Mongol legend Tsaghan Abughdn ("The White Old Man"), who is certainly, as stated on p. 117, a national Mongol deity; but from an iconographic point of view, as he appears in Grunwedel's drawing, he is nothing but a copy of the well-known Chinese God of Longevity. 1 In RALSTON, Tibetan Tales, p. 253. * On the lid of a Tibetan censer in the Field Museum (Cat. No. 122,522) are represented the full figures of two gazelles opposite and turned away from each other (the wheel of the law being placed between them), the well-known Buddhist motive symbolizing Buddha's first sermon in the Deer-Park (GRttNWEDEL, Buddhist Art in India, p. 143). One of these is provided with a single horn on its forehead ; the other, apparently conceived as the doe, is hornless. The former seems suggested again by a reminiscence of Ekacringa, but it is not known to me whether the Tibetans would name it bse-ru. Other Tibetan censers are surmounted by a monster of Chinese style, showing a horn on its nose and another on its forehead, manifestly derived from the two-horned rhinoceros. 124 CHINESE CLAY FIGURES ually forgotten by the people, the word bse or bse-ru of this meaning continued in literature; the people retained the recollection of its being a single-horned animal, and in their attempts at finding this creature, the legend of Hermit Single-Horn, the son of an antelope or gazelle, flashed into their minds ; so that the unicorn bse-ru was finally identified with a species of antelope named for this reason bse-ru. This unicorn bse-ru we now recognize also in the Chinese drawing of Cheng lei pen ts'ao (Fig. 13). Since the proof is now established that the interaction and intermingling of deer and rhinoceros have taken place in China, in Tibet, and in the West with the first conspicuous allusion in the Cy- ranides, 1 and that this process of adjustment and affiliation has radiated from the Indian legend of Single-Horn born from a gazelle, we are justi- fied in concluding that the foundation, or at least the commencement, of this transformation, must have arisen in India. The development of the matter in Tibet shows sufficiently that Ekacriiiga is disguised also under our Chinese illustration. So much about the latter. A most interesting psychological parallel to the representations of the rhinoceros in China is formed by the ostrich. We now know from the reproductions of CHAVANNEs 2 that in the T'ang period the ostrich was chiselled in stone in a very naturalistic manner on the imperial burial-places (Fig. is). 3 1 A counterpart of the rhinoceros of cervine character occurs also among the Arabs. In Ethiopic, the word charish corresponds to the monokeros of the Septuaginta (JOB, xxxix, 9) , and in all probability signifies the "rhinoceros. ' ' According to Qazwlnl, charish is an animal of the size of a ram, of great strength and swiftness, with a single horn on its forehead like the horn of the rhinoceros (karkadan). Some Arabic lexicog- raphers even take it for a marine animal, others identify it directly with the rhinoce- ros. HOMMEL (Die Namen der Saugetiere bei den sudsemitischen Volkern, p. 333, Leipzig, 1879), to whom this information is due, regards the Arabic word as a loan from Ethiopic. DamM, in his Lexicon of Animals, avails himself of this word in trans- lating the text of the Physiologus regarding the unicorn (K. AHRENS, Das Buch der Naturgegenstande, p. 43). What escaped Hommel is the fact that Cosmas Indico- pleustes (McCRiNDLE, Ancient India as described in Class. Lit., p. 157) states that the Ethiopians, in their language, call the rhinoceros arou or harisi. G. JACOB (Studien in arabischen Geographen IV, p. 166, Berlin, 1892) holds that Qazwlnl is the only Arabic author to discriminate between charish and the rhinoceros, and identifies the former with the Saiga-antelope of southern Russia. The rendering "unicorn" by the Seventy and the English Bible is erroneous. The Hebrew word, thus translated, is reem, corresponding to Assyrian rtmu. It is now generally interpreted as a wild buffalo, and on the basis of Assyrian monuments is ingeniously identified with Bos primigenius by J. U. DURST (Die Kinder von Babylonien, pp. 8-n, Berlin, 1899). The animal, called in Hebrew behemoth (Jos, XL, 15-24), and formerly taken for the rhinoceros (p. 83), is the hippopotamus of the Nile. The Bible does not mention the rhinoceros or the unicorn. * Mission archeologique, Nos. 458, 459, 472, 481. * These ostriches belong to the very best ever executed in the history of art. They are much superior to any representations of the bird by the Egyptians (O. KELLER, Die antike Tierwelt, Vol. II, p. 170), the Assyrians (P. S. P. HANDCOCK, Mesopotami- an Archaeology, p. 307), and the classical nations (!MHOOF-BLUMER and O. KELLER, Tier- und Pflanzenbilder auf Munzen und Gemmen, Plates V, 52; XXII, 33-36). HISTORY OF THE RHINOCEROS 125 It was the great general and explorer Chang K'ien, the first modern Chinese, who during his peregrinations to the west, among many other novel things, discovered also the ostrich for his compatriots. After he had negotiated his treaties with the countries of the west, the King of Parthia (An-si) sent an embassy to the Chinese Court and presented large bird's eggs, 1 which most probably were ostrich eggs. A live FIG. 15. Ostrich sculptured in Stone, T'ang Period (Sketch after Chavannes, Mission, No. 472). specimen (or specimens) of the "large bird of T'iao-chi" was despatched as tribute from the same country in 101 A.D., and termed in China "Parthian bird." 2 They are not made after any western artistic models, but constitute invincible proof for the fact that the Chinese artists in the T'ang era observed and studied nature, and worked after natural models. This case may be recommended for due considera- tion to the adherents of the preconceived dogma that all Chinese art is copied from that of the west, and that no art is possible outside of the sanctum of classical art. 1 Shi ki, Ch. 123, p. 6; HIRTH, China and the Roman Orient, p. 169. FORKE (Mitteilungen des Seminars, Vol. VII, 1904, p. 139) wrongly says that the Shi ki mentions "large birds (ostriches) with eggs as large as earthen pots as a peculiar feature of T'iao-chi;" this is not in the text of the Shi ki, which speaks only of large bird's eggs, but it is found in Ts'ien Han shu (Ch. 96 A, p. 6 a). The trade in ostrich eggs in the west is of very ancient date (O. KELLER, /. c., p. 168). 1 Hou Han shu, Ch. 118, p. 9; CHAVANNES, T'oung Poo, 1907, p. 178. M. CHA- VANNES advances the theory that the Chinese erroneously applied to the ostrich the 126 CHINESE CLAY FIGURES It was styled also "great horse bird." 1 Its resemblance to the camel was emphasized, and hence the name "camel-bird" was formed. Living ostriches were sent to China again in the T'ang period. In 650 Tokhara offered large birds seven feet high, of black color, with feet resembling those of the camel, marching with outspread wings, able to run three hundred li a day, and to swallow iron; they were styled camel- birds. 2 The T'ang artists, accordingly, were in a position to witness and to study live specimens of the bird; and the fact that they really did so leaks out in the realistic high-relief carvings referred to above. But what do we find among the latter-day draughtsmen who en- deavored to illustrate the creature for books? Fig. 1 6 shows the woodcut with which the P&n ts'ao kang mu of Li Shi-che'n is adorned. BRETSCHNEIDER (I. c.) , in a somewhat generous spirit, designated it as "a rude, but tolerably exact drawing of the camel-bird." FORKE 3 holds that this ostrich is pictured like a big goose, but with the feet of a mammal; and he comes far nearer to the truth. Li Shi-che'n, born in K'i chou in the province of Hu-pei, spent his life- name "bird of Parthia" (An-si, Arsak), but that in fact these birds originated from T'iap-chi, that is, Desht Misan or Mesene, where ruled Arabic princes who had all facilities for obtaining ostriches from Arabia. This theory does not seem necessary to me. As already observed by BRETSCHNEIDER (Notes and Queries, Vol. IV, p. 53; and Mediaeval Researches, Vol. I, pp. 144-145), the ostrich is described in Wei shu as a bird indigenous to Persia (compare also Sui shu, Ch. 83, p-. 7 b; Pei shi, Ch. 97, p. 8), and is again mentioned in the T'ang Annals as a Persian bird; there is, on the other hand, the testimony of the Persian authors and of Xenophon (Anabasis, i, 5), who saw the bird on the banks of the Euphrates; and up to the present time, ostriches are met with, though not frequently, in western Asia. HANDCOCK (/. c., p. 25) ob- serves that the ostrich appears in Mesopotamian art at a late period, though in Elam rows of ostriches are found depicte4 on early pottery, closely and inexplicably re- sembling the familiar ostriches on the pre-^dynastic pottery of ancient Egypt; it sometimes, however, assumes a conspicuous position in the embroidery of an Assyrian king's robe, and is found also on a chalcedony seal in Paris. Further references to Assyrian representations are given by O. KELLER (/. c., pp. 172, 594). In ancient Syria, the ostrich is well attested by the interesting description in JOB (xxxix, 13-18), Moses prohibited the flesh of the bird as unclean food, and by reliefs at Hiera- polis of Roman times. It further occurs in the Syrian version of the Physiologus. BREHM (Tierleben, Vol. Ill, p. 692) sums up, "In Asia, the area of the habitat of the ostrich may formerly have been much more extended than at present; but even now, as established by Hartlaub with as much diligence as erudition, it occurs in the deserts of the Euphrates region, especially the Bassida and Dekhena, in all suitable localities of Arabia, and finally in some parts of southern Persia. Vamb6ry even learned that it is still sometimes found on the lower course of the Oxus, in the region of Kungrad (?), and is named there camel or coffer bird." Also in the Encyclopaedia Britannica (Vol. XX, p. 362) it is said, "It is probable that it still lingers in the wastes of Kirwan in eastern Persia, whence examples may occasionally stray north- ward to those of Turkestan, even near the lower Oxus." 1 Ts'ien Han shu, Ch. 96 A, p. 6 b. In this passage the bird is noticed as a native of Parthia, and commented on by Yen Shi-ku. * CHAVANNES, Documents, p. 156. In the period K'ai-yuan (713-741) ostrich eggs were sent from Sogdiana (ibid., p. 136). L. c., p. 138. HISTORY OF THE RHINOCEROS 127 time as magistrate of the district of P'&ig-k'i in the prefecture of T'ung- ch'uan, province of Sze-ch'uan. The chances are that he had never seen the sculptures of ostriches in the mausolea of the T'ang emperors near Li-ts'uan, Shen-si Province; but, be this as it may, his woodcut proves that the T'ang tradition of the representation of the ostrich was wholly unknown to him, and moreover, that he himself had never be- held an ostrich. We have no records to the effect that ostriches were transported to China during the Ming period; and they were then probably known merely by name. Li Shi-che"n's production is simply a reconstruction based on the definitions of the texts (" marching with outspread wings, feet of a camel," etc.); the only exact feature is the two toes, which are mentioned also in the older descriptions of the bird; everything else, notably the crane's head, is absurd, and a naturalist of the type of Bretschneider should have noticed this. In the great cyclopaedia T'u shu tsi ch'Gng, published in 1726, we find a singular illustration of the ostrich, which is reproduced in Fig. 17 as an object-lesson in Chinese psychology. This accomplishment must open every one's eyes: here we plainly see that the illustrator had not the slightest idea of the appearance of an ostrich, but merely endeavored, with appalling result, to outline a sketch of what he imagined the "camel-bird" should look like. He created a combination of a camel and a bird by illustrating the bare words, as they struck his ears, without any recourse to facts and logic; he committed the logical blunder (so common among the Chinese from the days of the Sung period) of confounding a descriptive point of similarity with a feature of reality. All Chinese texts are agreed on the point that the bird is just like a camel, or conveys that impression. This case is most instructive in disclosing the working of the minds of the recent Chinese illustrators, and in exhibiting the value due to their productions. It would not do in the present case to deny that this figure is intended for an ostrich, to define it as a new animal species, a "bird-shaped biped camel " (something like an Avi-camelus bipes), and to conclude that the Chinese term t*o niao does not denote the ostrich. On the contrary, we have to con- clude that illustrations of this character are out and out valueless for our scientific purposes, that definitions of an animal cannot be deduced from them, but that all reasoning on the nature of the respective animal FIG. 16. Ostrich (from Ptn ts'ao kang mu). 128 CHINESE CLAY FIGURES FIG. 17. Alleged Ostrich (from T'u shu tsi ch'ing). HISTORY OF THE RHINOCEROS 129 can be based solely on the texts. 1 The illustrations are posterior in time and mere accessories, and, even if fairly sensible, of sheer secondary importance; in each and every case, however, if utilized as the basis for any far-reaching conclusion, their history, sources, and psychological foundation must be carefully examined. Another impressive lesson to be derived from the case of the ostrich is that China, which by virtue of a widely accepted school opinion appears to us as the classical soil of ultra-conservative perseverance of traditions, is very liable also to lose traditions, and even rather good ones. The excellent ostrich representations of the T'ang have not been perpetuated, but have re- mained as isolated instances. Indeed, they seem to have remained unknown to Chinese artists, archaeologists, and naturalists, and hidden away in seclusion and oblivion until discovered by M. Chavannes. It is this very China unknown to the Chinese, which, as research ad- vances, will become our most attractive subject of study. We referred above (p. 100) to the fact that the ancient illustrations to the Erh ya are lost, and that Kuo-P'o's sketches of the rhinoceros may have been nearer to the truth. In now raising the question whether any representations of the animal are handed down in the ancient monuments of China, we naturally remember the primeval form of writing that mirrors the stage of her primitive culture. The celebrated Catalogue of Bronzes, the Po ku t'u lu, published by Wang Fu in the period Ta-kuan (1107-1111), has preserved to us (Ch. 9, p. 23) two an- cient symbols which are veritable representations of the single-homed rhinoceros se (Fig. 18). They are placed on the ends of a handle of a bronze wine-kettle attributed to the Shang dynasty (B.C. 1766-1154). The explanatory text runs as follows : "The two lateral ears of the vessel are connected by a handle, on which are chased two characters in the shape of a rhinoceros (se). When it is said in the Lun yil that 'a tiger and rhinoceros escape from their cage,' 2 it follows that the rhinoceros is 1 And these must certainly be handled with a critical mind, as, for instance, a glance at the chapter "Ostrich " in the T'u shu tsi ch'tng will convince one. The first extract there given from the Ying yai sheng Ian of 1416 deals with the "fire-bird" of Sumatra, which is the cassowary (see GROENEVELDT, in Miscell. Papers relating to Indo-China, Vol. I, pp. 198, 262). Mo k'o hui si, a work written by P'6ng Ch'feng in the first half of the eleventh century (BRETSCHNEIDER, Bot. Sin., pt. i, p. 174), is quoted as making a contribution to the subject in question, because a bird able to eat iron and stone is mentioned there; this bird, however, called ku-t'o, occurs in Ho-chou, the present Lan-chou fu in Kan-su, is built like an eagle, and over three feet high! Accordingly we here have a wrong association of ideas, and the subject has nothing to do with the ostrich. The editors of the cyclopaedia blindly follow the uncritical example of Li Shi-ch6n, who embodied the same in his notes on the ostrich. Finally, Verbiest's K'un yii t'u shuo is laid under contribution, as he describes the "camel- bird" of South America. This is the Rhea belonging to the Ratite family, but distinguished from the true ostrich by its possession of three toes. 1 LEGGE, Chinese Classics, Vol. I, p. 306; and above, p. 74, note 4. 130 CHINESE CLAY FIGURES FIG. 18. Single-Horned Rhinoceros on a Bronze Kettle attributed to Sbang Period (from Po ku t'u lu, edition of 1603). FIG. 19. Bushman Sketches of Rhinoceros (from E. Cartailhac and H. Breuil, La caverne d'Altamira, pp. 180, 189). FIG. 20. Red Drawing of a Two-Horned Rhinoceros, from Font-de-Gaume (after Capitan and Breuil). HISTORY OF THE RHINOCEROS not a tame animal. Indeed, it inflicts injury on man; and for this reason the ancients availed themselves of it to fine a person a cup of wine, which is expressed by the phrase 'to raise the goblet of rhinoceros-horn.' l This goblet receives its name from the rhinoceros, and so it is proper also that there should be wine-kettles with the emblem of the rhinoceros. On the two ends of the handle of this vessel is pictured a rhinoceros with head and body complete, the latter having the shape of a glutton (f ao t'ie). This certainly indicates that it symbolizes a warning. In this manner all vessels were decorated during the Shang dynasty, and it is by such symbolic forms that they are distinguished from those of the Chou." Whatever the rough character of these two sketches transmitted by the Po ku t'u lu may be, 2 the single-horned rhinoceros is here clearly outlined with a naive and refreshing realism, such as could be spontaneously produced only by the hand of primitive man, who with a few forceful outlines recorded his actual ex- perience of the animal. Here we do not face the narrow-breasted academic and philological construction of the scholars of the Sung period, but the direct and vigorous impression of the strong-minded hunter of past ages, who was formed of the same stuff as the Bushman of southern Africa and palaeolithic man living in the caves of Spain and France. No bridge spans the chasm yawning between the Shang and Sung productions. The Shang rhinoceros breathes the same spirit as its companions on the rock paintings of the Bushman (Fig. 19), and in the palaeolithic cave of Font-de-Gaume in France (Fig. 20). The general form of the FIG. 21. Inscription on Bronze Kettle attributed to Shang Period, showing Pictorial Form of Sacrificial Bull (from Po ku fu IK). 1 Quotation from Shi king (see LEGGE, Chinese Classics, Vol. IV, p. 233). The rhinoceros-horn goblets are discussed below, p. 167. * Another cruder and more conventionalized symbol of the rhinoceros se, in which, however, the single horn is duly accentuated, is figured in the same work (Ch. i, p. 25 b), as occurring in the inscription on a round tripod vessel (ting) attributed to the Shang period. 132 CHINESE CLAY FIGURES animal is well grasped in the Chinese sketch, and the shape of the horn is correctly outlined. For the sake of comparison, and in order to show that the primitive Chinese man knew very well how to discriminate between a rhinoceros and an ox, the contemporaneous symbol for the sacrificial bull (hi niu), and designs of recumbent oxen (explained as such in the Po ku t*u lu) on the lid of a bronze vessel, are here added (Figs. 21 and 22). We arrive at the result, which will be corroborated by other evidence, that in the earliest stage of Chinese culture the animal se was the single-horned rhinoceros. 1 Before plunging into the Chinese sources relative to the rhinoceros, it will be well to re- member that all living species of rhinoceros are by most naturalists referred to a single genus, which is found living in Africa and south-eastern Asia, while formerly it was widely distributed over the entire Old World (with the exception of Australasia), ranging as far north as Siberia. 2 Three species exist in Asia, Rhinoceros unicornis, the great one-horned rhinoceros, at the present day almost entirely restricted to the Assam plain, but formerly extensively dis- tributed over India; 3 Rhinoceros sundaicus, called also the Javan rhino- ceros, the smaller one-horned rhinoceros, found in parts of eastern Bengal (the Bengal Sunderbans near Calcutta), in Assam, throughout Burma, the Malay Peninsula, Sumatra, Java, and Borneo; and Rhino- ceros (or Dicerorhinus) sumatrensis, the Asiatic two-horned rhinoceros, rare in Assam, ranging from there to Burma, Siam, the Malay Peninsula, 1 The later developments of the early forms of the symbol se may be viewed by those who are debarred from Chinese sources in F. H. CHALFANT, Early Chinese Writing, Plate II, No. 17 (Memoirs Carnegie Museum, Vol. IV, No. I, Pittsburgh, 1906). According to a communication of the late Mr. CHALFANT (Dec. 18, 1913), the ancient bone inscriptions twice reveal a character which may be identified with the word se, while the character for si has not yet been traced in them. 1 Hornless species formerly occurred in North America, where the group has existed since the latter part of the Eocene period. 'Chiefly after W. T. BLANFORD, The Fauna of British India; Mammalia, pp. 471-477- PIG. 22. Lid of Bronze Kettle attributed to Shang Period, with Designs of Recumbent Oxen (from Po ku t'u lu). HISTORY OF THE RHINOCEROS 133 Sumatra, and Borneo. 1 Judging from this remarkable case of dis- continuous distribution 2 and from historical records, there is every reason to believe that in ancient times this animal, like all the large mammals now facing extinction, was distributed over a much larger geographical area; and this fact is fully confirmed by palaeontological research, as well as by the records of the Chinese. For the purpose of our inquiry it should be particularly borne in mind that it is in the territory of Assam where we meet the three species together. "The Imperial Gazetteer of India " 3 states, in the chapter on Assam, "Rhinoceros are of three kinds: the large variety (unicornis), which lives in the swamps that fringe the Brahmaputra; the smaller variety (sondaicus), which is occasionally met with in the same locality; and the small two-horned rhinoceros (sumatrensis) , which is now and again seen in the hills south of the Surma Valley, though its ordinary habitat is Sumatra, Borneo, and the Malay Peninsula." Assam is inhabited by numerous tribes, a large portion of which ranges among the Indo-Chinese family. What now holds good for Assam, as will be recognized from a survey of Chinese sources, two millenniums and more ago was valid for the south-western and southern parts of China, the Tibeto-Chinese borderlands, and Indo-China in its total range; in short, the historical fact will be established that in the past the rhinoceros in its two main varieties, the single-horned and two-horned, had occupied the whole territory of south-eastern Asia. The greater part of the knowledge possessed by the Chinese in re- gard to the rhinoceros has been digested by Li Shi-che'n in his materia medica Pin ts'ao kang mu (Ch. 51 A, p. 5) completed in 1578 after twenty- six years' labor. He first quotes a number of authors beginning from the fifth century, and then sums up the argument in his own words. This discourse is also of value for zoogeography, in that it contributes materially to the possibility of reconstructing the early habitats of the rhinoceros in China. The text of this work is here translated in extenso, but rectified and supplemented from the materia medica of the Sung period, the Cheng lei pSn ts'ao, first printed in uo8. 4 l Al-BgrunI (973-1048) states that the rhinoceros existed in large numbers in India, more particularly about the Ganges (SACHAU, /. c.. Vol. I, p. 203). In the sixteenth century it occurred in the western Himalaya and also in the forests near Peshawar (YutE and BURNELL, Hobson-Jobson, p. 762). LINSCHOTEN found it in great numbers in Bengal (ibid., p. i); so also GARCIA AB HORTO (/. c., p. 66): multos in Cambaya Bengala finitima, et Patane inveniri tradunt. ABUL FAZL ALL AMI ( I 55 I ~i6o2), in his A in I Akbari written in 1597 (translation of H. S. JARRETT, Vol. II, p. 281, Calcutta, 1891), mentions the occurrence of the rhinoceros among the game in the Sarkar of Sambal (near Delhi). 1 Compare E. HELLER, The White Rhinoceros, p. 39. 1 Vol. VI, p. 20 (Oxford, 1908). * See T'oung Pao, 1913, p. 351. 134 CHINESE CLAY FIGURES KB flu Kft I mil IE IB IB a it HISTORY OF THE RHINOCEROS 135 Other texts of importance apt to throw light on the matter have been added from the T'u shu tsi ch'tng and several other works, so that the result is a fairly complete digest of what Chinese authors of the post- Christian era have to say about the rhinoceros and its horn. After this survey, we shall turn to the times of early antiquity, and discuss the subject in the light of such information as has been handed down to us from those days. Li Shi-che'n opens his discourse on the rhinoceros with the explana- tion of the name. "The symbol for the word si still has in the seal character chuan w&n the form of a pictograph, 1 and is the name for the female rhinoceros. The se is styled also 'sand rhinoceros' (sha si). The Erh ya i 2 says that the words se and tse (female) approach each other in sound like the two words ku ('ram,' No. 6226) and ku ('male'). In general, si and se are one and the same. The ancients were fond of saying se, the people of subsequent times inclined toward the word si. In the northern dialects the word se prevails, in the southern dialects the predilection is for si. This is the difference between the two. In Sanskrit literature the rhinoceros is called khadga." 3 Li Shi-che'n then proceeds to quote the ancient work Pie lu* which makes the following important statement in regard to the former localities where the rhinoceros occurred: "The habitat of the rhinoceros 1 This is indeed the case in the Shuo wSn (see p. 92). The names of the rhinoceros and the various kinds of its horn are here reproduced from T'u shu tsi ch'Sng (p. 134). 1 An appendix to the Erh ya by Lo Yuan of the twelfth century (BRETSCHNEIDER, Bot. Sin., pt. i, p. 37). 'Written with Nos. 1456 and 1558 (k'et-ga); compare EITEL, Hand-book of Chinese Buddhism, p. 76. (Other Sanskrit words for "rhinoceros" are ganfa, gandaka, gantfdnga.) The work Sheng shui yen fan lu, written by Wang P'i-chi about the end of the eleventh century (WvLiE, Notes, p. 195), seems to be the first to impart this Sanskrit name (see the Chinese text opposite) ; it further gives a Sanskrit word for the horn in the Chinese transcription pi-sha-na corresponding to Sanskrit vishdna ("horn"). The latter and the word khatfga were among the first Sanskrit words in Chinese recognized by Abel Re'musat (see S. JULIEN, M6thode, p. 3). * The Pie lu is not identical with the Ming i pie lu, as first stated by BRET- SCHNEIDER (Bot. Sin., pt. i, p. 42), but later rectified by him (in pt. 3, p. 2). It is an independent work, which must have existed before the time of T'ao Hung-king, and which was known to the latter and commented on by him. This is quite clear in the present case, as Li Shi-chen first introduces the Pie lu, and then proceeds, "T'ao Hung- king says." And since the latter starts with the phrase "at present," it is apparent that he had the words of the Pie lu before his eyes, and gave his definition in distinc- tion from the older work. This is also proved by the text of the Cheng lei ptn ts'ao published in 1108 by the physician T'ang Shen-wei (edition of 1523, Ch. 17, fol. 21), where the two quotations are separated and marked by type of different size. As in Bretschneider's opinion nearly all the geographical names occurring in the Pie lu refer to the Ts'in (third century B.C.) or Han periods, although some of them can be traced to the Chou dynasty (B.C. 1122-249), the above passage surely relates to a time antedating our era by several centuries; and it goes without saying, that as a matter of fact, in the age of the Chou and at a far earlier date, the two-homed rhinoceros must have been a live citizen in the south-western parts of China. 136 CHINESE CLAY FIGURES (si) is in the mountains and valleys of Yung-ch'ang and in Yi-chou; 1 Yung-ch'ang is the southern part of the present country of Tien (Yun- nan). " 2 The next author invoked by Li Shi-che"n is T'ao Hung-king (452- 536), a celebrated adept of Taoism and a distinguished physician, author of the Ming i pie lu, a treatise on materia medica. 3 He states, "At present the rhinoceros (si) inhabits the distant mountains of Wu-ling, 4 Kiao-chou, 6 and Ning-chou. 6 It has two horns; the horn on the forehead is the one used in fighting. 7 There is a kind of rhinoceros styled 'communicating with the sky' (t'ung t'ien), whose horn is in- tersected by a white vein running clear through from the base to the tip; the night dew does not moisten it. It is employed as a remedy, whereby its wonderful properties are tested. In the opinion of some, this is the horn of the water-rhinoceros, which is produced in the water. 8 The Annals of the Han Dynasty speak of the horn of 'the rhinoceros frightening fowl ' (hiai ki si) : when it was placed in the rice that served as food for the chickens, they were all scared and did not dare to peck; 1 PLAYFAIR, The Cities and Towns of China, No. 8596 (2d ed., No. 7527, i). In the Han period, Yi-chou was the name of a province occupying the territory of the present province of Sze-ch\tan, a part of Kuei-chou and Yun-nan (BRETSCHNEIDER, Bot. Sin., pt. 3, p. 565), while the southern part of Yun-nan is understood by the designation Yung-ch'ang. The Pie lu, accordingly, locates in south-western China the rhinoceros si, which, as follows from the comment of T'ao Hung-king, is the two-horned species. 2 This last clause is not contained in the text of the Cheng lei pin ts'ao, and is doubtless a later comment, presumably derived from T'ao Hung-king's edition of the Pen ts'ao king, which is listed in the Catalogue of the Sui Dynasty, and according to Bretschneider's supposition, embraced likewise the text of the Pie lu. 8 His biography is in Nan ski (Ch. 76, p. 4 b) and Liang shu (Ch. 51, p. 12). 4 PLAYFAIR, No. 8112 (2d ed., No. 7080): district forming the prefectural city of Ch'ang-t6, Hu-nan Province. 6 Northern part of the present Tonking (see HIRTH and ROCKHILL, Chau Ju-kua, p. 46). PLAYFAIR, No. 5239, 2 (4672, 2): in Lin-an fu, Yun-nan Province. Under the Tsin it was a province comprising Yun-nan and part of Kuei-chou (compare Hua yang kuo chi, Ch. 4, p. I, ed. of Han Wei ts'ung shu). 7 Thus the two-horned (so-called Sumatran) rhinoceros is here clearly mentioned. 8 The rhinoceros is fond of spending the hot hours of the day immersed in water, and thence the Chinese designation "water-rhinoceros" may take its origin. In this position particularly, the animal calls to mind the water-buffalo. In ancient times it was therefore dreaded as being able to overturn boats, which is quite believable; and soldiers crossing a river were encouraged to prompt action by their commander shouting the name of the animal (CHAVANNES, Les M6moires historiques de Se-ma Ts'ien, Vol. I, p. 225, Vol. IV, p. 37; FORKE, Lun-Heng, pt. II, p. 322; according to FORKE, the reading of the text is ts'ang kuang, but as quoted in T'u shu tsi ch'eng and P'ei v>2n yiinfu it is ts'ang se, as in Se-ma Ts'ien). The water-rhinoceros (shui si) is mentioned in Kuang chou ki (see BRETSCHNEIDER, Bot. Sin., pt. i, No. 377) as occurring in the open sea off the district of P'ing-ting, resembling an ox, emitting light when coming out of, or descending into, the water, and breaking a way through the water (quoted in T'u shu tsi ch'eng). HISTORY OF THE RHINOCEROS 137 when it was placed on the roof of a house, the birds did not dare to assemble there. 1 There is also the horn of the female rhinoceros, which is very long, with patterns resembling those of the male, but it is not fit to enter the pharmacopoeia." * 1 The allusion to the hiai ki si occurs in Ch. 108 of Hou Han shu (compare CHA- VANNES, Les pays d'Occident d'apres le Heou Han Chou, Toung Pao, 1907, p. 182; and HIRTH, China and the Roman Orient, p. 79), where this kind of horn is ascribed to the country of Ta Ts'in (the Roman Orient). The legend given in explanation as above is derived from the famous Taoist writer Ko Hung, who died about 330 A.D. ; and it is not accidental that the Taoist T'ao Hung-king here copies his older colleague, for the legend is plainly Taoistic in character. It is quoted in the commentary to Hou Han shu, but not in the text of the Annals. The view of Hirth, that it has arisen in consequence of a false etymology based on the Chinese characters transcribing a foreign word, seems to me unfounded. First, as Chavannes remarks, the foreign word supposed to be hidden in hiai-ki has not yet been discovered, and in all probabil- ity does not exist. Second, as will be seen from P'ei wen yiinfu (Ch. 8, p. 87 b), the term hiai ki si does not occur in Hou Han shu for the first time, but is noted as early as the Chan kuo ts'e at the time of Chang I, who died in B.C. 310, when the King of Ch'u despatched a hundred chariots to present to the King of Ts'in fowl-scaring rhinoceros-horns and jade disks resplendent at night (ye kuang pi). It is certainly somewhat striking to meet here these two names, which are identical with those in Hou Han shu, and occur there close together; and it cannot be denied that the passage of Chan kuo ts'e might be an interpolation. Huai-nan-tse, who died in B.C. 122, alludes to a rhinoceros-horn frightening foxes (5* kio hiai hu, quoted in P'ei wen yiin /, /. c., p. 89 a, "when placed in the lair of a fox, the fox does not dare return"), wftich is a case analogous in word and matter to the fowl-frightening horn. These notions must be taken in connection with the other legends regarding the rhinoceros, which all seem to spring from indigenous Taoist lore. The text of Ko Hung, as quoted in P'ei wen yiin fu and translated by Hirth and Chavannes, is fuller than cited above in the Pen ts'ao, while the final clause in regard to placing the horn on the roof does not occur in Ko Hung. The latter links the hiai ki si with the t'ung t'ien, which Hirth and Chavannes translate "communicating with Heaven." This is cer- tainly all right; but I prefer to avoid this term, because it may give rise to mis- understandings, as we are wont to think of Heaven as the great cosmic deity. A com- parative study of all passages concerned renders it clear that the rhinoceros is not associated with spiritual, but with material heaven; that is, the sky. It is the stars of the sky which are supposed to be reflected in the veins of the horn. This means that the designs of the horn gave the impetus to the conception of connecting the rhinoceros with the phenomena of the sky, again a thoroughly Taoistic idea, in which no trace of an outside influence can be discovered. Father ZOTTOLI (Cursus litteraturae sinicae, new ed., Vol. I, p. 301 ) renders the term t'ung t'ien si tai by " pene- trantis coelum rhinocerotis cingulum." Chao Ju-kua (HiRTH's and ROCKHILL'S translation, p. 103) attributes hiai ki si or t'ung t'ien si also to Baghdad (but I see no reason why these words should denote there a precious stone, instead of rhinoceros- horn). On p. 1 08 (note 10) the two authors represent the matter as though this refer- ence might occur in Ling-wai tai ta, but in fact it is not there (Ch. 3, p. I b); it must therefore be due to Chao Ju-kua, who seems to indulge in a literary reminiscence taken from Hou Han shu. The passage, accordingly, affords no evidence for a trade in rhino- ceros-horns from Baghdad to China, which per se is not very likely. In the illustra- tions to the Feng shen yen i (ed. of Tsi ch'eng t'u shu, p. 9, Shanghai, 1908), T'ung t'ien kiao chu (see W. GRUBE, Die Metamorphosen der Gotter, p. 652) is seated astride a rhinoceros (outlined as a bull with a single striped horn), apparently because his name Tung t'ien has been identified with t'ung t'ien si. 1 There are several additions to this text as edited in the Cheng lei pen ts'ao, the most interesting of which is that "only the living horns are excellent." This means the horn of a live animal slain in the chase, which was believed to be superior in qual- ity to a horn cast off and accidentally found (compare HIRTH and ROCKHILL, Chau Ju-kua, p. 233). Similar beliefs prevailed in regard to ivory. That coming from the tusk of an elephant killed by means of a pike was considered the best; next in quality 138 CHINESE CLAY FIGURES Li Shi-che'n does not refer to Ko Hung, the famous Taoist adept of the fourth century, 1 who is the first author to impart a fantastic account in regard to rhinoceros-horn. He is likewise the first to set forth its quality of detecting poison. His text is here translated, as given in T'u shu tsi ch'eng. 2 "Mr. Che"ng 3 once obtained a genuine rhinoceros-horn of the kind 'communicating with the sky,' three inches long, the upper portion being carved into the form of a fish. When a man carries such a piece in his mouth and descends into the water, the water will give way for him and leave a vacant space three feet square, so that he has a chance to breathe in the water. 4 The horn 'communicating with the sky* has a single red vein like a silk string running from the base to the tip. When a horn filled with rice is placed among a flock of chickens, the chickens want to peck the grains. Scarcely have they approached the horn to within an inch when they are taken aback and withdraw. Hence the people of the south designate the horn 'communicating with the sky' by the name 'fowl-frightening horn.' When such a horn is placed on a heap of grain, the birds do not dare assemble there. Enveloped by a thick fog or exposed to the night dew, when placed in a courtyard, the horn does not contract humidity. The rhinoceros (si) is a wild animal living in the deep mountain-forests. During dark nights its horn emits a brilliant light like torch-fire. The horn is a safe guide to tell the presence of poison : when poisonous medicines of liquid form are stirred with a horn, a white foam will bubble up, and no other test is necessary; when non-poisonous substances are stirred with it, no foam will rise. In this manner the presence of poison can be ascertained. When on a journey in foreign countries, or in places where contagion from ku was the ivory of an animal which was found shortly after it had died a natural death; least esteemed was that discovered in mountains many years after the animal's death (PELLIOT, Bulletin de I'Ecole frang aise d'Extrdme-Orient, Vol. II, 1902, p. 166). In Siam, the rhinoceros is still killed with bamboo pikes hardened in the fire and thrust into its jaws and down the throat, as described by Bishop PALLEGOIX (Descrip- tion du royaume Thai ou Siam, Vol. I, p. 75, Paris, 1854). 1 He died in 330 A.D. at the age of eighty-one; see GILES (Biographical Dic- tionary, p. 372); MAYERS (Chinese Reader's Manual, p. 86); BRETSCHNEIDER (Bot. Sin., pt. I, p. 42); and PELLIOT (Journal asiatique, 1912, Juillet-Aout, p. 145). * Chapter on Rhinoceros (hui k'ao, p. 3), introduced by the author's literary name Pao-p'u-tse, and the title of his work T&ng sht p'ien, which is not included in the Taoist Canon. 1 Presumably Ch6ng Se-yuan, a relative and spiritual predecessor of Ko Hung (L. WIEGER, Taoisme, Vol. I, Le canon, p. 16; PELLIOT, /. c., p. 146). 4 It is interesting to note that this belief is still upheld in the modern folk-lore of Annam: "Celui qui peut se procurer une corne de rhinoceros et la sculpte en forme de poisson, s'il la met entre ses dents, peut descendre sans danger, comme le rhi- nooSros ou le poisson, tout au fond de 1'eau" (P. GIRAN, Magie et Religion Annamites, p. 104, Paris, 1912). HISTORY OF THE RHINOCEROS 139 poison l threatens, a man takes his meals in other people's houses, he first ought to stir his food with a rhinoceros-horn. When a man hit by a poisonous arrow is on the verge of dying, and his wound is slightly touched with a rhinoceros-horn, foam will come forth from his wound, and he will feel relief. 2 This property of the horn 'communicating with the sky' of neutralizing poison is accounted for by the fact that the animal, while alive, particularly feeds on poisonous plants and trees provided with thorns and brambles, 3 while it shuns all soft and smooth vegetal matter. Annually one shedding of its horn takes place in the mountains, and people find horns scattered about among the rocks; 4 in this case, however, they must deposit there, in the place of the real one, another horn carved from wood, identical with that one in color, veins, and shape. Then the rhinoceros remains unaware of the theft. In the following year it moves to another place to shed its horn. 8 Other kinds of rhinoceros-horn also are capable of neutralizing poison, without having, however, the wonderful power of the t'ung-t'ien variety." Su Kung, the editor of the T'ang sin p$n ts'ao (the revised edition of the materia medica of the T'ang dynasty) states as follows: "The tse (No. 12,325) is the female rhinoceros. The patterns on its horn are smooth, spotted, white, and clearly differentiated. It is ordinarily called the 'spotted rhinoceros' (pan si). It is highly esteemed in pre- 1 See Toung Poo, 1913, p. 322. 1 The belief that the horn will check the effects of poisoned arrows is repeated in the Pei hu lu, written by Tuan Kung-lu around 875 in the T'ang period (PELLIOT, Bul- letin de l'Ecolefran$aise, Vol. IX, 1909, p. 223). The notes of this book regarding the horn are all based on the text of Ko Hung; instead of t'ung t'ien si, the term t'ung si is employed. 1 The animal feeds, indeed, on herbage, shrubs, and leaves of trees. 4 The supposition of the rhinoceros shedding its horn regularly has not been ascer- tained by our zoologists; but it is not very probable that it does so, nor have the Chi- nese made the actual observation. It is clear that their conclusion is merely based on the circumstantial evidence of detached horns occasionally found and picked up in the wilderness, which suggested to them the notion of a natural process similar to the shedding of cervine antlers. ' A similar story is told in regard to the elephant by Chen Kuan, who wrote two treatises on the medical virtues of drugs, and who died in the first part of the seventh century (BRETSCHNEIDER, Bot. Sin., pt. i, p. 44): "The elephant, whenever it sheds its tusks, itself buries them. The people of K'un-lun make wooden tusks, stealthily exchange them, and take the real ones away." K'un-lun is the Chinese designation for the Malayan tribes of Malacca, and was extended to Negrito, Papua, and the negroes of Africa (see HIRTH and ROCKHILL, Chau Ju-kua, p. 32). In this connec- tion we should remember also the words of PLINY (Nat. hist., vm, 3, 7), that the elephants, when their tusks have fallen out either accidentally or from old age, bury them in the ground (quam ob rem deciduos casu aliquo vel senecta defodiunt). It is not impossible that the great quantity of fossil ivory mentioned as early as by THEOPHRAST (De lapidibus 37, Opera ed. F. WIMMER, p. 345; compare the interesting notes of L. DE LAUNAY, Mine'ralogie des anciens, Vol. I, pp. 387-390, Bruxelles, 1803) may have given rise to this notion. 140 CHINESE CLAY FIGURES scriptions, but is not such an efficient remedy as the horn of the male rhinoceros." 1 Ch'6n Ts'ang-k'i, who lived in the first half of the eighth century, states in his work Pin ts'ao shi i ("Omissions in Previous Works on Materia Medica") as follows: "There are not two kinds of the rhinoce- ros, called the land and water animal. This distinction merely refers to finer and coarser qualities of horns. 2 As to the rhinoceros 'com- municating with the sky,' the horn on its skull elongates into a point after a thousand years. It is then adorned, from one end to the other, with white stars, and can exhale a vapor penetrating the sky; in this manner it can communicate with the spirits, 3 break the water, and frighten fowl. Hence the epithet 'communicating with the sky' is bestowed on it. Pao-p'u-tse 4 says, 'When such a rhinoceros-horn is carved into the shape of a fish, and one holding this in his mouth de- scends into water, a passage three feet wide will open in the water.' " 8 Su Sung, author of the T'u king -pin ts'ao, published by imperial order in the age of the Sung dynasty, has the following: " Of rhinoceros- horn, that coming from the regions of the Southern Sea (Nan hai) takes the first place; that from K'ien and Shu 6 ranks next. The rhinoceros resembles the water-buffalo, has the head of a pig, a big paunch, short legs, the feet being similar to those of the elephant and having three toes. It is black in color, and has prickles on its tongue. It is fond of eating thorny brambles. 7 Three hairs grow from each pore in its skin, 1 Li Shi-ch&n's text exactly agrees with that given in the Ch&ng lei p$n ts'ao. It is an interesting coincidence that the horn of the female rhinoceros (tse si kio) is men- tioned in the Annals of the T'ang Dynasty (T'ang shu, Ch. 40, p. 6 b) as the tribute sent from the district of Si-p'ing in Shen chou, the present territory of Si-ning in Kan-su. The Annals therefore confirm the statement of the contemporaneous P2n ts'ao. * It will be seen below that Li Shi-ch6n does not share this opinion. 1 The same paragraph is found in Li Shi, the author of the Su po wu chi (Ch. 10, p. 8 b; ed. of Pai hai), ascribed by tradition to the T'ang period, but in fact coming down from the Sung. He interprets the expression t'ung t'ien by the words, " It is capable of communicating with the spirits" (n&ng t'ung sh&n). According to him, "the horn communicating with the sky" is a thousand years old, long and pointed, overstrewn with white stars, the tip emitting a vapor. 4 Surname of Ko Hung, a famous Taoist writer, who died at the age of eighty-one about 330 A.D. (see p. 138). * The text in the Chtng lei pin ts'ao is somewhat fuller. It opens by saying that the flesh of the rhinoceros cures all poisons, especially poisoning caused by the bites of snakes and mammals. On Java bits of the horn are considered as an infallible antidote against snake-bites (P. J. VETH, Java, Vol. Ill, p. 289). At the close of Ch'fin Ts'ang-k'i's text it is added that the horn is called also nu kio (literally, "slave horn") and shi kio ("the horn, with which the animal feeds"); the word nu seems to be the transcription of a word from a non-Chinese language. * Ancient designations for the present territory of the provinces of Kuei-chou and Sze-ch'uan. 7 The entire definition, except the "prickles on the tongue," is derived from Kuo P'o (see p. 93). MARCO POLO (ed. of YULE and CORDIER, Vol. II, p. 285), speaking of HISTORY OF THE RHINOCEROS 141 as in swine. There are one-horned, two-horned, and three-horned ones." l the rhinoceros on Java, says, "They do no mischief, however, with the horn, but with the tongue alone; for this is covered all over with long and strong prickles [and when savage with any one they crush him under their knees and then rasp him with their tongue]." YULE comments that the belief in the formidable nature of the tongue of the rhinoceros is very old and widespread, though he can find no foundation for it other than the rough appearance of the organ. Dr. PARSONS (p. 9 in the pamphlet quoted above, p. 83) observes, "As to the tongue of the rhinoceros, the scribes assure us that it is so rugged that it can lick off with it the flesh from the bones of a man, but the tongue of the live animal examined by me is as soft and mild as that of a calf; whether it will grow rougher with the advancing age of the animal, I am unable to say." It is easy to see how the fable of the prickly tongue arose. The animal mainly feeds on herbage, and the alleged or real observation of its inclination for brambles led to the conclusion that its tongue must be thorn-proof and prickly. A similar belief seems to obtain in Siam: "On dit que ce monstrueux quadrupede fait ses d^lices des Opines de bambou" (Mgr. PALLEGOIX, Description du royaume Thai ou Siam, Vol. I, p. 156, Paris, 1854). 1 Now follows in the P2n ts'ao the quotation from the Erh ya translated above (p. 93). The text then following in the PSn ts'ao is purported to be a quotation from Ling piao lu i; but it is in fact abridged, and intermingled with extracts from Yu yang tsa tsu. For this reason I have abandoned at this point the text of the PSn ts'ao, and given separately translations of the two documents, as they are published in T'u shu tsi ch'eng (Chapter on Rhinoceros, hui k'ao, p. 4). In evidence of my statement, the text of the Ptn ts'ao here follows ; the main share in the confusion will probably be due to Su Sung, not to Li Shi-chen. " The Ling piao lu i by Liu Sun (of the T'ang period) says, 'The rhinoceros has two horns: the one on the forehead is called se si, the other, on the nose, is called hu mao si. The male rhinoceros also has two horns both of which are comprised under the name mao si ('hairy rhinoceros'). At present people uphold the opinion that it has but a single horn. These two kinds of horn are provided with grain patterns, and their price largely depends upon the finer or coarser qualities of these designs. The most expensive is the horn with floral designs of the rhinoceros 'communicating with the sky.' The animals with such horns dislike their own shadow, and constantly drink muddy water in order to avoid beholding their reflection. High-grade horns bear likenesses of all things. Some attribute the qualities of the t'ung t'ien horn to a pathological cause, but the natural reason cannot be ascertained. The term tao ch'a means that one half of the lines pass through in the direction downward; the term chtng ch'a means that one half 01 the lines pass through in the direction upward; the term yao ku ch'a means that the lines are inter- rupted in the middle, and do not pass through. Such-like are a great many. The Po-se designate ivory as po-ngan, and rhinoceros-horn as hei-ngan, words difficult to distinguish. The largest rhinoceros-horn is that of the to-lo-si, a single horn of which weighs from seven to eight catties. This is identified with the horn on the forehead of the male rhinoceros. It has numerous decorations conveying the impression of scattered beans. If the specks are deep in color, the horn is suitable to be made into plaques for girdle-ornaments; if the specks are scattered here and there, and light in color, the horn can be made only into bowls and dishes. In the opinion of some, the animal called se is the female of the si. [It resembles the water-buffalo, and is of dark color. Its hide is so hard and thick that it can be worked into armor.] I do not know whether this is the case or not." (There is here a confusion in Li Shi-chen's text. The passage enclosed in brackets does not occur in the text of the Ching lei pSn ts'ao, where it runs, "In the opinion of some, the animal called se is the female of the si; I do not know whether this is the case or not." The rest is evidently interpolated, and is derived from the Shuo win and its commentaries; at all events, it cannot be ascribed to Su Sung.) "Wu Shi-kao, a physician of the T'ang period, tells the fol- lowing story: 'The people near the sea, intent on capturing a rhinoceros, proceed by erecting on a mountain-path many structures of decayed timber, something like a stable for swine or sheep. As the front legs of the rhinoceros are straight, without joints, it is in the habit of sleeping by leaning against the trunk of a tree. The rotten 142 CHINESE CLAY FIGURES The Ling piao lu i ki l says, " The rhinoceros, in general, resembles an ox in form. Its hoofs and feet are like those of the elephant. It has a double armor and two horns. The one on the forehead is styled se si; the other, on the nose, which is comparatively smaller, is termed hu mao si. 2 The designs and spots in the anterior horn are small; many have extraordinary patterns. The male rhinoceros likewise has two horns, both of which are designated mao si ('hairy rhinoceros'), and are provided with grain patterns. 3 They are capable of being worked into plaques for girdles. 4 Among a large number of rhinoceros-horns there timber will suddenly break down, and the animal will topple in front without being able for a long time to rise. Then they attack and kill it.' " The conclusion is translated above in the text. 1 In the Pdn ts'ao, and otherwise, usually styled Ling piao lu i. According to BRET- SCHNEIDER (Bot. Sin., pt. i, p. 170), it is an account of the natural productions of China by Liu Sun of the T'ang dynasty. 2 HIRTH and ROCKHILL (Chau Ju-kua, p. 233), briefly alluding to this text, under- stand the terms se si and hu mao si as two different varieties of the rhinoceros. This point of view seems to me inadmissible, as Liu Sun distinctly speaks of the two-horned variety only, and then goes on to specify the two horns in the same animal, which differing in size and shape are, from a commercial and industrial standpoint, of dif- ferent value. The term Hu mao ('cap of the Hu ' ; the Hu in general designate peoples of Central Asia, Turks and Iranians) is a very appropriate designation for the anterior horn of this species, which is a low, flat, roundish knob, and indeed resembles a small skull-cap. In the Ming kung shi (Ch. 4, p. 8; new edition in movable types, 1910, in 8 chs.), a most interesting description of the life at the Court of the Ming dynasty (compare HIRTH, T'oung Pao, Vol. VI, 1895, p. 440), this cap is explained as coming down from the T'ang dynasty, and as having been used by the heir-apparent of the Ming ; it was made from sable and ermine skins, and worn in the winter on hunting- expeditions to keep the ears warm. It is mentioned in T'ang shu, Ch. 24, p. 8 (and presumably in other passages). 8 Li Shi-chSn (p. 150) expands this theme. Fang I-chi, who graduated in 1640, in his Wu li siao shi (Ch. 8, p. 20 b), states that only the rhinoceros-horn of Siam has grain patterns, while they are absent in the hairy (that is, the double-horned) rhi- noceros of Annam, which has flower-like and spotted designs. 4 In the Treasure-House of Nara in Japan are preserved objects carved from rhinoceros-horn coming down from the T'ang period, as leather belts with horn plaques, drinking-cups, Ju-i, and back-scratchers. The girdles studded with plaques carved from the horn seem to make their appearance in China under the T'ang dynasty; the assertion of BUSHELL (Chinese Art, Vol. I, p. 119) that they were the "official" girdles of the dynasty does not seem to be justified: at least, they are not enumerated in the class of official girdles, but seem to have been restricted to the use of princesses (compare the account of Tu yang tsa pien, translated below, p. 152). Interesting texts bearing on rhinoceros-horn girdles are communicated in T'u shu tsi ch'Sng (Chapter on Girdles, tai p'ei, ki shi, p. 9 b). Such girdles were made also in Champa: the Sung Annals (Sung shi, Ch. 489, p. 2) relate a tribute sent from there in the period Hien-tl (954-962) of the Hou Chou dynasty; it was local products including rhinoceros-horn girdles with plaques carved in the form of cloud-dragons. A rhinoceros-horn girdle sent from the Court of the Sung to that of the Khitan is men- tioned in Liao shi (Ch. 10, p. i). Under the Kin dynasty (1115-1234) the materials employed for official costume were ranked in the order jade, gold, rhinoceros-horn, ivory (Kin shi, Ch. 34, 3, p. 7). The emperor wore a hat-pin of rhinoceros-horn, and a girdle of black horn (wu si tai) ; the imperial saddle was decorated with gold, silver, rhinoceros-horn, and ivory. Officials of the second rank and higher were en- titled to a girdle of the t'ung si horn; those of the third rank, to a girdle of the hua si horn ; the rest, to plain rhinoceros-horn girdles (ibid., Ch. 43). They were in vogue also HISTORY OF THE RHINOCEROS 143 are few in which the lines pass through from one end to the other. These are pointed, and their designs are large and numerous. Those with small designs are styled too ch'a t'ung. 1 These two kinds are called also 'bottomless jade cups.' 2 If there is not sufficient space for the lines to pass through, and the white and black designs are equally distributed, then the price is considerably increased, and the horn will become the treasure of numberless generations. When I lived at P'an-yii, 3 1 made a thorough examination of what is current there con- cerning rhinoceros-horn. There is, further, the to-lo-si, the largest among the rhinoceros-horns, which may reach seven catties in weight. 4 This is the horn on the forehead of the male rhinoceros, which has numerous designs in the interior conveying the impression of scattered beans. If the stripes are deep in color, the horn is capable of being made into girdle-plaques and implements; if the stripes are dispersed and light in color, the horn may be employed to advantage for the making of cups, at the Court of the Ming emperors (Ta Ming hui lien, Ch. 5, p. 30), and were allowed to alternate with tortoise-shell girdles (Ming kung shi by Liu Jo-yu, Ch. 4, p. 3 b, new ed. of 1910). Under the Yuan dynasty a bureau for works in rhinoceros-horn and ivory was established. This was a sort of court-atelier, in which couches, tables, implements, and girdle-ornaments inlaid with these materials were turned out for the use of the imperial household. An official was placed in charge of it in 1263, and he received an assistant in 1268; the force consisted of a hundred and fifty work- ing-men (Yuan shi, Ch. 90, p. 5, K'ien-lung edition). According to Qazwml (1203- 83), the inhabitants of Sandabil (Kan-chou in Kan-su Province) were clad in silk and adorned with ivory and rhinoceros-horn (J. MARQUART, Osteuropaische und ostasiatische Streifzuge, p. 87, Leipzig, 1903). DE GOEJE is inclined to think in this connection of rhinoceros-horn set with gold and worn as amulet; but an instance of such a mode of use is not known in China, and it rather seems that it is in this case likewise the question of girdles decorated with plaques of ivory and rhinoceros-horn. The Mohammedan authors were well aware of the fondness of the Chinese for this material and its employment for girdles, and during the middle ages became the most active importers of the horn into China. The Arabic merchant Soleiman writ- ing in 851 relates that the inhabitants of China make from the horn girdles reaching in price to two and three thousand dinars and more, according to the beauty of the figure found in the design of the horn (M. REINAUD, Relation des voyages faits par les Arabes, Vol. I, p. 29). Hafiz el Gharb, who wrote at the end of the eleventh century, observed, "The most highly esteemed ornaments among the Chinese are made from the horn of the rhinoceros, which, when cut, presents to the eye singular and varied figures" (CH. SCHEFER, Relations des Musulmans avec les Chinois, p.io, in Cenlenaire de I'Ecole des langues orientales, Paris, 1895). 1 Too, "to reverse;" ch'a, "to insert;" t'ung, "to pass through." 1 Thus this phrase is explained in GILES'S Dictionary, p. 1326 b (tenth entry). * PLAYFAIR (2d ed.), No. 4927: one of the two districts forming the city of Kuang-chou (Canton). * HIRTH and ROCKHILL (Chau Ju-kua, p. 233), relying on Gerini, identify the coun- try To-lo or To-ho-lo, as written in "Tang shu, with a country situated on the Gulf of Martaban. The journey from Kuang-chou to that country takes five months. An embassy with tribute came from there to China in the period Chfing-kuan (627- 650) , and emphasis is laid on the great number of fine rhinoceroses. See also SCHLEGEL (Toung Pao, Vol. IX, 1898, p. 282) and PELLIOT (Bull, de I'Ecolc fran$aise, Vol. IV, 1904, P. 360)- 144 CHINESE CLAY FIGURES dishes, utensils, platters, and the like. 1 Then there is the horn ' frighten- ing fowl' with a white, silk-like thread; placed in the rice, it scares the fowl away. The ' dust-dispelling horn ' is utilized to make hairpins and combs for women; it keeps dust out of the hair. As to the 'water- dispelling horn,' when brought into the water of a river or the sea, it has the power of breaking a way across it. Exposed to a fog, and in the evening, it does not contract moisture. As to the 'resplendent horn,' this one, when put in a dark house, emits its own light. 2 Of all these various horns, I know only from hearsay, for I have not been able to procure and see them." The Yu yang tsa tsu by Tuan Ch'e'ng-shi of the ninth century 8 makes the following comments on the rhinoceros: "The variety of rhinoceros styled ' communicating with the sky ' dislikes its own shadow, and is in the habit of drinking muddy water. 4 When the animal is im- mersed in the water, men avail themselves of this opportunity to cap- ture it, as it is impossible for it to pull its feet out of the mud. The natu- ral structure of the horn is such that it is filled with figures resembling objects of nature. It is asserted by others that the designs penetrat- ing the rhinoceros-horn are pathological. 5 There are three varieties of design, styled tao ch'a ('lines inverted and inserted'), cheng ch'a ('straight and inserted'), and yao ku ch'a ('inserted like a barrel-shaped drum'). 6 They are styled 'inverted,' if one half of the lines pass 1 The colors indicated by the Chinese writers altogether answer the facts. In its exterior, the color of rhinoceros-horn is usually black or dark brown. A cross-section reveals various colors. A specimen kindly presented to the Museum by Mr. F. W Kaldenberg of New York exhibits in the interior a large black zone running through the centre and extending from the base to the tip, and filling the entire space of the extremity. In the lower, broad portion it is surrounded on the one side by a gold- brown section, about 3.5 cm wide and 21 cm long, and on the other side by a mottled light-yellow and greenish zone almost soap-like in appearance. This horn was found in the woods, and is in places eaten through by insects. The surface of the base exhibits the tips of the bristles, and appears like a coarse brush. The fibres running longitudinally, owing to the effect of weathering, can be easily detached. * As shown above (p. 138), optic properties are attributed to the horn as early as the time of Ko Hung. The subject is discussed in detail below (p. 151). 1 As now established by P. PELLIOT (T'oung Pao, 1912, pp. 373-375), this work was published about 860. 4 The Pin ts'ao adds, " In order to avoid beholding its reflection." This notion is doubtless derived from the animal's predilection for a mud-bath ; its favorite haunts are generally in the neighborhood of swamps (LYDEKKER, /. c., p. 31). 6 The Pin ts'ao adds, "But the natural reason cannot be ascertained." This is a comment of Su Sung. ' The meaning of these technical terms is not quite easy to grasp. The word tao (No. 10,793) is "to invert," ch'a (No. 205) means "to insert:" tao ch'a, accordingly, may mean "lines inserted in the horn in an inverted position;" and ching ch'a, "lines inserted straight." Yao ('loins') ku (No. 6421 ; in Pin ts'ao erroneously No. 6227) is the former name for a barrel-shaped drum (hua ku, see A. C. MOULE, Chinese Musical Instruments, p. 57, where an example from a verse of Su Tung-p'o is quoted). Yao K'uan, the author of the Si k'i ts'ung yu, written about the middle of the twelfth HISTORY OF THE RHINOCEROS 145 through in the direction downward. They are styled 'straight,' if one half of the lines pass through in the direction upward. They are styled 'drum-shaped,' if the lines are interrupted in the middle, without passing through. The Po-se designate ivory as po-ngan, and rhinoceros- horn as hei-ngan. 1 Wu Shi-kao, a physician from Ch'ng shi m6n, century (WvLiE, Notes, p. 160), makes the following remark: "The fundamental color of rhinoceros-horn is black. Is the color simultaneously black and yellow, the horn is styled 'standard throughout' (cheng t'ou). Is the horn yellow with black borders, it is styled 'inverted throughout' (tao t'ou). The horns of standard color are highly esteemed by our contemporaries. If the shape of the horn is round, it is designated as 'horn communicating with the sky' (t'ung t'ien si). In the south, there are counterfeits which may be recognized from gradually getting warm when rubbed. In view of the fact that rhinoceros- horn by nature is cold, it does not become warm when rubbed." 1 Su Sung[ makes the addition, "words difficult to distinguish." Po-ngan means literally "white ngan" (No. 57), and hei-ngan "black ngan, ' evidently transcrip- tions of Po-se words. PALLADIUS, in his Chinese-Russian Dictionary (Vol. I, p. 7), has indicated po-ngan ("ivory") and hei-ngan ("rhinoceros-horn") as Persian loan- words. Ivory, however, is called in Persian shirmaht; and rhinoceros, as well as the horn of it, kerkeden. It is true that Po-se is the Chinese name for Persia, which first appears in the Wei shu; but Persia is not meant in the above passage. P'ei wen yun fu (Ch. 8, p. 89 b) gives three quotations under the heading hei-ngan si. One from a book Sheng shui yen fan says that the Po-se call rhinoceros-horn hei-ngan; the refer- ence to the name of ivory is omitted, so that the clause "it is difficult to discriminate" makes no sense. The second is derived from the Leng chai ye hua of the monk Hui- hung, written toward the close of the eleventh century (WYLIE, Notes on Chinese Literature, p. 164), and says that "the men of the south (nan jen) designate ivory as po-ngan, rhinoceros-horn as hei-ngan." The third reference is taken from a poem of Tu Fu (712-770), who remarks that hei-ngan is a general article of trade of the Man. These texts render it probable that the country of Po-se here referred to is not Persia, but identical with the Malayan region Po-se mentioned by Chou K'u-fei in his Ling-wai tai ta, written in 1178 (Ch. 3, p. 6b; edition of Chi pu tsu chai ts'ung shu), and then after him in the Chufanchi, written in 1225 by Chao Ju-kua (translation of HIRTH and ROCKHILL, p. 125). The two authors seek it in or near the Malay Peninsula, though Negritos are not necessarily to be understood: the mere state- ment that the inhabitants have a dark complexion and curly hair is not sufficient to warrant this conclusion. GERINI identifies the name Po-se with Lambesi below Atjeh on the west coast of Sumatra, which seems somewhat hypothetical. Mr. C. O. BLAGDEN (Journal Royal As. Soc., 1913, p. 168) is inclined to regard Po-se as identical with Pase (or Pasai) in north-eastern Sumatra, but adds that there is no evidence that the place existed as early as 1178. The above text shows that the Po-se of the Chinese mediaeval writers were a Malayan tribe speaking a Malayan language, for the two transcriptions po-ngan and hei-ngan can be interpreted through Malayan. In the Hakka dialect, hei-ngan is hel-am; and hitam is the Malayan word for "black" (Javanese Ngoko hireng). Pei-ngan is in the Hakka dialect p'ak-am (compare Dic- tionnaire chinois-francais dialecte Hac-ka by CH. REY), in Cantonese pak-om, in Yang-chou puk-yd. In Javanese Krami "white" is petak, in Javanese Ngoko putih, likewise in Batak, in common Malayan puteh. We should expect that the two Malayan words, judging from the Chinese transcriptions, would terminate in the same syllable, which caused misunderstandings on the part of Chinese dealers. There is (or was) perhaps a certain Malayan dialect, in which the word for "white" ended in -am, or in which the words for "white " and "black " terminated in -t or -ih (compare Madagassy intim, inti, "black;" and puti, "white;" G. FERRAND, Essai de pho- ne^ique comp. du malais et des dialectes malgaches, pp. 24, 54, Paris, 1909). It is evident that neither the Malayan words for "ivory" (g&ding, Javanese gating) and "rhi- noceros-horn" (chulabadak or simply chula), nor the words for "elephant" (gdjah, Java- nese gajah) and "rhinoceros" (badak, Javanese warak), are intended here, but only the color names ' ' white ' ' and ' ' black , " with which the traders distinguished ivory and rhi- 146 CHINESE CLAY FIGURES while he served in the district of Nan-hai (in Kuang-tung), had occasion to meet there a captain who told him this story: 'The people of my country, intent on capturing a rhinoceros, proceed to erect on a mountain- path many wooden structures like watch-houses or posts for tethering animals. 1 As the front legs of the animal are straight, without joints, it is in the habit of sleeping by leaning against a tree. The rotten timber will suddenly break down, and the animal is unable to rise. 2 Another noceros-horn . The Malayan word badak seems to cover the entire Malayan area where the rhinoceros is found; it occurs on Borneo in the language of the Dayak (A. HARDE- LAND, Dajacksch-deutsches Worterbuch, p. 24, Amsterdam, 1859), and on Sumatra (M. JOUSTRA, Karo-Bataksch Woordenboek, p. 59, Leiden, 1907). Among the Malayans, the rhinoceros-horn (chula) is supposed to be a powerful aphrodisiac; and there is a belief in a species of "fiery" rhinoceros (badak api) which is excessively dangerous when attacked (W. W. SKEAT, Malay Magic, p. 150, London, 1900). The horn is carefully preserved, as it is believed to be possessed of medicinal properties, and is highly prized by the Malays, to whom the Semang generally barter it for to- bacco and similar commodities (SKEAT and BLAGDEN, Pagan Races of the Malay Peninsula, Vol. I, p. 203, London, 1906). There is nothing in these Malayan beliefs showing that complex series of ideas, met with in China. They may be a weak echo of Chinese notions conveyed by Chinese traders bartering among them for the horn. 1 Chii yi (Nos. 2974 and 13,205). I do not know but this may have to be taken as a compound with a more specific technical meaning. The two PSn ts'ao have changed this unusual term into "stables for swine or sheep." There is no doubt of what is meant, posts of rotten timber, which will easily break to pieces under the burden of the animal leaning toward it. 2 This story has passed also into the Arabic account of the merchant-traveller Soleiman, written in 851 A.D. (M. REINAUD, Relation des voyages faits par les Arabes et les Persans dans 1'Inde et a la Chine, Vol. I, p. 29, Paris, 1845): "The kerkeden (rhinoceros) has no articulation in the knee, nor in the hand; from the foot up to the armpit it is but one piece of flesh." In T'oung Pao (1913, pp. 361-4) the historical importance of this tradition is pointed out by me inasmuch as this originally In- dian story has migrated also to the West, where it leaks out in the Greek Physiologus (only the rhinoceros is replaced by the elephant) , and in CAESAR'S and PLINY'S stories of the elk. I wish to make two additions to these remarks. AELIAN (Nat. an., xyi, 20), describing the rhinoceros of India, called by him Kapr&favos, asserts that its feet have no joints and are grown together like the feet of the elephant (TOUS plv ?r66as t5iapt9pa>7Toi;s T teal f/*vKfvai: ed. of F. JACOBS). This passage, therefore, confirms my former conclusion that it was the rhinoceros which was credited in India with jointless legs; but we see that the same notion was like- wise attached to the elephant. It may be the case, accordingly, that the elephant with jointless legs was borrowed by the Physiologus straight from India. Mr. W. W. ROCKHILL (Diplomatic Audiences at the Court of China, p. 32, London, 1905) quotes a statement made to him by T. WAITERS on the kotow question with reference to Lord Macartney's embassy, as follows: "It was an opinion universal, and was told among the Chinese, that the Kuei-tse or foreigner was not built up like the jen [that is, man] or Chinaman, and particularly that he had no joints in his legs. So that, if the Kuei-tse was knocked down or otherwise put on the ground, he could not rise again. It was because the Emperor did not want to have possibly a death or at any rate an unseemly spectacle that he waived the kotow." Compare also Rubruck's story of "the creatures who have in all respects human forms, except that their knees do not bend, so that they get along by some kind of jumping motion" (W. W. ROCKHILL, The Journey of William of Rubruck, p. 199, London, 1900). The fabulous notion of the jointless legs of the rhinoceros may have arisen from the observation that the animal is indeed in the habit of sleeping in a standing position. Says E. HELLER (The White Rhinoceros, p. 41), "The hot hours of the day are spent by the white rhinoceros sleeping in the shade of the scattered clumps of trees or bushes which dot the grassy veldt. They seem to rest indifferently, either lying down or standing HISTORY OF THE RHINOCEROS 147 name for the rhinoceros is nu kio. There is also the ch8n ch'u, which is presumably a rhinoceros. The rhinoceros has three hairs growing out of each pore. 1 Liu Hiao-Piao asserts that the rhinoceros sheds its horn and buries it, and that people exchange it for a counterfeit horn." The story alluded to in the latter clause is better worded in the Pin ts'ao, which says, " It is told also that the rhinoceros sheds its horn every year, and itself buries it in the mountains. The people near the sea, with all secrecy, make wooden horns, and exchange these for the real ones, and so they go ahead continually. If they would go to work openly, the animal would conceal its horns in another place and defy any search." 2 Li Sun, who wrote an account of the drugs of southern countries (Hai yao pen ts'ao) in the second half of the eighth century, expresses himself in these words: "The rhinoceros 'communicating with the sky,' during the time of pregnancy, beholds the forms of things 3 passing across the sky, and these are reproduced in the horn of the embryo: hence the designation ' communicating with the sky.' * When the horn, placed in a water-basin during a moonlight night, reflects the brilliancy of the moon, it is manifest that it is a genuine horn 'communicating with the sky.' The Wu k*i ki 6 says, 'The mountain-rhinoceros lives on bamboo and trees. Its urinating is not completed in the course of a day. The I Liao 8 get hold of it by means of bow and arrow. This is up with lowered head. When at rest they stand with their noses almost touching the ground, their heads being elevated to a horizontal position only when alarmed." 1 The same is said in the P&n ts'ao in regard to the seal (compare G. SCHLEGEL, Toung Pao, Vol. Ill, 1892, p. 508). Compare p. 140. 1 In the text of the Ch&ng lei pn ts'ao, Su Sung terminates, " I do not know wheth- er at present they take horns in this manner or not." Compare the account of Ko Hung, p. 139. The Ch&ng lei p$n ts'ao reads "the destiny of things" (wu ming) instead of "forms of things" (wu king). * In the notes embodied in the Pdn ts'ao regarding the elephant (Ch. 51 A, p. 4) it is said that the patterns in the horn are formed while the rhinoceros gazes at the moon, and that the designs spring forth in the tusks of the elephant while the animal hears the thunder. A work Wu Ung hui yuan, as quoted in P'ei w&n yunfu (Ch. 21, p. 113 b), similarly says that the rhinoceros, while enjoying the moonlight, produces the designs in its horn, and that the floral decorations enter the tusks of the elephant when it has been frightened by thunder. These passages prove that it is material heaven to whose influence the formation of the natural veins in horn and tusk is ascribed. The rhinoceros gazing at the moon is represented in T'u shu tsi ch'ing (Fig. 10). 6 A work listed in the Tai P'ing yu Ian as being published in 983; but, as it is quoted here by Li Sun, it must have existed in or before the eighth century. 8 An aboriginal tribe belonging to the stock of the Man, according to T'ang shu (Ch. 43 A, p. 6 b) settled in Ku chou (PLAYFAIR, No. 3256) in the province of Kuei- chou. Compare p. 82 in regard to the possibility of killing a rhinoceros with arrows. 148 CHINESE CLAY FIGURES the so-called rhinoceros of K'ien.' l The I wu chi 2 says, 'In the sea- water of Shan-tung there is a bull that delights in the sounds of string and wind instruments. When the people make music, this bull leaves the water to listen to it, and at that moment they capture it.' 3 The rhino- ceros has a horn on its nose, and another on the crown of its head. The nose-horn is the one best esteemed. The natural histories (pen ts'ao) are acquainted only with the mountain-rhinoceros. I have not yet seen the water-rhinoceros." 4 K'ou Tsung-shi, a celebrated physician of the Sung period, reports in his Pen ts'ao yen i (completed in 1 1 16) 5 thus: " The designs in the horns of the river-rhinoceros and the southern rhinoceros are fine. The black rhinoceros-horn has designs clearly displayed, while the yellow rhinoceros-horn has very sparse designs. None equals the patterns in the horn of the Tibetan breed, which are high, and come out clearly at both ends. 6 If the forms of objects pictured in the horn are yellow, while the rest is black, the horn is 'standard color throughout' (cheng t'ou). If the forms of objects are black, while the rest is yellow, the horn is 'inverted throughout' (tao t'ou). If the black color is taken as stand- ard, and the forms of the design are imitative of real objects, the horn is a treasure; this horn is styled t'ung si ('penetrating rhinoceros'). It is an indispensable condition that the patterns come out clearly, and that the yellow and black be sharply differentiated. If both ends are moist and smooth, the horn is of the first quality." 7 1 The territory of the province of Kuei-chou, where the rhinoceros formerly occurred, as already attested by Su Sung (above, p. 140). 2 Several works of this title were in existence (see BRETSCHNEIDER, Bot. Sin., pt. i, p. 154). * The animal in question is certainly not a rhinoceros, and has crept in here by way of wrong analogy. In his notes on cattle, Li Shi-chen mentions a variety "ma- rine ox" (hat niu, Ch. 51 A, p. 7 a). This creature is described after the Ts'i ti ki by Fu Ch'en of the fifth century or earlier (BRETSCHNEIDER, Bot. Sin., pt. i, p. 201) as follows: "Its habitat is around the islands in the sea near Teng-chou fu (in Shan- tung); in shape it resembles an ox, it has the feet of an alligator (t'o No. 11,397, n t iguana, as GILES still translates, despite the correction of E. v. ZACH, China Review, Vol. XXIV, 1900, p. 197), and the hair of a bull-head fish. Its skin is soft, and can be turned to manifold purposes; its blubber is good to burn in lamps." The marine ox, accordingly, must be an aquatic mammal of the suborder of Pinnipedia (seals). There may be a grain of truth in the above story : the intelligence of seals is remark- able, they are easily tamed and susceptible to music. There is an interesting chapter on tamed seals in the classical treatise of K. E. v. BAER, Anatomische und zoologische Untersuchungen uber das Wallross (Mimoires de I'Acad. imp. des sciences de St. Peter sbourg, 6th series, Vol. IV, 1838, pp. 150-159). 4 The last clause is not in the text of Cheng lei pin ts'ao. 4 PELLIOT (Bulletin de I' Ecole franfaise d' Extreme-Orient, Vol. IX, 1909, p. 217). 4 The rhinoceros of Tibet has been discussed above, p. 116. 7 The Arabic authors assert that the interior of the Indian rhinoceros-horn fre- quently presents designs of a human figure, a peacock, or fish, and that the price paid in China is raised according to the beauty of these designs (M. REINAUD, Relation HISTORY OF THE RHINOCEROS 149 Li Shi-chSn himself, the author of the Phi ts'ao kang mu, sums up as follows: "The habitat of the rhinoceros is in the regions of the Si Fan, 1 the southern Tibetan tribes (Nan Fan), the southern portions of Yun- nan, and in Kiao-chou, and occurs there everywhere. There are three species, the mountain-rhinoceros, the water-rhinoceros, and the se si. There is, further, a hairy rhinoceros resembling the mountain-rhinoceros, and living in hilly forests; great numbers of it are captured by men. The water-rhinoceros makes its permanent abode in water, and is there- fore very difficult to capture. It has, in all, two horns. The horn on its nose is long, that on its forehead is short. The skin of the water- rhinoceros has a pearl-like armor, 2 but not so the mountain-rhinoceros. des voyages faits par les Arabes, Vol. I, p. 29). REINAUD (Vol. II, pp. 68, 69) com- ments on this point that the Chinese are satisfied to compare the designs with flowers and millet-seeds, and do not discover in them half of the things which the Arabs saw in them. It seems to me that the Arabs, in this case, merely reproduce the ideas of the Chinese. The philosophy of these designs was fully developed in the T'ang period. K'ou Tsung-shi speaks of real objects visible in the horn; and Wang P'i-chi, in his Shtng shui yen fan lu (p. 135), offers an elaborate contribution to this question. Ac- cording to him, "the designs in the horn from Kiao-chi are like hemp-seeds, the horn being dry, a bit warm, and glossy; the horn imported on ships and coming from the Arabs has patterns like chu yii flowers [this name applies to three different plants: BRETSCHNEIDER, Bot. Sin., pt. 2, No. 498], is glossy and brilliant with colors, some resembling dog-noses, as if they were glossed with fat; others with floral designs and strange objects, these horns being styled t'ung t'ien si; some like sun and stars, others like clouds and moon; some like the corolla of a flower, some like scenery; some have birds and mammals, others dragons and fishes; some have deities, others palaces; and there are even costume and cap, eyes and eyebrows, staff and footgear [conveying the illusion of the picture of a wanderer], beasts, birds, and fishes. When the horn is completed into a carving, as if it were a veritable picture, it is highly esteemed by the people. The prices are fluctuating, and it is unknown how they are conditioned." There is assuredly an inward relation between the statements of this account and the Arabic texts of Damlrt quoted by REINAUD (Vol. II, p. 69). It is hardly necessary to insist on the chronological point that Damlrl (1344-1405) wrote his zoological dictionary Hayat el-haiwan (C. HUART, Litt^rature arabe, P- 365, Paris, 1902) several centuries after Wang P'i-chi (end of eleventh century). From a psychological point of view, the dependence of the Arabs in this matter on the philosophy of the Chinese is self-evident. Neither the classical world nor ancient India has developed any similar thoughts ; and this subject is decidedly Chinese, with a strong Taoist flavor of nature sentiment. It must not be overlooked, either, that al-Berunl (SACHAU, Alberuni's India, Vol. I, p. 204) merely states that "the shaft of the horn is black inside, and white everywhere else," and that h'? is entirely reticent about figures in the horn. The Arabs interested in the trade of the horn to China imbibed this lesson, and propagated it themselves in catering to the taste of their customers. The question is whether, in the interest of the business, they did not help nature by art, and may have produced several of the more fanciful designs artificially. This, however, is no matter of great concern; and the fact remains that bristly fibres of various tinges compose the horn, and result in a natural play of design and color which is apt to arouse the imaginative power of a susceptible mind. I Western Tibetan tribes; from our standpoint, eastern Tibetans. I 1 take this to be identical with what our zoologists say in regard to the skin of the Asiatic species, which "has the appearance of a rigid armor studded with tuber- cles." The whole skin of the Javan species, as already remarked by B. CUVIER (The Animal Kingdom, Vol. I, p. 157, London, 1834), is covered with small compact angular tubercles. JOANNES RAIUS (Synopsis methodica animalium quadrupedum, p. 122, Londini, 1693) describes the skin of the rhinoceros thus: "Auriculae porcinae, 150 CHINESE CLAY FIGURES The se si is the female of the rhinoceros which is termed also 'sand- rhinoceros.' It has but a single horn on the crown of the head. The natural designs of the horn are smooth, white, and clearly differentiated, but it is useless as medicine, for the patterns on the horn of the male are big, those on the horn of the female too fine. In the beginning of the period Hung-wu (1368-1398) Kiu-chen 1 sent one as tribute, which was called one-horned (monoceros) rhinoceros. The view of Ch'n Ts'ang- k'i that there are not the two kinds of land and water animals, the view of Kuo P'o that the rhinoceros has three horns, and the view of Su Sung that the hairy rhinoceros is the male rhinoceros, are all erroneous. The term 'hairy rhinoceros' is at present applied to the yak. 2 The designs of the rhinoceros-horn are like fish-roe. On account of their shape they are styled 'grain patterns.' 3 Inside of the latter there are eyes, styled 'grain eyes.' If yellow decorations rise from a black back- ground, the horn is 'standard throughout.' If black decorations rise from a yellow background, the horn is 'inverted throughout.' If within the decorations there are again other decorations, the horn is 'double throughout.' The general designation for these is t'ung si, and they are of the highest grade. If the decorations are spotted, as it were, with pepper and beans, the horns are middle grade. The horn of the black rhinoceros, which is of a uniform black color and devoid of decorations, is the lowest grade. 4 If the horn of the rhinoceros 'com- municating with the sky ' emits light, so that it can be seen at night, it is molli et tenui cute vestitae; reliquum corpus dura admodum et crassa, velut squamis quibusdam crustaceis rotundis aspera." This is the reason why in some Chinese and early European sketches the ^ animal is covered with scales (see Figs. 3 and II, and Plate IX). 1 PLAYFAIR, No. 1295 (1278): in Annam (compare above, p. 81). 2 Li Shi-ch6n refers to the notes on this subject contained in the same chapter. This remark renders it plain that it was the notion of "rhinoceros" which was trans- ferred in recent times to the yak, and that the development was not in the reverse order, as assumed by Professor Giles. $ This and the following sentences, commenting on the natural designs of the horn, have been translated by S. JULIEN (in M. REINAUD, Relation des voyages faits par les Arabes, Vol. II, p. 68). 4 In the Memoirs on the Customs of Cambodja by Chou Ta-kuan of the Yuan period, translated by P. PELLIOT (Bulletin de I'Ecole jranqaise d 'Extreme-Orient, Vol. II, 1902, p. 167), it is said that the white and veined rhinoceros-horn is the most es- teemed kind, and that the inferior quality is black. The List of Medicines exported from Hankow, published by the Imperial Maritime Customs (p. 15, Shanghai, 1888), is therefore wrong in stating that the black and pointed horns are considered the best. A valuation for the horn is not given there. According to a report of Consul-General G. E. ANDERSON of Hongkong (Daily Consular and Trade Reports, 1913, p. 1356), rhinoceros-horns are imported into Hongkong to some extent, the price ranging from $360 to $460 per picul, or from about $1.30 to $1.65 gold per pound; they are largely of African production, and imported from Bombay. According to L. DE REINACH (Le Laos, Paris, no date, p. 271), rhinoceros-horns have in the territory of the Laos a market- value of 1 1 1-137 fr- the kilo, and rhinoceros-skins 60-70 fr. a hundred kilo. HISTORY OF THE RHINOCEROS 151 called 'horn shining at night' (ye ming si}: 1 hence it can communicate with the spirits, and open a way through the water. Birds and mammals are frightened at seeing it. The Shan hai king speaks of white rhino- 1 This idea may have been borrowed from the precious stones believed to shine at night (HiRTH, China and the Roman Orient, pp. 242-244; CHAVANNES, Les pays d'occident d'apres le Heou Han Chou, T'oung Pao, 1907, p. 181). Jade disks shining at night (ye kuang pi) are mentioned in Shi ki (Ch. 87, p. 2 b). The note of Li Shi- chn is doubtless suggested by the following passage of the Tu yang tsa pien, written by Su Ngo in the latter part of the ninth century (WvLiE, Notes on Chin. Lit., p. 194 ;ed. otPaihai, Ch. B, p. 9, or P'ei w$n yun fu, Ch. 8, p. 87 b): " In the first year of the period Pao-li (825 A.D.) of the Emperor King-tsung of the T'ang dynasty, the country of Nan-ch'ang [in Kiang-si; PLAYFAIR, No. 4562] offered to the Court a rhi- noceros-horn shining at night (ye ming si). In shape it was like the 'horn com- municating with the sky.' At night it emitted light, so that a space of a hundred paces was illuminated. Manifold silk wrappers laid around it could not hide its luminous power. The Emperor ordered it to be cut into slices, and worked up into a girdle; and whenever he went out on a hunting-expedition, he saved candle-light at night." We even hear of a luminous pillow (ye ming chtri) lighting an entire room at night ( Yun sien tsa shi, Ch. 6, p. 3 b, in T'ang Sung ts'ung shu, which quotes from K'ai-yuan T'ien-pao i shi). The story of Tu yang tsa pien may be connected with the curious tradition regarding WSn K'iao (Tsin shu, Ch. 67, p. 5), who by the alleged light emitted from a rhinoceros-horn beheld the supernatural monsters in the water (see P&TILLON, Allusions litte"raires, p. 227; S. LOCKHART, A Manual of Chinese Quotations, p. 280; and GILES, Dictionary, p. 794 b, who translate 'to light a rhi- noceros-horn,' which is not possible, as in this case the horn would burn down ; the horn was shining through its alleged own light) . An illustration of this scene by Ting Yun- p'fing is published in Ch'eng shi mo yuan and Fang shi mo p'u. The notion that the rhinoceros-horn is luminous at night, and is therefore styled "shining or bright horn" (ming si, or kuang ming si), and also "shadow horn" (ying si), is found in Tung ming ki (Wu-ch'ang print, Ch. 2, p. 2), embodied in a fabulous report on a country Fei-lo, said to be nine thousand li from Ch'ang-ngan in Indo-China (Ji-nan). This work relating to the time of the Han Emperor Wu, though purported to have been written by Kuo Hien of the Han, is one of the many spurious productions of the Leu-ch'ao period (fourth or fifth century), and teeming with anachronisms and gross inventions; some accounts in it are interesting, but devoid of historical value (see WYLIE, Notes, p. 1 9 1 ) . The assertion there' made, that the inhabitants of Fei-lo drive in carriages drawn by rhinoceros and elephant, is very suspicious; but the report that the horns sent from there were plaited into a mat, the designs of which had the appearance of reticulated silk brocade, is probably not fictitious; for this is confirmed by a passage of the T'ang Annals (Chapter wu king chi, quoted in T'u shu tsi ch'&ng), according to which a certain Chang Yi-chi had a mat made for his mother from rhinoceros-horn. Since the latter (the designation "horn," from a scientific standpoint, is a misnomer) is composed of agglutinated hair or bristles, it is possible to dissolve a horn into thread- like fibres; and the possibility of a technique employing these for the plaiting of mats must be admitted. 1 According to the more precise wording of the passage, as quoted in P'ei wen yun fu (Ch. 8, p. 88 a), the white rhinoceros occurs in the mountains of Kin-ku, inhabited by large numbers of other wild animals, also hogs and deer. The Shan hai king is an apocryphal work teeming with fables, and has little value for scientific purposes. The P'ei wen yun fu, further, quotes the Tung kuan Han ki (completed about 170 A.D.; BRETSCHNEIDER, Bot. Sin., pt. i, No. 990) to the effect that in the first year of the period Yuan-ho (84 A.D.) of the Emperor Chang of the Han dynasty the country Ji-nan (Tonking) offered to the Court a white pheasant and a white rhi- noceros. But this text, unreservedly accepted by HIRTH (Das weisse Rhinoceros, T'oung Pao, Vol. V, 1894, p. 392), must be taken with some caution, as it is identical with, and apparently derived from, the passage in Hou Han shu (Ch. 1 16, p. 3 b), according to which, in the first year of the period Yuan-ho (84 A.D.), the Man I beyond the boundary of Ji-nan offered to the Court a live rhinoceros and a white pheasant. The 1 52 CHINESE CLAY FIGURES "The work K'ai-yiian i ski 1 mentions the ' cold-dispelling' rhinoceros- horn (pi han si), whose color is golden, and which was sent as tribute by Tonking (Kiao-chi) , 2 During the winter months it spreads warmth, which imparts a genial feeling to man. The Po k'ung leu t'ie 3 speaks of the 'heat-dispelling' rhinoceros-horn (pi shu si) obtained by the Emperor Wen-tsung (827-840 A.D.) of the T'ang dynasty. 4 During the summer months it can cool off the hot temperature. The Ling piao lu i b records the horn of the 'dust-dispelling' rhinoceros (pi ch'en si), from which hairpins, combs, and girdle-plaques are made, with the effect that dust keeps aloof from the body. The Tu yang tsa pien 6 text of the official Annals is decisive, and it is easy to see that the word "live" could have been altered into "white" by the suggestion of the white pheasant. The T'ang leu tien, a description of the administrative organization of the period K'ai-yuan (713-741) of the T'ang dynasty, ascribed to the Emperor Yuan-tsung (compare PELLIOT, Bulletin de I'Ecole franc, aise A). HISTORY OF CHAIN MAIL AND RING MAIL 243 same name we meet this armor in the soldiery of the Indian Moghuls. 1 It is figured among the sketches of the Ain I Akbari, a history of the Emperor Akbar, written in 1597 by Abul Fazl Allami (1551-1602). 2 As this work has now become exceedingly rare, three illustrations from it are here reproduced from a copy in the writer's possession (Fig. 38). They are instructive from more than one point of view. First, they furnish actual proof of Persian chain mail, as well as helmet, having been transmitted from Persia into India. Second, as regards the manner of drawing, it will be noticed that the coat in Fig. 38 6 is striking- ly similar to the Chinese sketch of ring mail in Fig. 41. Both convey the impression of scale armor, but are explained as, and intended for, chain mail and ring mail respectively. 3 It is exceedingly difficult to produce a good sketch of either; and it is interesting to note that two draughtsmen, independent of each other, have had recourse to the same mechanical means of representing them. They teach, as many other cases, that caution and criticism are necessary in diagnosing types of armor after pictorial or other designs. 4 The helmet (Fig. 38 a) with nasal and coif of mail (mighfar) is the same as that still extant in India, and from there conveyed to Tibet (Plate XXVIII). IRVINE (p. 565) describes the zirih as a coat of mail with mail sleeves, composed of steel links, the coat reaching to the knees. There are six specimens in the Indian Museum. Armor in the collection of the Nawab Wazlr at Lakhnau is described in 1785 as follows: "The armor is of two kinds, either of helmets and plates of steel to secure the head, back, breast, and arms, or of steel network, put on like a shirt, to which is attached a 1 W. IRVINE, The Army of the Indian Moghuls (Journal Royal As. Soc., 1896, P- 565)- 1 Translation of H. BLOCHMANN, Vol. I, Plate XIII (Calcutta, 1873). 1 IRVINE (/. c., p. 564) remarks that from this figure it may be inferred that, in a more specific sense, baktar or bagtar was the name for fish-scale armor. Yet BLOCH- MANN'S explanation of this figure, according to the Ain I Akbari, is "chain mail with breastplate (bagtar)." 4 Chinese sketches of defensive armor certainly are far from being good or accu- rate; on the contrary, they are purely conventional in style, a fixed and ready-made motive or model being employed for each type of armor. Yet they are not much worse than corresponding designs from India, Persia, and mediaeval Europe. At all events, they are interesting, and in many respects even instructive. Whatever their defects may be, if we are willing to understand the symbolic language of the draughts- men, their productions allow us in the majority of cases to recognize what type of armor is intended by them, in the same manner as inferences as to the type of armor intended may be deduced from the terminology of the language. In cases where no actual specimens are at our disposal, the Chinese illustrations may still claim a pri- mary importance; where we have specimens to study, as in the case of chain mail and plate armor, the sketches of the Chinese afford opportunity for an instructive com- parison; and for this reason I have drawn upon these sources also. They may render us essential assistance in interpreting the types of armor represented in statuary and painting. 244 CHINESE CLAY FIGURES netted hood of the same metal to protect the head, neck, and face. Un- der the network are worn linen garments quilted thick enough to resist a sword. The steel plates are handsomely decorated with gold wreaths and borders, and the network fancifully braided." Thus Persian chain mail spread to India in the Moghul period. W. EGERTON l observes that Persian arms were generally worn by the upper classes in India, and that the blades of swords were often Persian, even though mounted in India; in fact, as Persian artificers were fre- quently employed at the principal native courts, it is difficult sometimes to say whether a piece of armor is Persian or Indian. Whether ancient pieces of chain mail are still preserved in Persia, I am unable to say. 2 Plates XXIII XXV illustrate a piece of mail com- plete with all paraphernalia, the shirt with long sleeves being open in front. It was obtained at Tiflis by Mr. Charles R. Crane of this city, and is said to have served as the parade armor of a chieftain of the Khew- sur.* It is doubtless of Persian manufacture, as proved principally by the Persian designs on the arm-guard (Plate XXV, Fig. 2). J. MouRiER 4 has already observed that the helmets with coifs of mail and the suits of chain mail found among the tribes of the Caucasus seem to be of Persian origin. The rings forming the texture of that mail con- sist of thin iron wire loosely twisted together, being neither welded nor riveted. This rather degenerate style of workmanship testifies to the fact that the suit in question was merely intended for ceremonial or pageant purposes: an energetic sword-blow would probably shatter the whole outfit. The iron casque of the well-known Persian form, called in Persian zirih-kulah, is provided with a sliding nasal (nose- guard), and with a couvre-nuque consisting of a long coif of mail guard- ing forehead, cheeks, neck, and shoulders. On Plate XXV the two-edged sword, arm-guard, hauberk, and gauntlet, completing the set, are shown. The Arabs have undoubtedly derived chain mail from the Persians. All the available historical evidence is decidedly in favor of Persian prior- I An Illustrated Hand-Book of Indian Arms, p. 142 (London, 1880). 'According to EGERTON (/. c., p. 141), armor is now no longer worn in Persia, except to add to the pageant of their religious processions, held annually in the month of Muharram, to commemorate the death of Hassan and Hussain, the Shiah martyrs. Many that are of modern manufacture have been made for ornament rather than use, and betray in their style the decline of the art. The best period, judging from the examples preserved, seems to have extended from the time of Shah Abbas to that of Nadir Shah. The armor of Shah Abbas is in the British Museum; it is figured in G. MIGEON (Manuel d'art musulman, Vol. II, p. 251, Paris, 1907). I 1 am under obligation to Dr. Charles B. Cory, the present owner of the armor, for his courtesy in placing it at my disposal. 4 L'art au Caucase, pp. 156, 157 (Paris, 1907). HISTORY OF CHAIN MAIL AND RING MAIL 245 ity. 1 Among the ancient Arabs of the pre-Islamic epoch we meet with leather and iron armor, 2 without any clear description of their appear- an.ce. The latter seem previously not to have consisted of mail, though th s cannot be stated positively; but according to the descriptions of the poets, chain mail comes into question in the majority of cases. 8 Tradi- tion ascribed its invention to King David, and the Koran (SQreXXI, 80; XXXIV, 10) sets forth that God himself taught David how to smelt iron, with which to make the rings, and to join them into a solid armor. This story certainly is devoid of historical value. The place SalQk in Yemen was of old renowned for its armor consisting of a double row of rings. Also " Persian armor " is mentioned in Arabic records, where- by garments lined with silk and cotton were understood. "Armor from Sogd" (Sogdiana) became known after the foreign conquests of the Arabs. 4 Possibly also scale armor was worn. 5 Chao Ju-kua narrates that the ruler of Basra, when he shows himself in public, is accompanied by more than a thousand mounted retainers in full iron armor, the officers wearing chain mail. 6 During the early middle ages of Europe, the horses of armies were not caparisoned. Only from the beginning of the thirteenth century, probably under the influence of the Crusades, were they pro- tected by chain-mail covers. 7 According to MAX JAHNS,* the chain mail (Par sen, Barschen), as it first appears during that time in the armature of the horse, is probably of oriental, and more specifically of Persian origin. Dr. BASHFORD DEAN, 9 the great authority on armor in this country, offers the following suggestive summary of this subject: "Chain mail marked a distinct epoch in the development of arms and armor: for it was light, flexible, and extremely strong. And it soon, therefore, came to supplant the 1 Compare the notes of C. H. BECKER (Der Islam, Vol. IV, 1913, pp. 310-311). 1 F. W. SCHWARZLOSE, Die Waffen der alten Araberausihren Dichtern dargestellt, pp. 325. 328 (Leipzig, 1886). 1 Ibid., p. 331. 4 Ibid., p. 334. 6 G. JACOB, Altarabisches Beduinenleben, p. 136 (Berlin, 1897). BECKER (/. c.) mentions also Arabic cotton armor (lubbada); what he calls ring mail (Ringpanser), I believe, strictly speaking, is chain mail. In the age of the T'ang (618-906) the soldiers of the Arabs were equipped with bow, arrows, long spears, and metal armor (Tang shu, Ch. 221 B, p. 8 b). Lien huan so-tse kia, literally, "armor of chains, the links of which are mutually connected" (see HIRTH and ROCKHILL, Chau Ju-kua, p. 137). 7 G. STEINHAUSEN, Geschichte der deutschen Kultur, p. 247; L. BECK, Geschichte des Eisens, Vol. I, p. 863. 8 Ross und Reiter, Vol. II, p. 137. 9 Catalogue of European Arms and Armor (The Metropolitan Museum of Art, Hand-Book No. 15, p. 21, New York, 1905). 246 CHINESE CLAY FIGURES cruder defences of Carolingian times. Some authorities maintain that this form of armor was borrowed from the Orient ; and certain it is that its development in the twelfth and thirteenth centuries was largely influenced by oriental models. If, however, this form of armor were derived originally from the East, it is a rather remarkable fact that its early appearance in Europe should be traced so clearly to the northern peoples, and that the 'byrnie' (briinne), or shirt of mail, should have become a characteristic part of the equipment of a Norseman. Never- theless it may still have been derived primitively from the East, since it is well known that the early excursions of the Viking carried them well into the Mediterranean, and that even by the eighth century they were well acquainted with many objects of oriental origin." The Arabs and Byzantines have transmitted chain mail to Europe; and a share in this movement may be attributed to the cultural exchanges between East and West during the crusades. At the time of Mohammed the Arabs had already adopted the Persian practice of protecting horse and man with armor, the armored horsemen and horses being designated mudjaffaf; that is, clad with the tidffaf, the Persian felt armor. 1 When we come to China, the situation is the same as in Europe and in India. Historical evidence is not lacking for the foreign origin of Chinese chain mail. Indeed, the first record alluding to it, the T'ang shu, 2 in its account of K'ang (Sogdiana, Samarkand), states that in the beginning of the period K'ai-yiian (713-741), Samarkand sent to China chain armor (so-tse k'ai) as tribute. 3 The famous poet Tu Fu, who 1 Compare C. H. BECKER (Der Islam, Vol. IV, 1913, p. 311). BECKER states that the history of defensive armor in the Islamic world still remains to be written; but his remarks render it sufficiently clear that the origin of these things is to be sought in Persia, and that they were transferred to Europe through the medium of the Arabs and Byzantines. The soldiers of the Byzantine army were protected for the most part by scale armor, though, judging from quite early monuments, ring or chain mail was sometimes used (O. M. DALTON, Byzantine Art and Archaeology, p. 684, Oxford, 1911). 2 Ch. 221 B, p. i b. 3 A tribute of armor from Samarkand is still recorded in the Ming shi under the year 1392 (see BRETSCHNEIDER, China Review, Vol. V, p. 123). It can of course be presumed only that the chain mail sent by Samarkand was of Persian origin; but this conclusion is most probable, as the culture of Sogdiana, the capital of which was Samarkand, was thoroughly Iranian. From what was said above on "armor from Sogd" it seems that among the Arabs Sogdiana was regarded as a famous seat of the manufacture of armor. In view of the fact that chain mail is an Iranian import in China it is curious that in the Persian legend of Alexander's expedition to China, the King of China presents to him among many other things a hundred long coats of mail (H. ZOTENBERG, Histoire des rois des Perses, p. 440). In T'ang shu (Ch. 220, p. 3 b), where an account of the foreign tribes of the east, including Koreans and Tungusians, is given, mention is made of a so kia ("chain cuirass"); the word k'ai is not used, and the question is probably of a leather corselet with rings attached to its surface. HISTORY OF CHAIN MAIL AND RING MAIL 247 lived about this time (712-770), alludes in a verse to a "metal-chain cuirass" (kin so kid). 1 Chain armor (so-tse kid)* is distinctly mentioned in the Wan hua ku, a work written at the end of the twelfth century, 8 in which are enumerated the designations for thirteen kinds of armor known at that period. Chain armor is there listed as the twelfth in the series ; and it is expressly stated that it ranges in the class of iron armor (t'ie kid). In all probability, however, this passage is taken from the T'ang leu Hen (the "Six Statutes of the T'ang Dynasty") drawn up by the Emperor Yuan-tsung in the first part of the eighth century (p. 189) ; and as the thirteen kinds of armor on record are said to have been made at that time in the Imperial Armory, we may presume that chain mail was turned out by the Chinese as early as the T'ang period, after models first introduced from Samarkand. In the Biography of Han Shi-chung, who died in 1151,* a "chain connected armor" (lien so kid) capable of resisting bows is credited to this general ; 6 but it would seem that this newly-coined term does not refer to a real chain mail, but rather to ring mail, in which rows of iron rings are fastened to a foundation of leather (see p. 252). According to the testimony of WILLIAM OF RUBRUCK, chain mail, which he styles haubergeon, was known to the Mongols. 8 In the year 1345, during the reign of the Emperor Shun, Djanibeg (1342-1356), son of Uzbeg, 7 sent to China, among other products, swords, bows, and chain mail coming from Egypt (Mi-si-rh). 8 Chain armor had no official recognition in China, and was never introduced into the army. It is conspicuously absent in the military regulations of the Ming dynasty, nor is it mentioned in the well-informed military work Wu pei chi. We have to go as far down as in the K'ien- lung period to renew its acquaintance. We meet it there again as a foreign import. In the Imperial State Handbook of the Manchu 1 P'ei wtn yunfu, Ch. 50, p. 70 (under so), or Ch. 106, p. 74 (under kia). There is also a quotation given there to the effect that "the finest of armors are designated chain mail," derived from a poetical work Erh loo fang shi hua, the date of which is unknown to me. 2 Entered in GILES'S Dictionary, p. 1264 c, with the same translation. 3 BRETSCHNEIDER, Botanicon Sinicum, pt. i, p. 160, No. 330. * GILES, Biographical Dictionary, p. 251. His biography is in Sung shi (Ch. 364, p.i). 1 Sung shi, Ch. 364, p. 6 b. W. W. ROCKHILL, The Journey of William of Rubruck, p. 261 (London, 1900). Rubruck reports that he once met two Mongol soldiers out of twenty, who wore haubergeons. He asked them how they had got hold of them; and they replied that they had received them from the Alans, who are good makers of such things, and ex- cellent artisans. 7 BRETSCHNEIDER, Mediaeval Researches, Vol. II, p. 15. 8 Yuan shi, Ch. 43, p. 5 b (K'ien-lung edition). 248 CHINESE CLAY FIGURES Dynasty (Huang ch'ao li k'i t'u shi, Ch. 13, p. 53) a piece of chain mail is illustrated (reproduced in Fig. 39) under the name so-tse kia. It is recorded that in 1759, after the subjugation of Turkistan, numerous captives were made, and innumerable spoils of arms obtained which were hoarded by imperial command in a building of the palace, the Tz'e kuang ko. Among these trophies were several pieces of chain armor ; and FIG. 39. Iron Chain Mail from Turkistan (from Huang ch'ao li k'i I'u shf). a document recording this event was draughted, and deposited be- neath those objects in the treasury. This shows that in the K'ien-lung period chain armor was foreign to the Chinese and considered an object of curiosity and rarity. The specimen consists of a jacket and trou- sers. The rings are said to be iron ; but it is not stated whether they are riveted, nor can this be gathered from the illustration. The shirt of mail is closed in front, and put on over the head. The collar, as ex- HISTORY OF CHAIN MAIL AND RING MAIL 249 plained in the text, is made of white cotton and tied up by means of a cord. Two specimens of chain mail secured in China are represented on Plate XXVI. Both are jackets with sleeves, having a short slit under- neath the neck, and being tied up by means of a leather band. Though identical in appearance, they are of different technique. The shirt of mail shown in Fig. i of the Plate consists of riveted steel rings; the one in Fig. 2, of welded iron rings. The former was obtained at Si-ning, Kan-su Province, with the information that it had previously hailed from Tibet; the latter, at Si-ngan, Shen-si Province. These two coats, accordingly, are technically much superior to the one from the Caucasus, in which the rings are merely of twisted iron wire not welded. It is thus clear that there are coats of mail widely varying in the technical process and in quality. To decide the question as to the locality where the two specimens were manufactured would require a larger compara- tive material than is at my disposal. The Tibetans, as will be seen presently, must be discarded as being unable to produce chain mail. The Chinese, as we noticed, may have themselves made it in the T'ang period; it is certain, however, that none is turned out in China at the present time. Altogether, these specimens are scarce; and modern Chinese accomplishments in iron and steel are so crude and inferior, that it is difficult to believe in the Chinese origin of the two pieces of mail. Particularly the mail in Fig. i of Plate XXVI represents such a complex and toilsome technicality, involving so great an amount of time and patience as can be credited only to a highly professional and skilful armorer, who was a specialist in this line; the process of riveting steel rings, moreover, is not practised by the Chinese. My personal impression in the matter, therefore, is that the two mails were fabricated in Persia or Turkistan, and thereupon traded to China. An offensive weapon deserves attention in this connection, because a chain is utilized in it, and its invention is ascribed by the Chinese to a foreign tribe. This is the t'ie lien kia (No. 1132) pang, a weapon con- sisting of two wooden cudgels, the one nearly three times the length of the other, their upper ends being connected by an iron chain (Fig. 40). The longer cudgel is round, and is held by its lower end in the hands of the soldier; the shorter one is square in cut, and provided at the end with a sharp iron point intended to hit the enemy's head. The chain allow- ing it ample freedom of motion, it is swung around in a wide circle, thus making it a fierce and powerful weapon. The Wu pei chi, illustrating and describing this instrument (Ch. 104, p. 14), states that its original home was among the Si Jung (the Western Jung), one of the general designations for the Turkish and Tibetan tribes living north-west 250 CHINESE CLAY FIGURES * f r *. L. * > t* JK rb M! FIG. 40. Flail-like Cavalry Weapon (from Wu pet chf). HISTORY or CHAIN MAIL AND RING MAIL 251 from China; that they made use of it, while riding on horseback, in fighting Chinese infantry; and that the Chinese soldiers learned to handle it, and are more clever at it than the Jung. Its shape is com- pared to a threshing-flail; and it may even have been derived from this FIG. 41. Ring Mail of Steel Wire (from Wu pet chi of 1621). implement, with which it agrees in mechanical principle. It is still known in Peking under the name of "threshing-flail," and is used in fencing. I saw this sport practised in 1902, and at that time secured a specimen for the American Museum, New York. In the time of the Emperor K'ien-lung it was still employed in the Chinese army. * 1 Huang ch'ao It k'i t'u ski, Ch. 15, p. 25 b. According to this work, the weapon is first mentioned in the Tung lien of Tu Yu, who died in 812, where it is said that it was manipulated by women on the walls to resist invaders. Ti Ts'ing, the famed general in the wars against the western Liao (biography in Sung shi, Ch. 290), who died in 1057, employed it on horseback. 252 CHINESE CLAY FIGURES Different from chain mail, though allied to it, is the ring mail. The Wu pei chi, as far as I know, is the only source to inform us of the existence of this type of armor in China (Fig. 41). The cut of this book is here reproduced, not only because it is unique in the representation of this specimen, but also because it is very instructive in showing us again how difficult it is to draw inferences from oriental illustrations as to the real type of armor intended by the artist. Any expert in armor, casting a glance at this sketch furnished by the Ming edition of the Wu pei chi, could voice no other opinion than that it is meant to represent a type of scale armor. But the author, as plainly stated in the heading, means to represent a ring armor made of steel wire; and the description added by him leaves no doubt of this intention. He states that "armor of connected rings wrought from steel wire was formerly made by the Si K'iang, and that the structure of the rings is identical with the large iron wire rings of his time, with openings as big as in a coin; in shape, it is like a sort of shirt, and it is held together above by a collar; it is not open in front, but put on over the head; spears and arrows can hardly ever pierce it and cause wounds." Unfortunately he omits to state what the foundation is to which the rings are fastened; but from the drawing, in which the rings are arranged in overlapping rows, it is necessary to con- clude that they were attached to a solid garment, in the same manner as our ring mail, which consisted of steel rings sewed edgewise upon leather or strong quilted cloth. The name K'iang (No. 1 264) mentioned in this text, as is well known, is a general designation for the multitude of ancient Tibetan tribes, at a time when they were still settled in the western parts of Chinese ter- ritory. A. WvLiE 1 has translated from the Annals of the Later Han Dynasty the records pertaining to them. They were exterminated by the Han dynasty. 2 The Chinese tradition tracing ring mail to Tibetan tribes is significant, though it is not necessary to adopt the opinion that the latter ever really made it. Yet the fact remains that ring mail still occurs among the Tibetans. There is even a Chinese source of the middle of the eighteenth century alluding to it. In the Si-tsang ki ("Records of Tibet"), a small but interesting work on Tibet in two volumes, published in 1751 by Chu K'i-tang (Ch. i, p. 23), three kinds of armor in use among the Tibetan soldiers are enumerated, the scale armor (liu ye, "willow-leaves"), the ring armor (lien huan, "connected 1 History of the Western Kiang (Revue de V Extreme-Orient, Vol. I, 1883, pp. 424- 478). 1 CHAVANNES, Les M&noires historiques de Se-ma Ts'ien, Vol. Ill, pp. 591, 595; and Trois g^neYaux chinois (T'oung Poo, 1906, pp. 256-258). HISTORY OF CHAIN MAIL AND RING MAIL 253 rings"), and the chain armor (so-tse). 1 This naturally carries us to Tibet and its relations to Persia in the matter of chain mail ; but before taking leave of China, it should be emphasized that chain mail remains the only type of armor borrowed and imported by her directly from a foreign country. With this exception, the making of armor, though foreign impulses cannot be denied, is purely indigenous, and also Chinese in its essential characteristics. From a negative point of view, its in- dependence from the west is exhibited by several features that are lack- ing in Chinese, but which occur in western armor: as, for instance, the curious nasal (or nose-guard), characteristic of Persian, Indian, and Turkish helmets (Plates XXV and XXVIII) ; and gauntlets, absent in China, but met in Persia, India, and Japan. The Persians seem to have had relations with Tibet at an early date. In the "Histoire des Rois des Perses, " translated (from an Arabic source composed between 1017 and 1021) by H.ZOTENBERG (p. 434), Alexander the Great is made to undertake an expedition into Tibet, whose king offers him submission and a tribute of a hundred loads of gold and a thousand ounces of musk. The two products of Tibet most eagerly solicited by the Persians are clearly emphasized in this legend. Among the wonders possessed by King Abarwlz figured the "malleable gold" extracted for him from a mine of Tibet (ibid., p. 700); this was a block of gold five hundred grains in weight, flexible like wax; when pressed in one's hand, it passed through the fingers and could be modelled; figures were fash- ioned from it, and it would then assume its former shape again. The Annals of the Sui Dynasty 2 have preserved a most interesting account of a country styled Fu, situated over two thousand li north-west of Sze-ch'uan. As I hope to show in detail on a future occasion, the question here is of a Tibetan tribe with a thoroughly Tibetan culture. The particular point that interests us in this connection is that this tribe of Fu possessed helmets and body armors of varnished hide, and that armor played a significant part in its funeral ceremonies. The corpse was placed on a high couch; it was washed, and dressed with helmet and cuirass; and furs were piled upon it. The sons and grand- sons of the dead man, without wailing, donned their cuirasses, and per- formed a sword-dance, while exclaiming, "Our father has been carried away by a demon! Let us avenge this wrong and slay the demon!" 1 As the Tibetans, even less than the Chinese, can be credited with the manuf acture of chain mail, and as Tibetan chain mail is plainly stamped as a Persian import, suspicion is ripe that also Tibetan (and consequently Chinese) ring mails are derived from the same source; but strict evidence for the antiquity of ring mail in Iran yet remains to be brought forward. 1 Sui shu, Ch. 83, p. 8. 254 CHINESE CLAY FIGURES This truly was the burial rite of a militant and valiant people, the dead being believed to continue their lives as warriors, and the survivors combating with their arms the demon who was supposed to have swept him away. A similar idea was symbolically expressed on the burial- places of the Tibetan heroes, who during the age of the T'ang had fallen in their bitter strifes with the Chinese. As related in the T'ang Annals, white tigers were painted on the red-plastered walls of the buildings belonging to their sepulchral mounds scattered along the upper course of the Yellow River: when alive, they donned a tiger-skin in battle, so the tiger was the emblem of their bravery after death. 1 The Tibetans were a warlike nation in the early period of their history, and at times the terror of their neighbors, even of China. The Annals of the T'ang Dynasty, 2 which call them T'u-po (Tibetan Bod), and describe at length their relations with the empire from the seventh to the ninth century, praise their armor and helmets as excellent, covering the entire body, and leaving openings for the eyes only; 3 so that power- ful bows and sharp swords cannot wound them very much. This pass- age, however brief, allows the inference that Tibetan armor of that period was of iron (for it is designated with the word k'ai, No. 5798) ; that it was a complete armor with brassards, ctdshes, and greaves; and that the helmet was provided with a visor. 4 The "gold" armor, 5 which King Srong-btsan sgam-po, according to T'ang shu, is said to have transmitted as a gift to the Emperor T'ai-tsung when he wooed the hand of a Chinese princess, is perhaps not to be taken too literally ; the word kin may simply mean "metal." 6 Among the eastern Tibetan tribes we have proof for the existence of iron armor as early as the sixth century. The Pei shi 7 imparts the interesting news that in the first year of the period Pao-ting of the Pei Chou dynasty (561 A.D.) the Pai-lan, a tribe of the K'iang, who in matters of customs and products agreed with the Tang-ch'ang, 8 sent 1 T'oung Poo, 1914, p. 77. 1 T'ang shu, Ch. 216 A, p. i b. 1 A striking analogy with the Persian helmet as described by Ammianus Mar- cellinus (above, p. 240). 4 Presumably of a similar type as the royal Persian helmet figured by J. DE MORGAN (Mission scientifique en Perse, Vol. IV, p. 320, Paris, 1897). 6 Thus translated by S. W. BUSHELL, The Early History of Tibet, p. 10 (reprint from Journal Royal Asiatic Society, 1880). 6 A golden (huang kin) armor, referring to the T'ang period, is mentioned in Ming huang tsa lu (Ch. B, p. 2). 7 Ch. 96, p. 9 b. 8 Regarding these tribes compare S. W. BUSHELL (The Early History of Tibet, p. 94), and W. W. ROCKHILL (The Land of the Lamas, p. 337). Tibetan armor has not infrequently been sent to China ; specimens are preserved, and may still be seen HISTORY OF CHAIN MAIL AND RING MAIL 255 envoys with a tribute of cuirasses made from rhinoceros-hide (si kia) and iron armor (fie k'ai). There is a somewhat vague Tibetan tradition relative to the period of the early legendary kings, to the effect that armor was first introduced into central Tibet from Lower K'ams (Mar K'ams) in the eastern part of the country. l It is difficult to decide as to what type of armor is to be understood in this passage, in which occurs the general word k'rab, the original meaning of which, as we tried to show (p. i9$), 2 must have been "scale armor." It may be permissible to think, in this case, of a style of hide armor, as it was in vogue among the Fu and the neighboring Shan and Man ; but the tradition which here crops out is somewhat weak and hazy. Coats of mail are frequently alluded to in Tibetan epic literature and historical records. In the History of the Kings of Ladakh they are mentioned under the reign of the seventeenth king, bLo-gros C'og-ldan, as being brought from Guge, eighteen in number; the most excellent of them receiving individual names, as was the case also with swords, saddles, turquoises, and other precious objects. 3 The usual types of armor in Ladakh were chain or scale armor. The fact that they are recorded as coming from Guge is significant, for Guge must have had ancient relations with Persia; 4 and the chain mail of Guge was most probably of Persian origin. The plain fact remains that the Tibetan blacksmiths do not turn out iron chain mail, nor are they capable of making it; so that they are most unlikely ever to have made it at any earlier time. The supposition of an import is therefore the only solu- tion of the problem. The Wei Tsang t'u chi, a description of Tibet by Ma Shao-yun and Mei Si-she"ng written in 1792, has the following note on the outfits of in many Lama temples. The Ming shi tells of a tribute of armor, swords, and products sent in 1374 by the country of Ngan-ting in the territory of the Kuku-nOr, which was classified among the Si Fan (BRETSCHNEIDER, China Review, Vol. V, p. 32). 1 CHANDRA DAS, in Journal Asiatic Society of Bengal, 1881, pt. i, p. 214. 1 B. HOUGHTON (Outlines of Tibeto-Burman Linguistic Palaeontology, Journal Royal Asiatic Society, 1896, p. 41), in pointing out the coincidence of Tibetan k'rab and Burmese k'yap, remarks that each word denotes originally a flat, thin thing or scale, and that hence they come to mean scale armor. "It is, of course, possible," he adds, "that this was possessed by the Burmans in Tibet, but on the other hand it is equally probable that the words have been applied independently on the introduc- tion of this particular kind of armor, (? from China)." This view seems forced. The words k'rab and k'yap are not loan-words from Chinese, but on equal footing with Chinese kia and kiai, and speak in favor of scale armor having been a very ancient means of defence in the Indo-Chinese group of peoples. 'Compare MARX, in Journal Asiatic Society of Bengal, Vol. LX, pt. I, 1891, pp. 122, 123. Also among the ancient Arabs, excellent armors were named (SCHWARZ- LOSE, Die Waffen der alten Araber, p. 69). 4 LAUFER, Toung Pao, 1908, p. 13. 256 CHINESE CLAY FIGURES the Tibetan army of that time: 1 "When the troops go on an expedition, they wear armor consisting of helmets and cuirasses. The latter are made of iron scales 2 or of chains. On the helmet of the cavalry is attached a red crest or a peacock-feather. From their waist hangs a. sword, on their back is slung a gun, and in their hand they carry a pike. On the infantry helmet is a cock's feather. They have hanging to their waist a sword, without counting a dirk. Under their arm is a bow and arrow, and in their hand a buckler of rattan or wood. Some also bear a pike in their hand. Their wooden bucklers measure one foot six inches across, and three feet one or two inches in length, and are painted with pictures of tigers, and ornamented with different-colored feathers; * outside they are covered with sheet iron." If the assumption is correct that Tibetan chain mail is Persian in origin, the scale armor would remain to be looked upon as the national body armor of Tibet, at least as the older type which preceded the in- troduction of chain mail. 4 In former times, it seems to me, the latter was traded over a direct route from Persia into Guge in western Tibet, on the same path along which religious ideas of the Zoroastrians poured in and exerted a deep influence on the shaping of the Tibetan Bon re- ligion, while during the last centuries northern India became the mart which supplied Tibet with this much-craved article. The Tibetan and Persian relations in matters of arms are expressed also by the identity of the Tibetan and old-Persian sword. Indeed, the Tibetan sword, as still in use at present, is the same as that re- 1 ROCKHILL, Journal Royal Asiatic Society, 1891, p. 215. 2 Mr. ROCKHILL has, "made of linked willow-leaf (shaped iron plates)." But the expression liu ye ("willow-leaf"), as we see from the regulations of the Ming dynasty, refers to scale armor, not to plate armor. Mr. WADDELL (Lhasa and its Mysteries, p. 168) speaks of cuirasses consisting of small, narrow, willow-like leaves about an inch and a half long, threaded with leather thongs, still worn by Tibetan soldiers, a few of whom also wear coats of chain mail. The Chinese physician Dr. Shaoching H. Chuan, who visited Lhasa with the Chinese Mission to Tibet in 1906- 1907 has written a very interesting and well-illustrated article on Lhasa under the title The Most Extraordinary City in the World (Nat. Geogr. Mag., 1912, pp. 959 995) ; on pp. 978 and 980 are good illustrations of Tibetan soldiers wearing chain mail. * In the Tower Armory there is a shield of the Angami-Naga, faced with bear- skin, the side ornamented with tufts of feathers (HEWITT, Official Catalogue of the Tower Armories, p. 100). Compare p. 210. 4 In ancient India, likewise, scale armor seems to represent the older type. The fukraniti describes solely this type of armor by saying that "armor consists of scales of the breadth of a grain of wheat, is of metal and firm, has a protection for the head, and is ornamented on the upper part of the body " (G. OPPERT, On the Weapons, Army Organization, and Political Maxims of the Ancient Hindus, p. 109, Madras, 1880). A suit of Tibetan scale armor is illustrated by A. GEORGI (Alphabetum Tibetanum, Rome, 1762, Plate IV) in the figure of a shaman, entitled do kion (that is, c'os skyong, "protector of religion"). HISTORY OF CHAIN MAIL AND RING MAIL 257 constructed by J. DE MORGAN l after a bas-relief of Takht-i-Bostan, both in its shape and in the style of its decoration, for which inlaid stones were employed. The history of the sword, however, is somewhat dif- ferent from that of chain armor, and is not connected with an importa- tion of swords from Persia into Tibet. The swords of the Turkish tribes of Central Asia, to which the Tibetan swords are related, must be taken equally into consideration ; and it seems that this type of sword is a common property of the whole group, of such great antiquity that the accurate history of its distribution can no longer be traced. 2 The Tibetans make (or rather, made) use also of the circular and convex rhinoceros-hide shield of Indian manufacture, ornamented with four brass bosses (Plate XXVII, Fig. i). 3 This shield is employed like- wise in Burma and Siam. The national Tibetan shield is made from rattan plaited in the basketry style of circular coils (Plate XXVII, Fig. 2). Of what type the shield of the ancient Tibetans (K'iang), adopted by the Chinese, was (p. 188), we do not know. Also the Tibetan helmet (Plate XXVIII), composed of steel sheets incrusted with gold and silver wire, forming floral designs, and with attached coif of mail and sliding nasal, is of Indo-Persian origin (com- pare Plate XXV). 1 Mission scientifique en Perse, Vol. IV, p. 321 (Paris, 1897). Compare this volume, p. 15. 1 The swords represented on the monuments of Turkistan belong to the same type (see A. GRttNWEDEL, Altbuddhistische Kultstatten, pp. 26, 27, and many other examples). 1 For Indian specimens see W. EGERTON, An Illustrated Handbook of Indian Arms, pp. 95, in, 118, 134 (London, 1880). Rhinoceros-hide shields are mentioned in the Ain I Akbari of Abul Fazl Allami (translation of H. S. JARRETT, Vol. II, p. 281, Calcutta, 1891). V. THE PROBLEM OF PLATE ARMOR "The skilful leader subdues the enemy's troops without any fighting; he captures their cities without laying siege to them; he overthrows their kingdom without lengthy operations in the field. With his forces intact he will dis- pute the mastery of the Empire, and thus, without losing a man, his triumph will be complete." SUN-TSE, Art of War (translation of LIONEL GILES). We had occasion to allude to plate armor l in the chapter on defensive armor of the Han period, stating that in all probability it existed in the China of those days; we referred also to its possible occurrence among the armor worn by the cataphracti of the ancients, and figured a Siberian petroglyph from the Yenisei representing a mounted lancer clad with such mail. We now propose to discuss this problem in detail, a problem of fundamental historical importance, as it reveals ancient relations between many peoples of Asia, and touches also the question as to the connection of Asiatic with American cultures. Classical and other archaeologists have not yet ventilated this problem, apparently for the only reason that they did not sharply enough discriminate between the various types of body armor. "Scale armor" was the catchword under which everything of this sort was pressed together. 2 But plate armor must be strictly differentiated from scale armor as a special type, which sprang up independently. The laminae forming plate armor are rectangular and flat, and mutually lashed together; and in the same manner the parallel horizontal rows are connected one with another. Such connection is absent in scale armor, in which each scale is individ- ually treated and attached to a background; the background is in this case a necessity, while in plate armor it is dispensable. The laminae of scale armor are arranged like roofing-tiles or the scales of a fish, one placed above another; while in plate armor the laminae, as a rule, are disposed one beside another, or but slightly overlapping. Plate 1 The word "plate armor" is used here throughout in the sense adopted by American ethnologists, armor consisting of horizontal rows of narrow, rectangular laminae (regardless of the material), the single laminae or plates being mutually lashed to- gether by means of thongs, and the various rows being connected in a similar man- ner. Students of European armor usually take the term "plate armor" to designate armor composed of large sheets of metal closely enveloping chest and back. This type is here styled "sheet armor." 1 In England, plate armor is usually styled "scale armor. " E. H. MINNS (Scythians and Greeks, p. 74, Cambridge, 1913), for instance, speaks of "a system of thongs plaited and intertwined as in Japanese and Tibetan scale armor." This, of course, is plate armor; scales are never intertwined. 258 THE PROBLEM OF PLATE ARMOR 259 armor is more flexible and lighter in weight, and hence recommended itself to all nations who became acquainted with it. Plate armor can be easily donned over or beneath any garment, and does away with the uncomfortable leather jerkin. For this reason it proved the most fa- vorite and enduring type of armor in China. It was capable of develop- ment and refinement, while scale armor always remained stationary. It is the ethnologists who were the first to place us on the track of this subject; and there are chiefly two scholars, Friedrich Ratzel and Walter Hough, who took the leadership in this research. Our best course will therefore be to begin by reviewing their studies of the subject, and then to see how their results compare with the new material now at our disposal. FRIEDRICH RATZEL J was the first to make a thorough investigation of the geographical dissemination of plate armor, as far as the material was accessible in his time (1886), among the tribes of north-western America and the Chukchi, also on the Society, Austral, and Gilbert Islands in the South Sea. He was particularly struck by the observa- tion that such armor was lacking in other parts of the world, and that its appearance in the Arctic regions was out of proportion to the general poverty of culture there prevailing. The belief in its independent existence among these peoples conflicted with his axiom that the in- dolence of inventive power is a fundamental law of the primitive stages of ethnic life. In order to explain the phenomenon of plate armor, Ratzel had recourse to Japan, where he deemed armor had reached its greatest development, 2 and where the threads of ancient tribal connec- tions indicated by these peculiar productions ran together; and he believed in a direct contact between Japan and the north-west coast of America in the distribution of plate armor, to the exclusion of the Asiatic Continent. Although the result of this investigation is seemingly historical, the methods and the point of view pursued are purely geo- graphical; and an historical mind cannot fail to notice the weak points of this argumentation. The existence of plate armor in Japan, for in- stance, is merely accepted as a fact given in space, without inquiry into its historical foundation and development, and without the knowl- edge of corresponding objects in China and other parts of Asia being much older. 1 Uber die Stabchenpanzer und ihre Verbreitung im nordpazifischen Gebiet (Sitzungsberichte der Bayerischen Akadcmie der Wissenschaften, 1886, pp. 181-216; 3 plates). * H. SCHURTZ (Urgeschichte der Kultur, p. 355) has adopted the opposite point of view, and interprets that the curious plate armor characteristic of the peoples of the Bering Sea has served as model for the Japanese armor made from lacquered pieces of leather, as certain traditional decorations in the former also seem to prove. This opinion is out of the question, for technical and historical reasons. 260 CHINESE CLAY FIGURES WALTER HOUGH, in his intensely interesting and valuable study "Primitive American Armor," 1 arrives, after a careful survey of the subject, at the conclusion that "plate armor in America is a clear case of the migration of invention, its congeners having been traced from Japan northeastward through the Ainu, Gilyak, 2 and Chukchi, across Bering Strait by the intervening islands to the western Eskimo. Here the armor spread southward from the narrowest part of the strait, passing into the slat armor of the North-west Coast, which is possibly a development of the plate idea. The plate armor also may have spread to the eastern coast of North America. Hence, it appears to be con- 1 Report of the U. S. National Museum for 1893, pp. 625-651 (Washington, 1895; 22 plates). 2 This is a debatable point. J. BATCHELOR (The Ainu of Japan, p. 287, London, 1892) says, "The Ainu also wore armor in their wars; but it was of a very light kind, consisting entirely of leather. Some of them, however, wore Japanese armor which they took from the dead in warfare. This is also one way in which they came by their swords and spears." It seems quite certain that the Ainu have never made any plate armor; and what is found among them of this class is plainly derived from the Japanese. Nor can the Gilyak be credited with plate armor. The only specimen of iron plate armor ever discovered in this tribe, and figured and described by L. v. SCHRENCK (Reisen und Forschungen im Amur-Lande, Vol. Ill, p. 573), is, as SCHRENCK says, of Manchu origin ; and he adds expressly that the iron armors, according to the unanimous statement of the Gilyak, originate from the Manchu. Dr. HOUGH, who has reproduced Schrenck's drawing of the helmet and of a piece of the armor, seems to have overlooked the description in Schrenck's text, though also on the plate the attribute "old Manchu" is added to both specimens, in contradistinction to the indigenous real Gilyak armor coat plaited from fibre. The Gilyak, therefore, cannot be cited, as Dr. HOUGH has done, as a stepping-stone in the migration of plate armor from Japan to the Eskimo. Also Mr. BOGORAS (The Chukchee, Jesup North Pacific Expedition, Vol. VII, p. 164), whose exactness and carefulness is other- wise deserving of the highest praise, has fallen into the same error by reproducing and describing Schrenck's drawing as "Gilyak armor," without paying attention to Schrenck's text. If, therefore, the statement of Bogoras should be correct, that the shape of the plates, and the manner of connecting them, in an iron armor of the Chukchi, are quite similar to those observed on the remnants of this "Gilyak armor," this would seem to say that the Chukchi armor in question would have to be con- nected with Chinese, and not with Japanese culture, as Mr. BOGORAS is tempted to believe; it will be seen on the following pages that other weighty reasons militate strongly against this Japanese theory. SCHRENCK, beyond any doubt, is correct in his statement; and his result agrees with my own inquiries among the Gilyak for armor, and also with my study of Chinese armor. Only SCHRENCK'S definition of "Manchu" must be modified into "Chinese." This error is excusable, as any in- vestigation of Chinese armor had not been made in his time. The Manchu can- not be credited with any original invention in the matter of armor: they adopted it, like so many other things, from the Chinese; and it can be shown step by step, substantiated by official documents, that the Manchu, as in numerous otjher matters, have also faithfully copied the military equipment established by the Ming dynasty. There is no Manchu type of armor which has not yet existed in, and could not be derived from, the Ming period. SCHRENCK'S Gilyak armor, accordingly, is plainly a modern Chinese specimen, that must forfeit any claim to the historical utilization, to which it has been submitted; it cannot be brought into relation with Japan, nor with die Chukchi, nor with the Eskimo. This ethnographical continuity asserted by HOUGH cannot be proved, nor does it in fact exist. RATZEL (/. c., p. 214) had just- ly emphasized the entire lack of plate armor among the peoples of Yezo, Saghalin, and the adjacent mainland. Thus the Japanese theories of Ratzel and Hough, though reaching the same end, materially differ in point of construction. THE PROBLEM OF PLATE ARMOR 261 elusive that plate armor in America had Asiatic origin." On p. 633 Dr. HOUGH states as follows: "The hoop or band armor mentioned as type 4 is found only on the Siberian side of this area and, as well as the plate armor, recalls well-known forms in Japan. This hoop armor is interesting as showing the reproduction of plate armor types in skin, being made of horizontal bands of sealskin instead of rows of ivory plates, the rings telescoping together when the armor is not in use." In describing Eskimo armor made of five imbricating rows of plates of walrus ivory, Dr. HOUGH observes that in the form, lashing, and ad- justment of the plates it is identical with certain types of Japanese armor. 1 His conclusions are the more remarkable, as the previous investigation of Ratzel was unknown to him, and his result has apparent- ly been attained independently. We are here confronted with the interesting case that two ethnographers of high standing have made a notable and praiseworthy attempt to apply an historical point of view to a purely ethnographical situation, with a result so tempting and seemingly convincing that some of the best representatives of our science have readily accepted it. 2 But in the light of a plain historical fact, the position taken by Ratzel and Hough in this question becomes untenable. 3 1 Compare also HOUGH (American Anthropologist, Vol. XIV, 1912, p. 40). 1 BOGORAS (/. c., p. 162), for instance, seems to accept Hough's results; the Chukchi hoop armor is, to him, "evidently an imitation in skin of plate armor" (repeated after HOUGH, p. 633). R. ANDREE (Globus, Vol. 69, 1896, p. 82) acceded to the theory of Hough. 1 This case well illustrates the difficulty of historical reconstructions built ex- clusively on the basis of observed data of purely geographical and ethnographical character. As soon as the Ijght of authenticated historical facts is obtained, our preconceived assumptions and conclusions will always be subject to considerable modifications. In my opinion it is therefore impossible to elaborate with assured results historical reconstructions founded on purely ethnological data. Our mind, owing to our scientific training, can evolve only a logical sequence of thoughts, and interpret given data in a highly logical manner only; but history itself is not logical; on the contrary, it is irrational and erratic, moving in zigzag lines, like lightning; it is a labyrinth of dark passages running in all directions; and, above all, it is more imaginative than the boldest flight of our fancy could possibly be. The unexpected, the unforeseen, has always happened ; and this is what cannot be supplied or supple- mented by the logic of our rational mind. Reconstructions certainly are justifiable and should be attempted, but must never be taken as a substitute for history, or even as real history; they will always remain more or less subjective and problemat- ical, and may be of value as a working hypothesis. It should never be forgotten, however, that the subjective criterion of conceivableness or plausibility, or of an appeal to our common sense, will but seldom prove before historical facts. The rule may even be laid down that whatever may appear to our conception as quite natural, self-evident, or logical, may hardly ever have happened that way, or need not have happened that way, but otherwise. Our knowledge of most subjects is still too meagre to allow at the present time of culture-historical reconstructions embracing a wide area of the globe. To these belongs also the theme of plate armor, tiie specific history of which must first be traced in the single culture zones where it occurs, before its general history can be built up with any encouraging result. Plate armor 262 CHINESE CLAY FIGURES In the north-east of China, beyond the boundaries of Korea, in the east conterminous with the ocean, the northern limit being unknown, we find from very remote ages the habitat of a most interesting people, the Su-she"n, who have greatly stirred the imagination of Chinese and Japanese chroniclers. They were the Vikings of the East, raiding on several occasions the coasts of northern Japan, and fighting many a sea-battle with the Japanese in the seventh century. 1 For a thou- sand years prior to that time, the Chinese were acquainted with this tribe and its peculiar culture: even Confucius is said to have been posted in regard to them, and to have been aware of the fact that they availed themselves of flint arrowheads, usually poisoned, which were then preserved as curiosities in the royal treasury of China. From Chinese records we can establish the fact that the Su-she'n lived through a stone age for at least fifteen hundred years down to the middle ages, when they became merged in the great flood of roaming Tungusian tribes. They had also stone axes, which played a rdle in their religious, worship. A mere supposition is that they belonged to the Tungusian stock of peoples; yet this remains to be ascertained. They may as well have been related to one of the numerous groups of tribes occupying ancient Korea, or, which is still more likely, to the so-called Palae- Asiatic tribes of the North-Pacific region; but the whole ancient eth- nology of north-eastern Asia remains as yet to be investigated. Under the year 262 A.D. it is on record in the Annals of the Three Kingdoms 2 that the Su-shen presented to the Court of China a tribute of a mixed lot of harness, altogether twenty pieces, including armor made of leather or hide, of bone, and of iron, with the addition of four hundred sable-skins. 3 On the iron armor, which was foreign to the culture of the certainly is not by any means so rigidly restricted as assumed by Ratzel and Hough; it will be seen that it takes its place in China, western Asia, ancient Siberia and Turkistan, where it is assuredly much older than in Japan. 1 Compare Jade, p. 59. The Han Annals state that the Yi-lou, another name for the Su-sh6n, were fond of making piratical raids in boats; the Wo-tsu settled in the north-eastern part of Korea, and bordering in the south on that tribe, "dreaded it so much that every summer they were wont to hide in the precipitous caves until winter, when navigation was impossible, at which time they came down to occupy their settlements" (E. H. PARKER, Transactions Asiatic Society of Japan, Vol. XVIII,, 1890, p. 201). In the same study of Parker (pp. 173 et seq.) a history of the Su-shSn will be found. 2 San kuo chi, Wei chi, Ch. 4, p. 133 (compare T'oung Pao, 1913, p. 347). * I am inclined to understand this passage in the sense that there were three dis- tinct kinds of armor, made entirely either of leather, or of bone, or of iron. It is impossible to presume that bone was used in connection with iron in the make-up of one and the same suit of armor. The iron armor, we are forced to conclude, must have formed an individual type in itself, and assuredly one alien to the culture of the Su-sh6n, who, we know with certainty, were not acquainted with the technique of metals for an extended period, and availed themselves of flint arrowheads. Before going to press, I notice from the work of R. and K. TORII (Etudes arche"ologiques,. THE PROBLEM OF PLATE ARMOR 263 Su-she'n, I shall comment later. Hide armor and bone armor formed the national harness of the Su-she'n, as we may infer from another memorable passage in the Annals of the Tsin Dynasty l relating to the period 265-419 A.D., where the characteristic arms of the tribe are enumerated as wooden bows, stone crossbows, hide and bone armor. 2 It is remarkable that the Chinese do not ascribe bone armor to any other of the numerous tribes, with whom they became familiar during their long history, and whose culture they have described to us. In all likelihood, the term "bone armor" occurs in their records only in those two passages; and it is not at all ambiguous. There is but one thing that can be understood by it, the well-known type of bone armor, as it still occurs among the tribes occupying the northern shores of the Pacific on the Asiatic and American sides, particularly among the Chukchi and Eskimo, and in that region exclusively. 3 The Eskimo ivory plate armor represented on Plate XXIX will give some idea of what the Su-she'n Journal of the College of Science, Vol. 36, Tokyo, March 29, 1914, p. 73), which has just reached me, that the two Japanese authors understand this passage in exactly the same sense. 1 Tsin shu (compiled under the T'ang dynasty by Fang K'iao and others), Ch. 97, p. 2 b. 1 The question in this passage, accordingly, is of the armor, offensive and de- fensive, possessed and made by the Su-shen in the beginning of the middle ages. Hide and bone armor are attributed to them, while iron armor is not mentioned. The text might be construed to mean that the Su-shen possessed but a single type of armor, composed of both bone and leather; that is, plates of bone lashed together by means of hide thongs; bone armor is unthinkable without such a ligament, but this consideration need not preclude the assumption that the Su-shen fabricated also pure bide armor. The ethnographical fact that in the culture-area to which this tribe belonged hide and bone armor still occur side by side, must be equally considered in this question; and for this reason we may well understand the passage of the Tsin Annals in the sense that the Su-shen had hide or leather armor, and bone armor. But this point of view is of minor importance. The same passage in the Tsin shu indicates a tribute sent by the Su-shen toward the end of the period King-yuan (260-264) an d consisting of arrows, stone crossbows, armor, and sable-skins. What kind of armor it was on this occasion is not specified; but the general word kia refers to a hide armor or cuirass. J. KLAPROTH (Tableaux historiques de 1'Asie, p. 85) attributes "cuirasses made from skin and covered with bone" to the Yi-lou; the latter are identical with the Su-shen, and the text from which Klaproth translated must be the same as that of the Tsin shu referred to above. The text relative to the Yi-lou inserted ia Hou Han shu (Ch. 115, p. 2 b) makes no allusion whatever to armor, but I am not inclined to infer from this silence that the Yi-lou or Su-shen lacked armor in the Han period. 1 As stated by me in Toung Poo (1913, p. 349), the plates of this bone armor were presumably carved from walrus ivory, in the same manner as in the present Eskimo and Chukchi plate armor. Dr. W. HOUGH of the U. S. National Museum in Washington, to whom I addressed the question as to whether ivory or ordinary bone was utilized to a larger extent in these pieces has been good enough to write me as follows: "The Eskimo armor in the Museum and such suits as I have seen are mostly made of walrus ivory, and so far as I can remember, there are no combinations of ivory and bone in the same piece. On the other hand, there are fragmentary parts of armor from St. Lawrence Island and from the Alaskan mainland which are made of bone; just what bone I cannot say, probably the whale." 264 CHINESE CLAY FIGURES tribute armor was like. 1 The point here at issue, then, is the fact that the entry of the Chinese annalist, under the year 262, regarding the presentation of bone armor on the part of the Su-she"n, is the earliest recorded reference to bone armor in history, capable of throwing a flashlight on events in the North-Pacific culture area, so glaringly devoid of any records. The date 262 is of far-reaching consequence. Certainly, like all dates where inventions or culture ideas are involved, it is a mere symbol, that requires a certain latitude in its translation. The tribute of 262 indicates that bone armor had been made prior to that date by the Su-she"n, or generally within the culture-zone to which they belonged; and since complex inventions of such character require time to mature, and the laborious efforts of several generations, it is justifiable and reasonable to conclude that the beginnings of the invention go back to a far earlier period. Plate armor of bone must therefore be infinitely older than could heretofore be supposed from the mere circumstantial evidence of present geographical distribution; and it follows also that the geographic area of bone armor must have been much more extended in ancient times, and reached farther south along the shores of Asia. In other words, the culture area under consideration, as it now presents itself to our eyes, must have occupied a larger territory in the times of which we speak, a conclusion confirmed to me also by other reasons; and the Su-shen must have either ranged among the representatives of North-Pacific culture, or have been strongly influenced by it. If as early as 262 the Su-she"n were in possession of bone plate armor, this type of harness cannot be explained as having been made in imitation of Japanese plate armor for the plain reason that Japanese plate armor was at that time not in existence. Metal armor in Japan cannot be pointed out before the close of the eighth century. Fragments of armor consisting of scales of bronze incrusted with gold, and preserved in the Museum of Tokyo, are assigned to about the year 800 A.D. by BASHFORD DEAN, Z our great authority on Japanese armor; while frag- ments of iron plate armor are not older than about 1050 and noo; that 1 The number of perforations in the plates is not always six, as in the specimen illustrated. A large number of detached Eskimo ivory plates in the Field Museum (Cat. No. 34,154) exhibits on an average twelve perforations, two and two being close together. Sometimes a third perforation is added to the two in the corners, and some- times an additional perforation is drilled through the centre of the upper or lower side. A very interesting specimen in our collection (Cat. No. 34,153) is a pair of Eskimo cuisses (leg-guards) of mastodon ivory, 16.5 cm long, with rows of perforations along the top and bottom edges. These objects were obtained by A. M. Baber from the Asiatic Eskimo on the Tchukotsk Peninsula. 1 Catalogue of the Loan Collection of Japanese Armor, pp. 20, 28 (New York, 1903)- THE PROBLEM OF PLATE ARMOR 265 is, they belong to the latter part of the Fujiwara period (900-1100). Before this time, padded coats and hide cuirasses were the usual means of body protection; the latter sometimes assumed the form of scale armor, the scales being cut out of pieces of boiled leather. 1 The Chinese Annals of the Sui Dynasty, 2 in the interesting account on Japan, state that the Japanese (Wo) make armor of varnished leather (tsi p'i wei kid) and arrows of bone. At that time, which, from the standpoint of Japanese development, is designated as the protohistoric or semihistoric period, defensive armor cannot have played any signifi- cant r61e in ancient Japan, as it is conspicuously absent in her two oldest records, the Kojiki (composed in 712 A.D.) and the Nihongi (720 A.D.). S In the year 780 an order was issued by the government that leather ar- mor should be used, because the kind hitherto worn (that is, padded coats) was continually requiring repair. This order permitted, further, the use of iron instead of leather, and advised that all armor should be gradually changed to metal. 4 It is therefore clear that at the time, when our Su-she"n account of bone armor is at stake, the Japanese did not possess any metal or any plate armor, and that it is even question- able whether they then availed themselves of defensive armor at all. We are hence prompted to the conclusion that bone plate armor, being at least from six to eight hundred years older than Japanese plate armor, cannot have been made as a reproduction of the latter, and that Japan cannot be made responsible for it. Thus the whole theory of a con- nection of American and Northeast-Asiatic plate armor with Japan must naturally collapse. If the opinion should be correct of those who believe that American- Asiatic plate armor must have been made in imitation of a form of iron 1 Catalogue of the Loan Collection of Japanese Armor, p. 38 (New York, 1903). According to W. GOWLAND (The Dolmens and Burial Mounds in Japan, p. 47, Westminster, 1897), no bronze armor has as yet been found in the dolmens of Japan; and iron armor, too, is by no means of very common occurrence. 1 Sui shu, Ch. 81 , p. 6 b (also Pei shi, Ch. 94, p. 72). It is notable that the account of Japan in the Annals of the Later Han Dynasty (Ch. 1 15, p. 5 b) makes no mention of body armor, but points out only the shield and the use of offensive weapons, such as spear, wooden bow, and arrows with bamboo shafts and bone heads. Arrows with iron heads employed in Japan are first reported in Tsin shu (Ch. 97, p. 3). 1 O. NACHOD, Geschichte von Japan, Vol. I, p. 155 (Gotha, 1906). But shields are several times mentioned as offerings. The Annals of the Later Han Dynasty, as pointed out, confirm the existence of shields. The idea generally entertained that Japan has had a bronze and an iron age, in my opinion, is erroneous. The bronze and iron objects found in the ancient graves have simply been imported from the mainland, and plainly are, in the majority of cases, of Chinese manuf acture. Many of these, like metal mirrors, certain helmets and others, have been recognized as such; but through comparison with corresponding Chinese material, the same can be proved for the rest. Ancient bronze objects are so scarce in Japan that, even granted they were indigenous, the establishment of a "bronze age' would not be justified, nor is there in the ancient records any positive evidence of the use of bronze. 4 BASHFORD DEAN, 1. c., p. 27. 266 CHINESE CLAY FIGURES armor, two other theoretical considerations could be advanced. There remain the Chinese and the ancient Turks of Siberia and Central Asia; and it might be argued that Chinese or Siberian harness of iron plate could have furnished a suitable model for the Arctic harness-maker. To such a point of view, however, serious objections could be raised; and here again, first of all, on purely historical grounds. The utiliza- tion of iron in the making of armor, as we noticed in Chapter III, does not become apparent in ancient China till as late as the first cen- turies of our era, its beginnings being justly laid by the Chinese in the period of the Later Han dynasty (25-220 A.D.; see p. 210), and thus it appears from inward evidence. This primeval iron armor, in all likelihood, was not yet a true iron plate armor, but merely a hide cuirass reinforced by iron laminae; rectangular iron plates may have then existed, but the matter is still problematical. Even presuming that iron plate armor might have obtained during the epoch of the Later Han, for which there is as yet no positive evidence, we should be forced to infer that the developments of the ancient Chinese iron armor and the northern bone armor, in this case, have necessarily been contem- poraneous events. The tribute of the Su-shdn bone armor in 262 A.D. is separated from the closing year of the Han period in 220 A.D. only by the brief span of forty-two years; that is, the average duration of a generation. If, accordingly, these two developments should have run parallel to each other in point of time in two widely different culture areas which otherwise had not a single point in common, the inference would have to be drawn that these two developments have taken place independently, and may have each been prompted by factors coming from a different quarter. In the present state of our knowledge it is safe to assume that bone armor in north-eastern Asia is as old as, or even older than, any iron plate armor in China or Korea. If an outward impetus to the making of bone armor in that region must be assumed, I am disposed to believe that it came from the interior of Siberia. 1 In regard to ancient Siberian armor, our information is exceedingly scanty. Only traces of plates of armor have been dis- covered in graves on the Berel, 2 and a famous petroglyph on the Yenisei depicts to us a horseman armed with lance and mail-clad (Fig. 35). The long continuity of the iron age in Siberia renders it impossible at 1 For evidence see below, p. 274. 1 W. RADLOFF, Aus Sibirien, Vol. II, p. 130. Also in Siberia iron armor may have formed the exception, while hide, as the cheaper material, always maintained its place. MARCO POLO (ed. of YULE and CORDIER, Vol. I, p. 260) says concerning the Tartar (that is, Mongol) customs of war, "On their backs they wear armor of cuirbouly [boiled leather], prepared from buffalo and other hides, which is very strong." THE PROBLEM OF PLATE ARMOR 267 the present time to fix a date for these antiquities with any degree of certainty ; but a general deduction may be hazarded. There are good reasons for assuming that the Chinese derived their iron armor from Turkish and Iranian peoples, first, because their knowledge of smelt- ing and forging iron came from them; and, second, because their own inventiveness in defensive and offensive armor was rather poor, and be- cause others of their weapons, like swords and daggers, were adopted from the same group (p. 215). The sudden appearance of iron armor in the Later Han dynasty speaks in favor of this view; and as only copper plate armor was known in the preceding period of the Former Han dynasty, it seems very likely that iron armor among the Turkish tribes was not much older than in China. As previously stated, the Su-she"n sent iron armor along with skin and bone armor to China, but only the latter two types formed their national armor, according to the later re- port of the Annals of the Tsin Dynasty. The occasional introduction of iron armor, consequently, did not suppress among them the employment of skin and bone armor; and although iron armor was known to them at the end of the third century, they adhered, for several centuries down- ward, to bone and hide, that seem to have represented a more efficient means of defence at that time than iron armor, the making of which must still have been in a primitive and experimental stage. On the other hand, in opposition to this theory of a foreign influence, it must be emphasized that the culture types of north-eastern Asia, on the whole, have strong and pronounced characteristics which have hardly any parallels in the rest of the Asiatic world, and that owing to geograph- ical conditions the entire area has remained purer and more intact from outside currents than any other culture group in Asia. The profound researches of Bogoras and Jochelson have shown us that in language, folk-lore, religion, and material culture, the affinities of the Chukchi, Koryak, Yukagir, and Kamchadal go with Americans, not with Asiatics. In fact, Turkish-Mongol influence on these tribes is exceedingly small; Chinese influence, if any, amounts to a minimum ; l and the alleged Japa- 1 While the Chinese, owing to political circumstances, were comparatively well acquainted with the tribes inhabiting Manchuria, Korea, and the Amur region, their knowledge of the tribes beyond has always been very limited. Their first acquaint- ance with the Ainu dates from the year 659 A.p., when some members of this tribe accompanying a Japanese embassy made their appearance at the Court of the Emperor Kao-tsung (650-683) of the T'ang dynasty; they are described on this occasion as "forming a small country on an island in the ocean, having beards four feet long, being clever archers, and sticking arrows through their hair; they have a man hold an arrow (according to another reading, a vessel) which they use as a tar- get at a distance of ten paces, without missing their aim " (Tang shu, Ch. 220, p. 1 1 ; and Yen kien lei han, Ch. 231, p. 47). They are called by their Japanese name Yemishi (Chinese, Hia-i). This embassy is mentioned under the same year also in the Japanese Nihongi (AsTON, Nihongi, Vol. II, p. 260), where it is said that the 268 CHINESE CLAY FIGURES nese influence is a chimera. Plate armor, if due in that region to a stimulus received from outside, would represent a somewhat isolated instance of historical contact in the line of warfare; 1 and whatever the psychology of this first stimulus may have been, I venture to deny that it ever operated in the haphazard and purely external manner indicated by Ratzel and Hough, a certain independent course of development in that area cannot be absolutely denied. While I am very far from contesting that historical interrelations may have been at play in the dissemination of the plate idea in north- eastern Asia, I wish to maintain for the present an attitude of reserve toward this point. The downright failure of the Japanese hypothesis should put us on our guard; and, the imitation theory, I confess, be it formulated with reference to the Japanese, Chinese, or Siberians, does not strike me very favorably. Whatever we may now be inclined to assume in that direction, it will remain mere assumption in our present state of knowledge; and it must be upheld that no imitation theory, with whatever modifications, can be backed up by certain facts. In other words, the problem is not yet susceptible of a definite solution. There is, however, not only an historical, but also a technical side to this question, and we should not entirely lose sight of the technical point. We observe in various culture-groups that plate armor is never a primary type of armor, but occupies a secondary place in point of Japanese took with them a Yemishi man and woman of Michinoku to show to the T'ang Emperor. In the Description of the Tributary Nations of the Ts'ing Dynasty (Huang Ts'ing chi kung t'u, Ch. 3), published under the patronage of the Emperor K'ien-lung, the Ainu are figured and briefly characterized under the name K'u-ye. This is the Gilyak designation Kuhi for the Ainu, identical with the Huye of Du HALDE (Description de 1'empire de la Chine, Vol. IV, p. 15; compare also L. v. SCHRENCK, Reisen und Forschungen, Vol. Ill, p. 129). On some Chinese maps Saghalin is still designated as "Island of K'u-ye." The Gilyak came to the notice of the Chinese at a very late date; they do not seem to be mentioned earlier than in the Se w$n hien t'ung k'ao (published in 1586) under the name Ki (or K'i)-li-mi (Gilami), the name given this people by its Tungusian neighbors (compare A. WYLIE, Chinese Researches, pt. 3, p. 249, who alludes to this passage without identifying the tribe). In the Chinese work previously quoted, the Gilyak are pictured and described under the term Fei-ya-k'a as inhabiting the country to the extreme east of the Sungari, the littoral of the ocean, and scattered over the islands (compare L. v. SCHRENCK, /. c., pp. 100-103). 1 A very interesting case was established by FRANZ BOAS in his study Property Marks of Alaskan Eskimo (American Anthropologist, 1899, pp. 601-613). Property marks are very frequently used by these tribes on weapons employed in hunting with the object of securing property-right in the animal in whose body the weapon bearing the mark is found. It is a remarkable fact that these mjarks occur only among the Eskimo tribes of Alaska, but are not known from any other Eskimo tribe. This fact, taken in connection with the form and occurrence of such marks among the north-eastern tribes of Asia, suggests to Boas that this custom, like so many other peculiarities of Alaskan Eskimo life, may be due to contact with Asiatic tribes. This case is very plausible, and would merit a more profound historical investigation in connection with the practice of tamga now disseminated throughout Siberia. THE PROBLEM OF PLATE ARMOR 269 time; it is always preceded by plainer types, usually cuirasses of hide or cotton, and scale armor. Cuirasses of rhinoceros-skin were utilized in China for thousands of years, before any metal harness became known. In China as well as in Egypt we clearly recognize the inter- mediary stages of hide and plate armor, the surface of the hide being first reinforced by irregular, scale-like metal pieces (first of copper, later of iron), which gradually assumed the standard rectangular plate shape; and then, by removing the hide foundation, the pure metal plate armor sprang up as a new and independent type. The history of defensive and offensive weapons, moreover, is closely interrelated; the eternal game of modern war industry first inventing bullet-proof naval armor-plates, and then the bullets to pierce them was in full swing even in the stages of primitive life. The growing perfection of metal weapons constantly forced man to devise new means of increasing the power of his defensive armor, and this accounts for the coming into existence of ever-varying new types. I am certainly not competent on any subject of American ethnology, and must leave it to our Ameri- canists to reason out the case for themselves. But this much may be said. Nearly everywhere in North America, even in the eastern area, we generally find the type of hide armor, the indigenous development of which is admitted by Dr. Hough and cannot seriously be challenged; thus hide armor may have been the oldest form of body protection in war also in this region. 1 We meet there also the intermediary stages, as, for instance, the wooden cuirass of the Thompson River Indians, covered with elk-hide, described by JAMES TEiT, 2 and the application of wooden slats, of reeds, of bone plates to the exterior or interior of the cuirass, to strengthen it more efficiently, the secondary development. Finally those materials were exclusively utilized in its construction, leading up to pure plate armor as a tertiary and ultimate stage. No fundamental difference can be found in the employment of wood and bone, or ivory, which simply present purely technical changes of mate- rial; and American-Asiatic bone plate armor, after all, might be con- ceived as quite a natural development, which may have arisen inde- pendently, without the contact of an outside culture. Its coming into existence could be explained by the trend of indigenous thought and the 1 "The American savages were acquainted with body armor when they were first encountered. Wherever the elk, the moose, the buffalo, and other great land mammals abounded, there it was possible to cover the body with an impervious suit of raw- hide" (O. T. MASON, The Origins of Invention, p. 390). * The Thompson Indians of British Columbia (Jesup North Pacific Expedition, .. II, p. 265). See also A. P. NIBLACK, The Coas (Report U. S. National Museum, 1888, pp. 268-270). Vol. II, p. 265). See also A. P. NIBLACK, The Coast Indians of Southern Alaska 270 CHINESE CLAY FIGURES inventiveness of the aborigines, which may have resulted in a large variety of ingenious armor spread over an extensive area. 1 There remain other considerations to be made which would seem to confirm this impression. The cut, the style, and the mode of wearing armor in the North-Pacific region are different from those in eastern Asia. The peculiar Chukchi fashion of having the left side covered up and the left arm and hand hidden in the armor, while only the right arm remains free for action, 2 is a striking feature, which is entirely lacking in any other part of Asia. At any rate, I am inclined toward the opinion that the type of bone plate armor under consideration is not exclusively due to an impact of foreign influence. In some form unknown to us it may have pre-existed, before any metal plate armor had reached the Far East ; while I am quite willing to admit that at some later period the regular, rectangular shapes of the ivory plates, and the peculiar method of lashing them together, may be the outcome of an adaptation to some imported model. The memorable passage in the Chinese Annals concerning the Su- sh&i may elucidate still another problem. Their gifts to China in 262 consisted not only of bone armor, but also of iron armor. BOGORAS 3 has shown that ancient iron armor, made of small pieces of iron with fastenings of narrow leather strips, was until recently very common among the Reindeer Chukchi; and he makes it probable that iron was known among them before the arrival of the Russians. And here the Su-she~n come again to our assistance in dispelling the Japanese spectre ; for the question of the origin and manufacture of Chukchi iron armor suggests to Mr. BOGORAS "a connection with the Japanese which does not exist at present," and which in all probability has never existed. Mr. BOGORAS is unable to furnish any evidence for such an alleged inter- course, which is certainly not proved by the occasional occurrence of a modern Japanese article of trade in that region. 4 The facts in the case 1 1 do not mean to say, of course, that the development has actually and ob- jectively taken place that way, but only wish to point out that it may be thus construed in our minds. 2 HOUGH, Plate V; BOGORAS, The Chukchee, p. 163 (shows also a suit of left- handed iron armor). The Chukchee (Jesup North Pacific Expedition, Vol. VII, No. I, p. 54). 4 The statement of Bogoras that the armor and helmet figured on p. 164 are Japanese seems to me to require further proof. It rather conveys the impression of being un- Japanese. Bogoras alludes to the advance of the Japanese to Kamchatka without citing sources in support of this opinion. I presume he must have had in mind the passages of G. W. STELLER (Beschreibung von dem Lande Kamtschatka, pp. 3, 249) saying that the Japanese were long known as traders to the inhabitants of the littoral of the Okhotsk Sea (on the Kamchadal name of the Japanese, see L. v. SCHRENCK, /. c., p. 192). Kamchatka was vaguely known to the Japanese of the eighteenth century, as we see from KLAPROTH'S Apercu ge"n6ral des trois royaumes THE PROBLEM OF PLATE ARMOR 271 are that the Japanese never have penetrated much beyond Saghalin Island, where the southern portion inhabited by the Ainu was their main field of exploitation, while the northern part remained a terra incognita to them. The Japanese have exerted no influence on the culture of the Gilyak settled there, 1 nor is there any Japanese trace on the mainland in the region of the Amur. Even without such considerations, how- ever, the point of view taken by Bogoras in this matter can no longer be upheld. The fact that the Su-she'n possessed knowledge of iron armor in 262 goes to prove that iron armor around that time was within the boundaries of the North-Pacific culture-zone. 2 Again, it must be called to mind that the. Su-she'n iron armor cannot have been of Japanese origin, as iron armor was not then in existence in Japan; neither can it be set in relation with Chinese iron armor, as it would be absurd to suppose that the Su-she'n should have sent Chinese iron armor as tribute to the Chinese Court. Their tribute certainly consisted of curious and valuable objects which were new and impressive to the Chinese. As the Su-she'n were not able to make iron armor, not being acquainted with the technique of smelting and forging iron, they con- sequently must have received it in the channel of trade from an iron- producing region, such as we find in ancient times in the interior of Siberia, in Central Asia, 3 and in the beginning of our era also in (p. 195, Paris, 1832). The Italmen, the ancient Kamchadal, knew the Japanese chiefly as importers of iron needles, and styled these Sis (plural Sisin: I. RADLINSKI, Slownik narzecza Kamczadal6w, p. 72, Cracow, 1892) after Sisam, the Ainu designa- tion of the Japanese. But it is altogether the simple question of a superficial trading relation along the coast by way of the Kuriles; and there is no trace of Japanese influence whatever on the culture of the Kamchadal. 1 Likewise L. v. SCHRENCK (Reisen und Forschungen im Amur-Lande, Vol. Ill, P- 570). * This chapter, as it now stands, was in substance written in the autumn of 1912, an abstract of it having been read at the meeting of the American Anthropological Association held in Cleveland, December, 1912 (see Science, 1913, p. 342, or Am. Anthr., 1913, p. 960). A confirmation of the above conclusion is now furnished by the highly interesting study of R. and K. TORII (I. c., p. 72), who found in eastern Mon- golia a metal (seemingly iron) plate of an armor (4 x 2.5 cm) with four apertures in the long sides. It is correctly diagnosed by the two Japanese authors, who remark that such plates are now dispersed among the ruins left by the Tung Hu ["Eastern Hu," a general Chinese designation for the populace of eastern Siberia], especially in the region of the Shira Muren. This archaeological discovery bears out the fact that iron armor anciently did exist in eastern Siberia, and that it was of the type of plate armor. Thus the supposition is gaining ground that the iron harness in the possession of the Su-shen was iron plate armor, and existed in that region side by side with bone plate armor. Messrs. Torii, in this connection, remind us of the fact that the Wu-huan, according to the Annals of the Later Han Dynasty, are capable of making their bows and arrows, also saddlery, and turn out their own arms from forged iron. 1 It is known that L. v. SCHRENCK (/. c., Vol. Ill, p. 569) attributes to Japanese influence the knowledge of iron-forging among the Ainu and Gilyak. This being an affair of recent origin is certainly not a serious case; these tribes purchase Japanese pig-iron, and work it up into blades for knives. Schrenck's point of view that iron- 272 CHINESE CLAY FIGURES Korea. 1 These considerations are instructive also in that they reveal the baselessness of what might be styled "the Japanese mirage of American ethnology." Not only objects of material culture like plate armor, but also motives of myth and legend, have been traced from America directly to Japan, as, for instance, by the late PAUL EmiENREiCH. 2 This method seems to me inadequate for historical reasons. The primeval culture type of Japan, as we know it, is a comparatively recent production, very recent when contrasted with the great centres of culture developed on the mainland of Asia, and recent even in comparison with all in- digenous cultures found on the American Continent. I mean to say that most phenomena of culture, inclusive of myth and religion, are by far older on this continent, and still preserved in an older form, than any corresponding phenomena in Japanese culture, even if the latter are reduced to their oldest attainable condition. The Kojiki and Nihongi, the main text-books of Japanese mythology, do not present a pure source of genuine Japanese thought, but are retrospective records largely written under Chinese and Korean influence, and echoing in a bewilder- ing medley continental-Asiatic and Malayo-Polynesian traditions. But more than that, it may be safely stated at the present time that the history of American cultures has never had, and never could have had, any relation with Japan, which always was beyond the pale of American-Asiatic relations, and that American ethnology offers no point of contact with Japan. The threads of historical connection run- ning from America into Asia do not terminate in Japan, but first .of all, as far as the times of antiquity are concerned, in a territory which may be defined as the northern parts of modern Manchuria and Korea. From ancient times the varied population of this region has shared to some extent in the cultural elements which go to make up the character- forging among the Gold on the Amur is due to the adjoining Manchu-Chinese, how- ever, is entirely erroneous, as this art doubtless is much older in that region than the rule and influence of the Manchu, and points decidedly in the direction of the Turkish Yakut. Many iron objects of an ornamental character in use among the Gold can be plainly recognized as Yakutan in origin, and Yakut are constantly living and trad- ing in their midst. Neither the Japanese nor the Chinese need be invoked to explain iron-forging in eastern and north-eastern Siberia, as it is much older in the interior of Siberia, where there have been at all times better blacksmiths, forging better iron-work than was ever turned out in China. 1 The Annals of the Later Han Dynasty (Hou Han shu, Ch. 115, p. 5 b) relate that the country Shen-han in Korea produced iron, that the Wei, Wo (Japanese) and Ma-han went there to purchase it on the market, and that iron was the means of barter in all business transactions. There was no iron in the country of the Shi- wei, and they received it from Korea in exchange for sable-skins (Pei shi, Ch. 94, p. 9 b) . The considerable beds of iron ore in Kang- wun Province are still worked by the natives, who scrape it up from the surface of the ground, and smelt it in furnaces by means of charcoal (H. B. HULBERT, The Passing of Korea, p. 274). 1 Die Mythen und Legenden der sudamerikanischen Urvolker, pp. 77 et seq. (Berlin, 1905). THE PROBLEM or PLATE ARMOR 273 istics of the North-Pacific culture-province. It does not suffice for the study of American-Asiatic relations to take into consideration only the present ethnological conditions, as has been done, but the ancient ethnology .of that region must first be reconstructed. From this point, the further contact, if any, may be given, and as our knowledge advances, may eventually be established at a future date (I speak only hypo- thetically) with ancient China on the one hand, and ancient Siberia on the other, relations which would all refer to pre- Japanese times, and move outside of the current of Japan. The early existence of bone armor is one of the examples proving that this view seems to be on the right track, and entitling us to speak of an historic antiquity in North- Pacific culture. A pragmatic history of the development of plate armor cannot yet be written, as the subject has not been thoroughly investigated by specialists in the antiquity of western Asia, and as there are doubtless many missing links still unknown to us. Meanwhile the following in- dications which I have been able to trace may be welcome. In Assyria, plate armor is unmistakably represented on monuments of King Sargon (B.C. 722-705) in connection with foot-archers, whose coats consist of six or seven parallel rows of small rectangular plates. 1 It seems that in Assyria plate mail sprang up during that period, for in the reign of Salmanassar II (B.C. 860-825) the bowmen sculptured in stone are frequently clad with long coats reaching from the neck to the ankles and girdled below the chest, the coats being covered with an irregular checkered design, but not with rows of rectangles. 2 Further, we find metal plate armor in ancient Egypt; 3 there a cuirass of thickly wadded material was covered with metal plates. It is ascribed to the reign of Ramses II, who ruled in the thirteenth century B.C. Also the Shardana armor described by OHNEFALSCH-RiCHTER 4 consisting of bronze plates, two of which are mutually joined by means 1 P. S. P. HANDCOCK (Mesopotamian Archaeology, pp. 350-2), who speaks only of coats of mail. 1 Ibid., pp. 260, 350. * An illustration of it may be seen in A. ERMAN'S Life in Ancient Egypt (p. 545, London, 1894). As a rule, the helmet and body armor did not consist there of metal, being more probably made, as many of the pictures seem to indicate, of thickly wad- ded material, such as is worn even now in the Sudan, and forms an excellent protec- tion. In rare instances, however, defensive armor may have been covered with metal plates. No special investigation of this subject has as vet been made in regard to the two culture zones of Assyria and Egypt; but these indications, however brief, will suffice to show that plate armor must have been widely distributed in ancient times, and that a mere consideration of present conditions alone, as attempted by Ratzel and Hough, cannot bring about the solution of the problem of its history. 4 Zeitschrift fur Ethnologic, Vol. XXXI, 1899 (Verhandlungen, p. 360). 274 CHINESE CLAY FIGURES of hinges, and sewed to a foundation of linen or leather evidently belongs to this category. The most valuable contribution to the question is presented by a number of single bone plates of rectangular shapes, found in barrows about Popovka on the Sula in southern Russia. Five of such plates are reproduced by E. H. MiNNS. 1 As these have perforations (one, two, or three) only at the top and base, we must suppose that they were sewed on to a foundation of cloth or leather; they could not have been lashed together freely without such a background, as in the Chukchi and Eskimo plate armors discussed above. 2 Those with pointed top and a single perforation, having the one side curved and the other straight, formed the ends of a plate-row. This find attests the fact that bone plate armor anciently existed in the western part of the Old World among Scythian tribes; and this case shows that in regard to Northeast- Asiatic and American bone plate armor we need not resort to the theory of explaining it as an imitation of iron in bone. If imitation it is, it may have been Scythian (or Siberian) bone armor (a single piece or several) , which by trade found its way to north-eastern Asia. In the territory of the Scythians we find plate armor not only of bone and horn, but also of bronze and iron; and it seems to me that the adoption, on the part of the Scythians, of the Iranian tactics of cataphracti (p. 220) gave the impetus to the introduction among them of this type of armor. The rock-carving of the mounted lancer on the Yenisei (Fig. 35) demon- strates that plate armor, presumably of iron, had penetrated into Siberia during the iron age. I suspect the institution of cataphracti of being largely responsible for the wide dissemination of this type of armor; it was peculiarly adapted to fighting on horseback, and the Iranian mode of tactics, as we saw in Chapter III, expanded into the Roman Empire, and was adopted by the Huns, to be continued by the Turks (T'u-kue) under the T'ang dynasty. When tactics and cavalry organization spread over the boundaries of Iran, the armature of the cavaliers was necessarily bound to migrate along the same path. The fresco paintings discovered in Turkistan furnish many valuable contributions to the history of body armor, and particularly of plate armor. A. STEIN 3 was the first to correctly recognize this type of armor in a Buddhist statue excavated by him at Dandan-Uiliq. The figure, standing over the body of a prostrate foe, is clothed with a coat of mail reaching below the knees and elaborately decorated. "The gay colors 1 Scythians and Greeks, p. 188 (Cambridge, 1913). 2 In these, perforations likewise run along the long or vertical sides of the plates. 3 Sand-buried Ruins of Khotan, p. 272 (London, 1904); and Ancient Khotan, Vol I, p. 252, Vol. II, Plate II (Oxford, 1907). THE PROBLEM OF PLATE ARMOR 275 of the successive rows of small plates which form the mail, alternately red-blue and red-green, were remarkably well preserved, and not less so all the details of the ornaments which are shown along the front and lower edge of the coat and on the girdle around the waist. Even the arrangement of the rivets which join the plates of mail, and the folds of the garment protruding below the armor, are indicated with great accuracy. There can be no doubt that the artist has carefully re- produced here details of armor and dress, with which he was familiar from his own times." l A rich material for the study of plate mail in the art of Turkistan is offered by the fascinating work of A. GRUNWEDEL,* who himself has clearly recognized and pointed out this armor type. 3 The fact that the plates are painted blue clearly proves that they were wrought from iron. The coats are tight-fitting, and open in front; the sleeves are likewise bedecked with plates, and the shoulders with pauldrons. A further example will be found in the work of A. v. LE Coq. 4 The T'ang period (618-906) is responsible in China for a far-reaching innovation in the line of armor, which has persisted at least down to the end of the eighteenth century, the combination of armor with the military uniform, resulting in a complete armor-costume. Up to that time, armor and garment had been distinct and separate affairs. The ancient hide harnesses were worn over the ordinary clothing or uniform, and were naturally put on only when making ready for battle; while 1 The comparison made by Stein (Ancient Khotan, p. 252) between this armor and that on a Gandhara relief figured by GR^NWEDEL (Buddhist Art of India, p. 96) is not to the point. The two suits of armor are of entirely different types, the former being plate armor; the latter, as correctly interpreted by Grunwedel, scale armor. Stein did not recognize this difference, nor did V. A. SMITH (History of Fine Art in India, p. 122), who copied him on this point. Among the finds made by A. Stein (Ancient Khotan, pp. 374, 411) at Niya, there is a single piece of hard, green leather, shaped and perforated very much like the metal plate of an armor. Stein suggests that "it probably belonged to a scale armor" (he means plate armor), and thinks that this supposition is confirmed by the metal plates of an armor coming from Tibet (p. xvi). This is possible; I do not believe, however, that an entire suit of armor was ever made in Turkistan in this manner, but that only certain parts of an armor suit were of this technique. There would be no sense in producing a complete suit by means of such separate leather laminae, a very toilsome and cumbrous process; any plain hide coat would probably present a more enduring protection than such an affair. Indeed, this technique is known to us from Japan: thus a shoulder-guard believed to date from prior to 1 100 (BASHFORD DEAN, Catalogue of the Loan Collection of Japanese Armor, Fig. 12 B) is made from bands of laminae of boiled leather interlaced with rawhide. Leather laminae, of course, do not present any original state, but are a secondary development, being the outcome of an imita- tion of metal laminae. 1 Altbuddhistische Kultstatten in Chinesisch-Turkistan (Berlin, 1912). 1 L. c., p. 201, and Figs. 451, 452, 456, 460, 512, 513, 628. 4 Chotscho, Plate 48 (Berlin, 1913). 276 CHINESE CLAY FIGURES during the march they were rolled up and carried. l Scale, chain, ring, and plate armor were all a great burden on the body owing to their heavy weight, and a serious obstacle to the mobility of troops. The reform is attributed to Ma Sui, who was president of the Board of War under the Emperor T'ai-tsung of the T'ang dynasty, and who died in 796. 2 He conceived the idea of combining armor with the costume (styled k'ai i, "armor clothing") in three grades differentiated according to length; and the soldiers thus clad were enabled to run, and to advance comfortably. The helmets he made in the form of lions. 3 This in- novation is illustrated by an interesting passage in the Ch'u hio ki, 4 where some new names for the parts of armor are given, derived from the names of clothing. "The skirt attached to the armor is called shang (No. 9734, "the clothes in the lower parts of the body"); the inner side of an armor is styled lei (No. 6843) ; 5 and the coat of the armor (kia i, No. 5385) is termed kao (No. 5949)." 6 The general expression for clothing, i-shang, finds here application to armor: the upper portion of the armor is directly styled i ("upper clothing"), and the term kao used with reference to it plainly indicates that a robe made of some textile material was worn over the mail to cover it all round. This state of affairs is confirmed by the Wan hua ku, 7 where, besides cuirasses and six kinds of iron suits, are enumerated armor made from white cotton stuff (pai pu kia), that made of black silk taffeta (tsao chiian kia), and even wooden armor (mu kia). 8 1 As expressly stated by Sun-tse (see L. GILES, Sun Tzu on the Art of War, p. 58, London, 1910). 2 GILES, Biographical Dictionary, p. 569. 3 T'ang shu, Ch. 155, p. i b. 4 Compiled by Su Kien in the early part of the eighth century (BRETSCHNEIDER, Botanicon Sinicum, pt. I, p. 143, No. 76). 5 COUVREUR (p. 473 c) explains this word as mailles d'une cuirasse. 5 Ordinarily "a quiver," but originally a case to place any arms in; hence COU- VREUR (p. 304 a) enveloppe de cuirasse, de bouclier, de lance (see p. 176). In the above case, the costume worn over the armor is thus called, because, like a case, it envelops the armor. 7 See above, p. 196. 8 Wooden armor existed perhaps under the Later Han dynasty, though alluded to only in a metaphorical sense. In the Chapter Wu king chi (Hou Han shu), ice- crusts covering trees (mu ping) are likened to wooden armor (mu kiai) ; and the com- mentary explains kiai as symbolizing military armor (P'ei wdn yunfu, Ch. 69, p. 42) ; thus the existence of wooden armor at that time might be presupposed as being in- strumental in this comparison. "Wooden armor" can be nothing but wooden slat armor, as described by W. HOUGH (Primitive American Armor, /. c. t pp. 632, 636) among the North- American Indians. Another type is presented by the wooden armor of the Thompson Indians described by JAMES TEIT (The Thompson Indians of British Columbia, Jesup North Pacific Expedition, Vol. II, p. 265) as consisting of four boards an inch and a half thick, two for the front and two for the back, which reached from the collar-bone to the hip-bone; these boards were laced together with buckskin, and the whole covered with thick elk-hide; while the same tribe made also THE PROBLEM OF PLATE ARMOR 277 We do not know from the literary records how the armor credited to Ma Sui was constructed in detail ; but it was doubtless the forerunner of the armor-costumes, as we find them duly sanctioned by the emperors of the Sung, Ming, and Manchu dynasties; those, in my opinion, go back to types established in the T'ang period. Ma Sui's invention was a coat of cotton or silk, the exterior or interior of which was covered with rows of small iron or steel plates. Indeed, plate mail is well represented on Chinese clay statuettes of the T'ang period, in accordance with what we find in the art of Turkistan. The nearest approach to Ma Sui's contrivance may be recognized in the clay figure of a soldier (five of these are in our collection) on Plate XXX. These figures coming from graves of Shen-si Province are clad with an ordinary long-sleeved coat; in front and back, over the chest, and along the lower edge, we notice a row of plates emerging. 1 Plates, accordingly, strengthen the front and back of the coat, and are covered with the same material as the latter consists of. The whole affair is tightly held together by two bands adorned with bosses. The two clay figures on Plate XXXI represent two identical speci- mens of the same type of warrior, coming from Shen-si Province. The left hand, which is raised as if brandishing a weapon (spear), is unfor- tunately broken off in both pieces. The expression of lively motion and the quality of modelling are remarkable. In the grim faces slightly bent and turned sideways, the demoniacal power of these armored knights watching over the grave is well represented. The helmet- mask is formed by a bird's head with a strong flavor of the Indian Garuda; a horn or crest in the centre of the head is broken off. The well-developed eyebrows of the bird's faces terminate in spirals arranged on the foreheads; the beak is strongly curved; the interval between the eyes is filled with a pigment of indigo. The helmet covers the back of the head, nape and chin. A shawl is elegantly draped around the shoulders, and tied in a knot over the chest, the two round iron breast- plates being visible beneath it. An animal head is brought out in relief in the middle, apparently a metal clasp holding the two sheets of the armor together. 2 An apron, a sort of undivided braconniere, consisting of three horizontal rows 8 of long, rectangular iron plates is worn over corselets from narrow strips of wood from half an inch to an inch in thickness or of rods, going entirely around the body; the strips of wood were placed vertically, and laced together with bark strings; such vests were generally covered with one or two thicknesses of elk-skin. 1 Compare Plate XVIII. 1 Sheet armor is discussed in Chapter VI. 1 It is interesting to compare it with the clay statuette found by GR^NWEDEL, /. c. t Fig. 460. 278 CHINESE CLAY FIGURES the coat (Plate XXXI, Fig. i); the plates are distinctly represented by parallel rows of lines executed in black ink and continued on the back (Fig. 42) ; the lines are somewhat rounded at the top, and leave no doubt of the real shape of these armor-plates. In Fig. 2 of the same Plate these lines are omitted, or may have been worn out. FIG. 42. Back of Clay Statuette represented on Plate xxxi, Fig. 1. As those two statuettes represent the typical armed warriors of Shen-si Province, so the pair on Plate XXXII illustrates the character- istic types current in Ho-nan, and is for this reason inserted here, though not vested with plate armor. Of powerful martial appearance, "armed at point exactly, cap-a-pie," these heroes valiantly lean on the hilts of their straight swords resting between their feet, not dissimilar to a THE PROBLEM OF PLATE ARMOR 279 mediaeval Roland. They are protected by iron sheet armor, 1 over which a jerkin is thrown, two circular spaces being cut out on the thorax, and exposing the iron plastrons or breastplates. The helmet envelops the occiput, nape, and cheeks, and is held by a broad leather mentonnire. The baggy trousers are fastened with garters over the upper parts of the thighs. Many of these figurines, as indicated by the remains of pigments, must originally have been well painted, the pig- ments being spread on a background formed by a thick coating of white pipe-clay. 2 In the two figures in question, judging from the traces of pigments, the helmet was colored a crimson-red, the face pink, the eye- balls black, and likewise the big mustache with turned-up tips; the breastplates were vermilion, and the garment surrounding them light green. The sleeves on the upper arms are still decorated with parallel black stripes; those on the lower arms are painted a crimson color, the hands pink. Geometric ornaments that are but partially preserved were painted in red on the portion of the coat beneath the girdle. Plate armor is met also on contemporaneous Chinese sculpture in stone. There is in the Museum's collection a marble slab dug up in the environment of the city of Hien-yang, Shen-si Province (Plate XXXIII). It represents a mock-gate which denoted the en- trance to a tomb. The two door-leaves countersunk in the slab are divided by a faint line in the centre, and kept closed by means of a bolt carved in relief. On each leaf is delicately traced the figure of a guardian completely armored with plate mail, and holding a sword. On the lintel two phenixes surrounded by rich foliage are chiselled out in flat relief. Plate armor was officially adopted by the Sung dynasty. In 1134, the Imperial Armory had four model pieces constructed, which were founded on the principle of the plate. The first of these, an armor suit, consisted of 1825 plates (styled ye, "leaves," written without the classi- fier 'metal') polished and burnished on both sides; the e"pauli6res (pauldrons) were protected on the inner side by 504 plates; each of these plates weighed one fifth of an ounce plus six fen. The second, also a coat, was formed of 332 plates, each plate of the weight of two-fifths of an ounce plus seven Jen. The third piece, a lower garment, was com- posed of 679 plates of the shape of a tail-feather of a hawk, each plate weighing two-fifths of an ounce plus five fen. The fourth piece was a helmet consisting of 310 plates, each weighing one-fifth of an ounce plus five fen; the total weight of the helmet, inclusive of its appurte- 1 See Chapter VI. 2 The same process is applied to T'ang pottery vessels, as will be seen in Part II. 280 CHINESE CLAY FIGURES nances sheltering the nape and the forehead, amounted to one catty and one ounce. The leather straps wound around the head weighed five catties, twelve ounces and a trifle more than a half. Each suit had a weight of forty-nine catties and twelve ounces. The weight of an armor naturally depends upon the weight of the individual wearer; in the army, however, concern about the individual would not be feasible, and would incur heavy expense as well as waste of material. It was therefore thought advisable to reach a compromise, and to standardize the weight of the armor at from forty-five to fifty catties, with the strict under- standing that in no case should it exceed fifty catties. 1 In regard to the Mongols, we mentioned the employment of hide and hide scale armor in their armies (pp. 190, 197). There are also accounts to the effect that plate mail was known to them. In the earliest Euro- pean document regarding the Mongols, written by MATTHEW PARIS under date of 1240, giving the first description of this new people, they are described as "men dressed in ox-hides, armed with plates of iron, . . . their backs unprotected, their breasts covered with armor;" their backs remained unprotected so that they could not flee. 2 WILLIAM OF RUBRUCK, travelling from 1253 to 1255, makes us acquainted with sundry types of armor in use among the Mongols, the haubergeon (chain mail), scale hide armor, and iron plate armor, the iron plates being introduced from Persia. 3 But the Franciscan Friar John of Pian de Carpine (or Latinized, Piano Carpini), who travelled to the Court of Kuyuk Khan (1245-47) as ambassador of Pope Innocent IV, is that mediaeval writer who has left to us the clearest and most complete description of Mongol plate armor. At the same time he is the first European author to give any description of Eastern plate armor at all. In his "Libellus historicus" (Cap. XVII) 4 he describes the defensive armor of the Mongols, and states that the upper part of their helmet is of iron or steel, while the portion guarding the neck and throat is of leather. Whereas the majority wear leather armor, some have their harness completely wrought from iron, which is made in the following manner. They beat out in large numbers thin iron laminae a finger broad and a full hand long. In each they bore eight small apertures, through which they pull three straight leather thongs. Thereupon they arrange these laminae or plates one above another, as 1 See Sung ski, Ch. 197, p. 6. 1 W. W. ROCKHILL, The Journey of William of Rubruck, p. xv (London, 1900). 1 Ibid., p. 261. He mentions also iron caps from Persia. 4 In the new edition of G. PULL, pp. 86-88 (Studi italiani difilologia indo-iranica, Vol. IX, Firenze, 1913). C. R. BEAZLEY, The Texts and Versions of John de Piano Carpini, pp. 89, 124 (London, 1903, Hakluyt Society). THE PROBLEM OF PLATE ARMOR 281 it were, ascending by degrees, and tie the plates to the thongs mentioned by means of other small and tender thongs drawn through the apertures. And in the upper part they fasten a single, small thong, doubled on each side, and sew it on to another, that the plates may be well and tightly connected. Thus a uniform protection is effected by these plates, and such-like armor is made for their horses as well as for their men. It is so highly polished that a man may mirror his face in it. In regard to shields, Carpini observes that they have them made of wickerware or small rods (de mminibus vel de virgulis factum), but that they carry them only in camp and when on guard over the emperor and the princes, and then only at night. The armament of the Mongols was not uniform; and this complex and expensive structure of plate armor was probably within the reach of but few. Their ordinary armor was a cuirass of boiled-leather scales. According to Carpini, the leather was that obtained from an ox or some other animal; and the scales were a hand broad. 1 Three or four of these were held together by means of pitch, and connected with one another by means of cords. In double or triple rows they were laid around the trunk. The complete set of armature consisted of four parts, the front piece, reaching from the neck down to the lower part of the thighs, and well adapted to the form of the body; the back protector, and an apron encompassing the back and abdomen; and the brassards and cuishes. The back of the upper arm was guarded by two iron plates hinged together. The plate idea has remained the basic principle of the officially recognized body armor down to the end of the eighteenth century. The changes were those of style and ornamentation only, while no funda- mental innovations were added in the Ming and Manchu periods. The Statutes of the Ming Dynasty (Ta Ming hui lien) contain the following regulations relative to plate armor: "In 1374 it was ordered that in- stead of the threads, by means of which the armor-plates were held together, leather thongs should be used. In 1376 the General Staff was ordered to make war-suits of cotton (mien htia chan i), and to apply to them four colors, red, purple, dark blue, and yellow; for Kiang-si and other places, to make war-coats with different colors on the exterior and interior, and to cause the officers and petty officers to change their uniforms accordingly. In 1383 orders were given for harness, each set to be made as follows: for the colletin (neck-guard) thirty plates, for the body armor two hundred and nine plates, for the plastron (breast- plate) seventeen plates, for the pauldron (e"pauliere) 2 twenty plates. 's complete text is followed here; this portion is lacking in the former editions of Carpini. 1 In Chinese, "arm-pit plates" (chi wo ye). 282 CHINESE CLAY FIGURES FIG. 43. Illustrations showing the Conventional Chinese Style of Drawing Plate Mail (from T'u shutsich'tng). THE PROBLEM OF PLATE ARMOR 283 FIG. 44. Illustrations showing the Conventional Chinese Style of Drawing Plate Mail (from T"u shutsi ck'txg). 284 CHINESE CLAY FIGURES All these pieces are soaked with lime, and united by means of soft, tanned leather thongs passing through the perforations of the plates. Along the maritime coast of Che"-kiang and in Kuang-tung, the guards stationed there have to utilize black-lacquered iron plates perforated and connected by cotton strings; for the rest, however, their armor is made in the style of the 'brilliant armor' (ming kid). 1 " In 1435 (tenth year of the period Suan-te 1 ) the ordinance was issued that each coat had to be fixed at a length of four feet and six inches, with a supply of two catties of cotton and velvet; for the making of the trousers, half a catty of cotton and velvet should be used; the wadded boots should be from nine inches and a half up to one foot, or one foot and two inches long. Now, the regulation was provided to make wide coats and trousers, and to employ for these fine, closely woven, broad, and white cotton stuff dyed blue, red, or green; the sleeves should be wide and long; and the materials employed, like cotton and velvet, should be of solid quality. The wadded boots should be fine, thick, and strong. In the finished garment a written entry was to be made by the government officers who inspect the troops and examine their equip- ments; they shall enter the family name and surname of the tailor, the cost-price, the measurements in feet and inches, the weight, the number of strips of cloth used in the skirts, with seal attached. At fixed terms, every year before the seventh month, the uniforms were to be furnished. "In the year 1496 (ninth year of the period Hung-chi under the Emperor Hiao-tsung) it was ordered that for the covers of the armor 2 thick and dark blue and white cotton stuff should be employed, that for the ' armor with nails ' (ting kid) small studs with lacquered heads should be used. It was further settled that, for each set of a blue cotton stuff iron armor, iron to the quantity of forty catties and eight ounces should be required, and that each set of the finished armor should weigh twenty-four to twenty-five catties. In 1 503 order was given that the guards stationed in southern China should exchange their iron armor for that made of water-buffalo skin sewed together by means of cotton ropes." Figs. 43 and 44 are here inserted to illustrate the conventional Chinese style of representing plate mail. 3 The Manchu dynasty adopted the military institutions of the Ming in their entire range, and in particular the defensive armor, without making any new additions in the line. Plate XXXIV illustrates a 1 A technical term frequently employed in the Annals; it presumably refers to highly varnished and polished plates of iron or steel. * In Chinese, "the face of the armor" (kia mien). ' Compare note 4 on p. 243. THE PROBLEM or PLATE ARMOR 285 horseman's suit of armor, as it was in vogue during the K'ien-lung period (1736-1795). It is complete with leggings and helmet. The lower garment is covered by four parallel rows of very thin, light and elastic steel laminae of rectangular shape, 9 cm long and i cm wide, rounded at the upper end, perforated at both apex and base, and sewed on to a foundation of cloth, the lower ends being hidden in a fold, where they are riveted by means of studs with broad, gold-plated heads. They are not mutually joined, but one overlaps another to a slight degree. In the upper garment the steel plates are invisible, being inserted as an I interlining (between the lining and the silk on the exterior) , and fastened by means of rivets, so that their gilt heads appearing on the surface indicate the hiding-places of the plates. 1 Dragons, all together six, rising from the sea and standing erect, are embroidered with gold threads on the front and back of the coat, on the two separate shoulder-pieces, and on the two side-pieces underneath the arms. The casque, composed of two steel sheets and surmounted by a black velvet plume, has chased dragons in front, and is provided with silk protectors enveloping occiput, neck, ears, and chin. The uniform of an artillery-man (Plate XXXV) consists of a coat, lower garment, and pair of leggings of wadded black satin lined with light-blue silk, and studded with gold-plated, riveted bosses. These bosses, of a merely decorative character, are the survivals of the iron or steel plates which, as in the preceding harness, are wrapped up in the interior of the garment or are fastened to the lining. The plates are retained in this specimen only for the protection of the shoulders, but have a decorative rather than a positive value. They are arranged in rows of three, two rows being in front and two at the back on each 1 It is singular that the students of plate armor have never turned their atten- tion to China, although it was very clearly described as early as by GERBILLON (in Du HALDE, Description of the Empire of China, Vol. II, p. 340, London, 1741): "All the soldiers who were in the camp, headed by their officers, repaired to the place appointed, armed with their casques and cuirasses. The Emperor put on like- wise his cuirass and helmet, being accompanied with his eldest and third sons; but this latter was not armed, being too young to bear the weight of a Tartarian cuirass. This cuirass consists of two pieces; one is a sort of under petticoat which is girt about the body, and reaches below the knee when they are standing, but covers all their limbs when they are on horseback: the other piece is like the coats of armor of the ancients, but the sleeves are longer, reaching to the wrist. The outside of both these pieces is of satin, for the most part purple, embroidered with gold, silver, and silk of various colors. Next to this satin, lined with some pieces of taffety, are ham- mered plates of iron or steel, finely burnished, which are placed like scales on the body of a fish, whence they probably took the notion. Each plate, which is about an inch and half long, and a little more than an inch in breadth, is fastened to the satin by two small nails, the heads, being round and well polished, appearing without. Some few put another piece of taffety within-side, which covers the iron plates. These cuirasses have this conveniency that they do not deprive the body of the liberty of turning and moving easily; but then they are exceeding heavy." 286 CHINESE CLAY FIGURES shoulder, and connected by a broad, quadrangular plate resting on the shoulder. Each lamina is of steel and gold-plated, and chased with a four-clawed dragon soaring in clouds. From the lower ends of the plate rows project two gold-plated arms, likewise chased with figures of dragons and encircling a round metal plaque (of white copper or tootnague, with brass rim) . A plaque of the same material and size is fastened to the back. Such circular plaques are known as hu sin king (No. 2170), literally "mirror guarding the heart;" that is, a protective amulet. The helmet is identical with the one previously mentioned, and heavily lined with quilted material. The archer's suit of armor (Plates XXXVI, XXXVII) is made of black silk, the interior being covered with broad steel plates, each secured by means of two rivets only, so that the plates are loose and movable. Their disposition on the shoulders is at variance with that in the preceding specimen. There is but one row of three brass plates in front of each shoulder, extending in length as much as the two rows in the previous armor. There are three narrow plates arranged side by side on the surface of the shoulders , and three on the back much shorter than those in front. The three rows covering either shoulder are inter- laced and riveted together. Each of these shoulder-plates is decorated with two rampant dragons playing with a flamed ball. The coat is embroidered with six dragons all together. In i go i I saw a very interesting and ancient suit of plate mail in the Mahakala Temple, which is situated within the walls of the Imperial City of Peking. The suit is of yellow silk, to which iron plates are attached both outside and inside, those on the exterior being very nar- row slips, those on the interior being four times broader and occupying the interval left by the outside plates ; so that by this alternating process a complete plating is insured. On Plates XXXVIII XL is represented what may be styled a parade or ceremonial armor. It is the uniform belonging to a guard-officer of the first rank, detailed on duty in the Imperial Palace. 1 These military officers were divided into seven ranks, each distinguished by a special coat and helmet, and an equipment with appropriate insignia. Their outfits are minutely described in the State Handbook of the Manchu Dynasty. The cut, the style, and the main characteristics of body armor are well preserved in this costume, which is magnificently em- broidered with heavy gold thread, and studded with gilt bosses. Drag- ons', tigers', and lions' heads are the prevailing motives of ornamenta- tion. The disposition of the shoulder-plates is identical with that 1 This is ascertained from the descriptions and illustrations of the official costumes given in Huang ch'ao li k'i t'u shi and Ta Ts'ing hui tien t'u. THE PROBLEM OF PLATE ARMOR 287 in the suit of the artillery-man, except that the dragons are here em- bossed, and the clouds are treated in open-work, all metal pieces being heavily gilded. Five similar plates are suspended from the ends of the shoulder-pieces. The steel helmet (Plate XXXIX) is a gaudy and elaborate affair of admirable workmanship. It is surmounted by a high crest terminat- ing in a pair of eagle-feathers painted with dragons in gold, and is adorned with twelve black sable-tails, 1 seven of which are preserved. Dragons are lavished on it, being chased in the plated brass mountings, or cut out of the same material in full figure, or represented in inlaid feather- work. 2 The bow-case and quiver figured on Plate XL belong to the accoutre- ments of the same official. They are of leather, dressed with red velvet; the upper corners and lower portion of the bow-case are finished with black leather. The metal fittings, of gilt bronze, fastened to the centre and corners of both objects, are of very elegant forms and delicate workmanship. The quiver, in addition to these ornaments, is decorated with three symbols meaning "longevity" (shew). The arrows are stuck into the folds in the interior formed by layers of brown felt. Reference has been made above (p. 272) to the early mining of iron in Korea, and the barter carried on in this metal from there to the neigh- boring tribes. Metal armor (k'ai kid) seems to have prevailed in the kingdom of Kokurye (Kao-kii-li) at an early date. 3 The Annals of the Sui Dynasty 4 state in regard to the kingdom of Sinra in Korea that its defensive and offensive armor is identical with that of China, which would mean that Sinra had derived its armor from China. The Books of the T'ang Dynasty mention a kind of armor, seemingly peculiar to the state of Pek-tsi in Korea, under the name "armor of bright lustre" (kuang ming k'ai), which must have been iron armor. Such a suit was presented in 622 to the Emperor of China, and in 637 iron armor (fie kid), together with carved axes, was sent as tribute to the Emperor T'ai-tsung. 5 Metal armor is alluded to likewise in the Annals of Korea. 6 When the Japanese plundered the royal palace of Kokurye, in 562, 1 This is the required number according to the official statement. 1 From the blue plumes of the kingfisher, Halcyon smyrnensis (in Chinese, fei-ts'ui). 1 Liang shu, Ch. 54, p. 9 b; Nan shi, Ch. 79, p. I b. 4 Sui shu, Ch. 81, p. 4 (also Pei shi, Ch. 94, p. 7). * T'ang shu, Ch. 220, pp. 4, 7. See, for instance, Ta tung ki nien (published at Shanghai, 1903), Ch. I, p. 69 b. The Koreans possess a considerable literature on military art (M. COURANT, Biblio- graphic cor^enne, Vol. Ill, pp. 63-89). 288 CHINESE CLAY FIGURES they obtained among other treasures two suits of armor. l We have no exact information as to what these ancient suits of armor were like, and can base our conclusions only on such specimens as we find in the country at present. Among these are some of considerable age; that is, ranging within the time of the last two centuries or so. We have two main types of harness from Korea, padded armor 2 and plate mail. A very interesting specimen of the latter type is in the Museum collection (Plates XLI, XLII). It is a rough-looking coat of strong twill, lined with blue cotton, and covered with hemp cloth of loose texture imprinted all over with charms by means of wooden blocks. 3 The designs are effaced to such a degree that the details can no longer be recognized: birds' heads, floral designs, trees, arabesques, are con- spicuous; Sanskrit letters, which occur in other specimens, are absent. The buttons in front are of bone; the sides are open, and provided with rows of buttons. Both front and back are strengthened by seven parallel rows of rectangular steel plates (averaging 10.2 X 7.5 cm), very flexible, each coated on both faces with a black varnish. The plates are not mutually connected, but merely imbricated, a feature not yet ob- served in Chinese plate mail. Each plate is clinched to the cloth foundation by means of two rivets with flat heads. They are driven through, and appear on the exterior as big iron nail-heads. A number of plates have additional perforations that are not utilized, but which show that the plates could have been tightly sewed on to the back- ground had not the wearer of this armor preferred to have them loose and movable. The shoulders are covered on the interior by two rows of 1 ASTON, Nihongi, Vol. II, p. 86. 2 A Korean armor consisting of many thicknesses of coarse cotton cloth is figured by W. HOUGH (The Corean Collections in the U. S. National Museum, Report U. S. Nat. Mus., 1891, Plate XXVIII, and Primitive American Armor, /. c., p. 645) ; the sur- face of portions of the coat is printed with prayer formulas (dhdrant) in Sanskrit, and such are inscribed also on the helmet. This practice seems to be derived from China: the helmets used by the imperial house during the Manchu dynasty were chased with Sanskrit characters (see Huang ch'ao li k'i t'u shi, Ch. 13, or Ta Ts'ing hui tien t'u, Ch. 61). A modern Korean helmet is illustrated by E. ZIMMERMANN (Koreanische Kunst, Hamburg, no date, Plate VI). It is a leather helmet of conical shape, surmounted by a bunch of horse-hair and a metal ball in open-work, and adorned with dragons and a hydra about to attack, wrought in gilt metal; fur-lined ear- warmers covered with metal studs are attached to it, the whole style being that of the Manchu dynasty. The costume on Plate VII, explained as the official robe of a minister, is in fact a pseudo-armor, as shown by the rows of metal bosses and the two applique' dragons playing with balls; it is similar to the one on our Plate XLIII. Generals' and soldiers helmets are figured and briefly described by F. H. JENINGS (Korean Headdresses in the National Museum, Smithsonian Miscellaneous Collec- tions, Vol. 45, 1904, pp. 161-163). Good specimens of these are also in the Field Museum. 1 Much in the style of Tibetan cloth prints which are attached to flag-poles set up on the roofs of houses in order to bring luck to the inmates. THE PROBLEM OF PLATE ARMOR 289 plates, and are reinforced on the outside by iron bars, one for either shoulder, each bar consisting of two parts hinged together, so that easy motion is secured. The casque (Plate XLII) consists of two sheets of iron riveted together, with a projecting visor and frontal covering the forehead. The couvre-nuque and the ear-protectors attached to the casque are made from the same hemp cloth as the harness; they are likewise printed with designs, and stuffed with iron plates, which are kept in place by means of the clinches appearing on the surface. The top of the helmet is surmounted by an iron trident and a tuft of red- dyed horse-hair. There is no doubt that this Korean armor represents a very primitive type of plate mail, and conveys to us an excellent idea of what the ancient Chinese plate mail may have been like. 1 On Plate XLIII is shown the Korean court costume of a high official, which is a pseudo-armor in imitation of Chinese style. The cloak-like robe consists of red cloth trimmed with otter-fur, and lined with light- blue Chinese silk. It is strewn with regular rows of brass bosses rep- resenting purely decorative survivals or reminiscences of plate armor. Three globular buttons close the garment in front; the two lower ones are hidden under a broad sash of figured blue silk. Around the neck are laid twelve maple-leaves cut out of brass and riveted to the cloth (in the illustration hidden by the ear-protectors of the helmet). The epaulets are adorned with full figures of gilt, embossed dragons hunting for the flamed jewel; they are worked in sections, which are cleverly connected by hinges, so that the shoulders are not handicapped in any motion. The helmet is an elaborate affair, composed of strong, com- pressed, glazed leather, lined with soft leather. The surface is divided by means of four metal bars into four compartments, two of which are each adorned with a dragon, the two others each with a phenix on the wing, all of gilt bronze. On the sides, silver phenixes filled with dark- blue enamel 2 are added. The most interesting point concerning our subject is the fact that the ear-muffs and nape-guard, likewise of red cloth trimmed with otter-fur, have thin copper plates concealed between the outside material and the lining. They are kept in place by copper nails with gilt heads. A quilted cap of blue silk is worn next to the skull, 1 W. E. GRIFFIS (Corea, the Hermit Kingdom, p. 101) figures what he calls "a Korean knight of the sixteenth century." I have no judgment on the authenticity and alleged dating of this illustration, but in itself it is interesting in that the laminae forming the plastron and reinforcing the sleeves and brassards are arranged in hori- zontal (not, as usual, vertical) position. "Many of their suits of armor," GRIFFIS says, "were handsomely inlaid, made of iron and leather, but less flexible and more vulnerable than those of the Japanese, which were of interlaced silk and steel on a background of tough buckskin, with sleeves of chain mail. The foot-soldiers on either side were incased in a combination of iron chain and plate armor." 1 A process still extensively applied in China to silver jewelry. 290 CHINESE CLAY FIGURES under this helmet. Below, on the same Plate, is illustrated the black- varnished wooden chest in which the suit is stored, with a special conical compartment for the casque. This arrangement is also in imitation of a Chinese practice. Japanese plate armor has so frequently been described l that it is not necessary to dwell on this subject. What is important for the purpose of our investigation is the fact that it does not arise in Japan earlier than the first part of the ninth century ; 2 that is, in the T'ang period, when it was perfectly known in China. It is therefore certain that the idea has penetrated into Japan from China and Korea, whatever subsequent developments, changes, and improve- ments plate mail may have undergone in Japan. Armor composed of horizontal rows of small iron plates, presumably of Chinese origin, seems to occur occasionally in Tibet. A specimen recently presented by the Dalai Lama to the King of England is now preserved in the British Museum. 3 Looking backward at the remarkably wide distribution of plate armor, we cannot fail to recognize in this fact a certain degree of histori- cal coherence. This coherence, without any doubt, exists in the T'ang period between Turkistan and China on the one hand, and between China, Korea, and Japan on the other hand. But the T'ang epoch de- notes only the culminating point in this development, that period in which we observe plate mail wrought to its greatest perfection. Metal plate mail is a complex affair of difficult and refined technique, a down- right product of higher civilization, which is witnessed by the fact that it is conspicuously absent among all primitive cultures of Asia, Africa, and ancient Europe. Certainly it did not come into existence all at once as a finished product of industry. It ran through many experi- mental stages, and took time to develop and to mature. The elegant specimens of the T'ang, granting the muscles free motion and aiming at aesthetic qualities, were preceded by those of coarser and cruder work- manship ; as we see, for instance, in the Korean specimen on Plates XLI and XLI I. There is a great deal of probability in the supposition that such existed, both in China and among the Iranian and Turkish tribes of 1 First by PH. H. v. SIEBOLD, Nippon, Vol. I, p. 333. 1 J. CONDER, The History of Japanese Costume (Transactions Asiatic Society of Japan, Vol. IX, 1881, p. 256). According to this author, the employment of plates and scales of iron in armor was finally established as late as the epoch Tensho (1573- 1592). See chiefly BASHFORD DEAN, Catalogue of the Loan Collection of Japanese Armor. 1 It is figured on Plate III of the Ethnographical Guide published by the British Museum. See also A. STEIN, Ancient Khotan, Vol. I, p. xvi. Armor of small steel plates riveted on red velvet appears also in Europe (see, for instance, BASHFORD DEAN, Catalogue of European Arms, p. 48), but this subject is not within the scope of the present investigation. THE PROBLEM OF PLATE ARMOR 291 Central Asia, ages before the T'ang, presumably as early as the era of the Han (p. 214). Iranians surely were the mediators between the west and the east in this matter, in the same manner as they acted in the transmission of chain mail, caparisons for horses, l and the great principles of cavalry tactics. Up to this point the territory is fairly well reconnoitred. But thus far we are entirely ignorant of when and how plate mail may have arisen in Iran, nor do we positively know whether it existed there at all; if it did, the possible connection with the plate mail of ancient Egypt and Assyria remains a subject for in- vestigation. Altogether the impression remains that plate armor, the last offshoots of which we encounter in the farthest north-east corner of Asia and the farthest north-west of America, took its origin from western Asia. This field is entirely beyond my competency ; and it is the sole object of these notes to point out the existence of the problem, and to leave its final solution to the ambition of others. 1 See Chapter VII. VI. DEFENSIVE ARMOR OF THE T'ANG PERIOD In the preceding notes we had occasion to refer repeatedly to de- fensive armor of the T'ang period (618-906). Mention has been made of the fact that cuirasses of rhinoceros-hide were then still in existence (p. 189), and also that those of buffalo-hide then came into vogue (p. 162). Plate mail reached its climax at that time (p. 277), and chain mail was introduced from Iranian regions (p. 246). The types of armor utilized under the T'ang must have been of a large variety. The Statutes of the T'ang Dynasty, drawn up by the Emperor Yuan-tsung (7 I 3~755)> 1 enumerate thirteen classes of armor manufactured by the Imperial Armory (wu k'-u) : six of these were of iron, and of the types of plate, scale, and chain armor; others were of white stuffs, black silk, hide, and even of wood (p. 276). How the military uniform was then combined with armor has also been set forth (p. 2 7 5) . Besides the means of protection officially recognized in the army, there were other plain and cheap contrivances for the use of the people, such as are still com- mon in the country. Thus we hear in the Annals of the T'ang Dynasty in regard to a certain Ch'eng K'ien that he made defensive armor from layers of felt. 2 The most curious armor of which we read in that period was a kind made from sheets of paper laid in folds, which could not be pierced by the strongest arrows; this invention is credited to Shang Sui-ting. Under the Sung dynasty, paper armor was officially recognized, for we hear that in the year 1040 the troops stationed in Kiang-nan and Huai-nan (in An-hui Province) were ordered to fabricate thirty thousand suits of paper armor, to be distributed among the garrisons of Shen-si Province. The localities mentioned are celebrated for their paper manufacture, and were accordingly obliged to contribute to a demand which could not be filled in Shen-si. The Wu pei chi (Ch. 105, p. 17) of 1621 has preserved for us an illustration of such paper armor (Fig. 45), arranged in triangular scales slightly rounded at the base. These suits were especially favored under the Ming in southern China by the soldiers fighting the Japanese, who then invaded the Chinese coasts. 3 The favorite brand of paper for this purpose in recent times 1 See above, p. 189. * P'ei w$n yun fu, Ch. 40, p. 86. In 1286, according to Yuan shi, the country of Ma-fa sent a tribute of saddles, bridles, and felt armor. 1 The same work illustrates also armor of plaited rattan; but it is not known at what time this type of armor sprang up in China. 292 DEFENSIVE ARMOR or THE T'ANG PERIOD 293 was the famed Korean paper highly prized in China and Japan for its toughness and durability, and forming part of the annual tribute sent from Korea to Peking. In the treaty of 1637, concluded after the Manchu invasion, the figure was stipulated at five thousand rolls of large and small paper. 1 A good deal of Korean paper was utilized by FIG. 45. Paper Armor (from Wu fei chi). the tailors of the Chinese metropolis as lining for the coats of officials and gentlemen. It served also for the covering of window-frames. A sewed wad of from ten to fifteen thicknesses of it made a protective armor for the troops. It is said to have resisted a musket-ball, but not a rifle-bullet. 1 1 W. W. ROCKHILL, China's Intercourse with Korea, p. 25 (London, 1905). A notice on Korean paper is contained in the Wei Ho (Ch. 12, p. i b). 1 W. E. GRIFFIS, Corea, the Hermit Nation, p. 153 (New York, 1904). Paper and cotton armor still exist in southern China. Consul BEDLOE (quoted above, p. 1 80) offers the following remarks on this subject: "Parallel to this alternating of leather and wool in the north was that of paper and cotton cloth in the south of China. It seems ridiculous to call such combinations armor, and yet they make an armor superior in many instances to steel. Thirty thicknesses of alternate calico and paper will resist a pistol bullet or one from a rifle at a distance of a hundred yards. A spearman who thrusts his weapon into a man clad in this kind of garment 294 CHINESE CLAY FIGURES The most interesting source for the study of T'ang defensive armor is naturally offered by the clay figures and figurines; and these reveal to us a new style of armor, that of sheet armor, which is thoroughly characteristic, not of the life, but of the art, of this period. The type of clay image which comes here into question is of the great- est interest, as it originated in the ivaitic worship of India, and be- came widely diffused over Tibet, Turkistan, China, and Japan. We may in general classify the manifold variations of this type among the so-called Dharmapala ("Protectors of Religion"), guardian deities adopted by Buddhism, and more specifically designate it as Yama, the God of Death, who still plays such a prominent r61e in Tibetan Lamaism. J. EDKiNS 1 holds that he may be pointed to as the most remarkable example of the influence of Hindu mythology on the popular mind of China. Among the clay figures of the T'ang period we find two fundamental types of this Hindu god, a zoomorphic and an anthropomorphic form. The zoomorphic form is doubtless the older one, and is closely associated with the Lamaist representation of Yama as Dharmaraja (" King of the Law ") , figured with the head of a bull, and dancing on the back of this animal. 2 Old ZIEGENBALG, who wrote in 1713 at Tran- quebar on the coast of Coromandel, gives the following description of his image as found in southern India: "Yama is represented as being quite black, with a horrible face, and a crown on his head, and al- together surrounded by fire. In his mouth he has a lion's teeth, and in his four hands he holds respectively a club, ropes, a trident, and a wine-jug, from which he gives wine to the dying to mitigate the bitter- ness of death. On the whole he is adorned like the king, and rides on a black buffalo. The poets have written many stories about him, which these heathens receive with undoubted credence." 3 On Plate XLIV we see him modelled in clay, with most powerful can neither wound his enemy nor extract his weapon, and if his enemy is an archer or is armed with a long sword or javelin, he is likely to lose his life for his mischance. The suit of a famous Yun-nan bandit consisted of sixty thicknesses of cotton cloth and paper, and made him practically invulnerable. These suits are comparatively light, are very durable, and of course, extremely cheap." Heavy quilted cotton armors are still occasionally worn by Chinese in this country under their garments, when the members of secret societies are on the war-path. The writer was once shown a wonderful specimen in the Police Department of New York, which weighed so heavily upon the unfortunate Chinaman that he was unable to run, and was easily captured after a shooting-affair. 1 Chinese Buddhism, p. 219 (London, 1893). 2 PANDER and GR^NWEDEL, Pantheon des Tschangtscha Hutuktu, p. 62; GRttN- WEDEL, Mythologie des Buddhismus, pp. 62, 168, 174. 1 B. ZIEGENBALG, Genealogy of the South-Indian Gods (translated into English by G. J. METZGER), p. 192 (Madras, 1869). DEFENSIVE ARMOR OF THE T'ANG PERIOD 295 expression and lively motion, standing on the body of a sow. 1 The animal is represented in the agony of death, with wide-open muzzle and with its facial muscles distorted, stretching forth its four feet. The terrific god has the head of a bull, exactly as in the corresponding Tibetan images, with two curved horns, bushy eyebrows, and protruding eyeballs painted black; his mouth is wide agape, and shows the esopha- gus. Palate and face are coated with a red pigment. Hands and feet are provided with sharp eagle-claws. The head is surrounded by flames. 2 A projecting crest is attached to the spine, and there is a tail at the end of it. Another representation (Plate XLV), likewise with horned bull- head, shows him in the same posture, standing over the back of a re- clining bull, a snake winding around his left arm. In another clay figure (Plate XLVI) he is clad with a leopard-skin, and standing in the same attitude as the two preceding ones, but without a bull; the bearded face, though of human traits, bears a grim, demoniacal expression, and is painted red, beard and mustache being in black outlines. The erect ears are animal-like, as are the hands and feet; the head is sur- mounted by a long, slightly twisted horn, somewhat similar to that on the clay figures of sphinxes. Between the animal and the human types, there is an intermediary form with some features borrowed from both. In Fig. i of Plate XLVII, his head is still modelled in the style of the bull-faced Yama, with horns and flames, but he is equipped with an armor in the same manner as the human forms; and the plume surmounting his head-dress is identical with the one in the figures of knights (Fig. 2 of the same Plate). The statuette on Plate XLVIII, belonging to the same intermediary type, displays all these features brought out still more clearly, the two- horned bull-like head with a certain assimilation to human traits, the high plume and pommels of the elaborate head-dress, animal-heads protruding from the sleeves, breastplates, an apron, and a skirt con- sisting of two flaps; thus he is standing over the figure of a demon.* A demon of exactly the same type is modelled in the glazed statuette on Plate XLIX. The god, however, is here represented as a purely human form, a knight clothed with heavy armor, pressing his right hand on his hip, and raising his left. The figure, except the head, is coated with 1 Why in this particular case a sow, and not as usual a cow, is represented, I do not know. The interpretation itself is indubitable, the animal being modelled in a most naturalistic style and thoroughly characterized by the anatomy of the head and the crest on the skull and spine. 1 The tips of two of them are broken off. * Compare in Indian art Kubera standing on a Yaksha (GR&NWEDEL, Buddhist Art in India, p. 40; and Mythologie des Buddhismus, p. 15). 296 CHINESE CLAY FIGURES soft lead glazes in four colors, green, blue, brown, and yellowish white; the demon is glazed yellowish white with brown hair. The plastron of the knight's armor is blue, the circular portions are white, the knobs in the centre are blue. 1 Besides the god in the garb of a knight trampling down a demon, we meet again a similar type of knight standing on the back of a reclin- ing bull (Plate L). 2 The positions of feet and hands are quite stereo- typed. The right foot is set on the head of the bull, the left on its croup; the left arm is akimbo, and the right hand is raised as if throw- ing a weapon (Plates LI and LIII, Fig. i). Or, the left foot rests on the bull's head, the right on its croup, while the left arm is akimbo, and the right hand raised for attack (Plates LIII, Fig. 2, and LIV). It will be noticed how the conventionalization of this type gradually advances. Somewhat more artistic features adhere to the statuette on Plate LII, which, with the exception of the head, is glazed in three colors, green, brown, and yellowish white; the bull is lost, and may be supplemented from the preceding figure in Plate LI. 3 The bull, as previously pointed out, alternates with the demon (Plate LIII, Fig. 2). In Plate LIV, Fig. 2, a human body is plainly fashioned; so that in this case we have the same motive as in the Lamaist images, in which a human corpse serves as basis for certain Tantrik deities. The flat miniature figure on Plate LV is very curious, in that it is cast from lead; it shows Yama in the same pose as the preceding ones, and standing on a bull. Finally we see the ultimate stage of develop- 1 The method of glazing in the T'ang figures is very interesting: the idea under- lying the application of glazes, if more than one glaze is enlisted, seems to centre upon the tendency of reproducing the colors of costume or armor. In the majority of cases, probably in all human figures, it is only the costume which receives the col- ored glaze, while head and hands remain uncoated. In the figurines of women it is sometimes merely the central portion which is glazed, the dresses usually being of green and brownish-yellow tinges, while the remaining portion is covered with a white plaster. In the case of monochromes, the glazing as a rule extends to the whole figure. 1 A curious analogy to this type is offered in European mediaeval art by the brasses of English lords in full armor standing on the back of a lion or another animal, and by the monument of Count Otto IV of Henneberg, and other German statues (for illustrations see, for example, BASHFORD DEAN, Catalogue of European Arms and Armor, Figs. 17-22; or Encyclopaedia Britannica, Vol. I, p. 587). 1 A type similar to this one is figured on Plate XIV of the Catalogue of Early Chinese Pottery, published by the Burlington Fine Arts Club (London, 1911), except that in this figure both feet are straight on the same plane. The modelling of the head, the position of the left arm, the armor, and the style and colors of the glazing, are identical in both figures. The pose of the right arm, however, must have been different in our figure, in accordance with the drawn-up right foot; it doubtless has to be supplemented correspondingly with the left arm in the figure on Plate XLIX ; that is, the arm was raised, and the hand either formed into a clenched fist, or the palm stretched outward. Also in the specimen referred to, which is in the possession of Mr. G. Eumorfopoulos of London, the face and hands are unglazed, while the re- mainder is glazed in cream, orange-yellow, and green colors. DEFENSIVE ARMOR OF THE T'ANG PERIOD 297 ment in clay figures without the mythological attributes of the bull or demon; these are purely armored knights or guardians. But the derivation of this type is unmistakable. The demoniacal expression in the face of the tall figure (Plate LVI) the mouth is agape, as if he were represented shouting reveals his affinity with the group of the God of Death. His style of hair-dressing is the same as that in the figure on Plate L, and he is armored in the same manner as the preceding images. Such a demon-like creature is disclosed also by the warrior on Plate LVI I, with very elaborate body armor consisting of large plas- tron and dossiere of metal, connected by leather straps running over the shoulders. It is plainly visible how the two breastplates join together in the middle. He wears a high collar and turned-up sleeves, animal- heads being brought out on the upper arms; the waist is narrow and tied by leather straps, and an apron of plate mail is hidden under the garment. Finally we come to clay figures which are plainly knights or guardians armored ca.p-d.-pie, without any mythological reminiscence (Plates LVIII-LX). In Japan, types exist which are related to the Chinese clay figures already described. These are of highly artistic qualities, and show us that in the T'ang period a Buddhist school must have flourished, the tradition of which embraced the whole of eastern Asia. Two examples are here selected. The one is a clay figure, originally colored, in the Todai temple in Nara, founded in the middle of the eighth century (Fig. 46). 1 This remarkable statue is justly dated by the Japanese in the eighth century (T'ang period). Head-dress and armor, as well as pose of hands and feet, closely agree with those of the Chinese types; here we observe that the raised hand was indeed grasping a weapon. The Japanese name this figure Dhritarashtra, one of the four Maharaja or Lokapala of Hindu mythology guarding the world-mountain Sumeru. Another very similar statue (Fig. 47) , 2 likewise and justly attributed to the eighth century, is named VirQpaksha, the third of the four guardians of the world. Both are posed on the bodies of demons. 8 The four Lokapala are conceived as kings and heroes, and hence represented as 1 The sketch is reproduced from the Kokka, No. 170, 1904. 1 From the Kokka, No. 42. In the same manner Vajrapapi is represented (Kokka, No. 28, Plate V). 1 The Japanese identifications are doubtless based on correct traditions, but I am not inclined to transfer these interpretations to the Chinese figures standing on demons as those mentioned before. We noticed that in some of these the bull-face of Yama is still preserved, and that consequently this figure is Yama: hence we may infer that also the anthropomorphic figures standing on demons are derived from the same type. Compare also the four wood-carved Lokapala posed on crouching demons in Kokka, No. 165, 1904. 298 CHINESE CLAY FIGURES FIG. 46. Japanese Colored Clay Statue of Dhritarashtra, Eighth Century (after Kokka). DEFENSIVE ARMOR OF THE TANG PERIOD 299 FIG. 47. Japanese Statue of Virupaksha, Eighth Century (after Kokka). 300 CHINESE CLAY FIGURES armored; at the same time they are regarded as "protectors of religion" (Sanskrit dharmapdla), and for this reason are shown in so-called terrific forms. 1 Analogous types of Lokapala are met in the contemporaneous stone sculpture of China, for instance, in the caves of Lung-me'n. 2 A marble relief (Plate LXI) in the Museum collection shows an armored Virupak- sha leaning on a two-edged sword, and holding a miniature Stupa (tope) in his left hand. 3 The armor is very clearly represented: the breast- plates tightly envelop the thorax, and are held in place by means of broad leather suspenders running over the shoulders and connecting with the dossire. The metal buckles fixed to the edge of the plastron are plainly visible, and tongues are passed through perforations of the straps. The ends of these straps reach the centre of either breastplate, and are strengthened at this spot by an additional piece of leather. The belt is a broad leather band starting in a rosette from the sternum, the end being turned upward from beneath the girdle. It is of especial interest that similar clay figures representing Loka- pala (the term is perhaps too narrow, and should rather be Dharmapala) have been discovered in Turkistan. 4 These are likewise enveloped by suits of armor much resembling those of the Chinese and Japanese clay statuettes. It is therefore obvious that in this case the question is not of any national type of armor which the Chinese applied to the clay figures, but that this armor was already peculiar to the latter when they were received in the channel of Buddhist art and reproduced by the potters of China. The art displayed in the caves of Tun-huang on the boundary of Turkistan and China may be made directly responsible for the transmission of this particular type from Turkistan to China; for there we find a statue of a Dharmapala standing on a demon, and with exactly the same characteristics as our Chinese clay figures. 6 Was this armor ever a living reality in China, or did it merely remain an artistic motive? It is not very likely that it ever became of any practical use among the Chinese. It is not described in the official records of the T'ang dynasty; at least, in the records at our disposal no armor is 1 Styled in Sanskrit krodha, in opposition to fdnta, the mild forms. A mild form of Yama seated on the back of a bull was painted by the Buddhist monk Eri, who died in 935 (reproduction in Kokka, No. 133, 1902). 1 CHAVANNES, Mission, No. 353. Besides the hero and warrior type of Lokapala, we have in the same period a nude type clad only with an apron, and with fine modelling of strong, well-developed muscles (ibid., Nos. 358, 359). An excellent marble of the latter type is in the collection of Field Museum. * Styled in Chinese "King of Heaven lifting a StQpa" (To t'a rien wang). * A. GR^NWEDEL, Altbuddhistische Kultstatten in Chinesisch-Turkistan, p. 205. 5 A. MAYBON, L'art bouddhique du Turkestan oriental, p. 55 (L'art dtcorattf, 1910). DEFENSIVE ARMOR OF THE T'ANG PERIOD 301 described that could freely be recognized in it. Sheet armor, indeed, was never peculiar to China, but is plainly of western origin. Above all, this type of armor, even if it should have sparsely existed here and there in China during the T'ang, has left no trace in any later period; it does not survive in any harness of the Ming and Manchu periods; and this is a signal fact, as otherwise the T'ang tradition in regard to armor was still alive in that recent age. 1 Buddhism, however, may have influenced Chinese armature to a certain degree. A peculiar kind of armor styled " lion-armor " (ni k'ai) is attributed to the T'ang period. 2 The helmet and the coat are roughly figured in T'u shu tsi ch'eng (Fig. 48) ; but only the former is explained by a note to the effect that for each single piece five or six catties s of 1 In Japan, however, specimens of such armor, though very rare, do occur. BASHFORD DEAN (Catalogue of Japanese Armor, p. 52) has figured one exactly corre- sponding to the sheet armor of our clay statuettes. It is said to date about 1500, and "this form simulates the naked body and is known as the Hotoke-do (saint's breastplate), an Indian saint being often represented with the body naked." This term means "Buddha's breastplate (Hotoke= Chinese Fu, "Buddha"), and clearly indicates that this armor was made in imitation of that represented on Buddhist statues. Among modern Indian armor, a very similar type is still found (W. EGER- TON, Illustrated Handbook of Indian Arms, Plate XII, No. 587, and p. 124). A somewhat different type of iron sheet armor is figured by W. GOWLAND (The Dol- mens and Burial Mounds in Japan, p. 48, Westminster, 1897; the same also in YAGI SHOZABURO, Nihon Kokogaku, n, p. 153, Tokyo, 1898; and N. G. MUNRO, Prehistoric Japan, pp. 396, 417, Yokohama, 1908). It is likewise a harness composed of plastron and dossiere which are formed of horizontal plates of iron skilfully forged and clinched together with iron rivets. Gowland makes the interesting and correct observation that both body armor and helmet are entirely different in form and con- struction from those of historical times, but that they agree very closely with the armor represented on the terra-cotta figures called haniwa. It is very interesting that the two TORII, in the publication previously mentioned (Etudes archgologiques, Journal College of Science, 1914, p. 73), figure such a haniwa with the description "cuirasse de style europe"en trouve'e en Musashi, Japon." The Japanese authors, accordingly, are struck by the "European" character of this armor. It is now obvious that it has reached the East by way of Turkistan: consequently the haniwa adorned with this style of armor cannot be older than the age of the T'ang dynasty. Again we see in this example that the chronology of Japanese antiquities is in need of revision. * AMIOT (Supplement a 1'art militaire des Chinois, Memoires concernant les Chinois, Vol. VIII, p. 373, Paris, 1782) was the first to describe this armor, but from a different source. Amiot styles it "cuirass in imitation of the skin of the animal called ni (resembling, it is said, the lion)." 1 The T'u shu tsi ch'eng, deviating from its ordinary practice, does not state the source of this passage, which is evidently not extracted from a contemporaneous record of the T'ang period, which, however, seems to go back to a tradition of that time. The catty (kin) of the T'ang period is not identical with the present one. In the Museum collection there is a spherical bronze weight of the T'ang period (Cat. No. 1 16,892) inlaid with gold speckles and engraved with an inscription (the grooves of the characters being laid out with gold foil) yielding the date 672. The weight is stated in this inscription as being i pound (catty) 8 ounces, while it is 2 pounds in our weight. According to the present Chinese standard, it weighs I pound 11.32 ounces, or 27.32 ounces. Consequently I ounce of the T'ang period is equal to 1.138 modern Chinese ounce, and I pound of the T'ang period is equal to 18.24 ounces modern. 302 CHINESE CLAY FIGURES FIG. 48. "Lion-Armor" and "Lion-Helmet" of the T'ang Period (from T'u sku tti ch'titg). FIG. 49. "Lion-Helmet" of the T'ang Period (from Wu pei chi). DEFENSIVE ARMOR OF THE TANG PERIOD 303 pure iron mixed with one catty of steel are required, and that a skin cut in five layers, to the weight of two catties, is laid around this founda- tion. 1 The term ni k'uei is not interpreted in this passage; but this word ni occurs only in the combination suan-ni, designating the lion. We noticed above (p. 276) that Ma Sui of the eighth century fashioned helmets in the shape of lions. A lion-helmet (suan-ni mou) is mentioned as having been in the possession of General Han Shi-chung, who died in usi. 2 A similar helmet with the same designation is illustrated also in the Wu pei chi 3 (Fig. 49) ; and the descriptive text there given is identical with that of T'u shu tsi ch'eng; nevertheless the illustration of the latter is not derived in this case from that book, as the knob of the helmet and the number of plate-rows in the attached coif of the helmet are different, being six in the T'u shu, and five in the Wu pei chi. It will be noticed that the triangles on the plates are alternately drawn point upward or downward, and that the T'u shu begins with points down- ward, the other book with points upward: the two sketches must there- fore come from different sources. Still more curious is the fact that the Wu pei chi 4 illustrates an armor of a different design under the same name, T'ang ni k'ai (Fig. 50) . While the two drawings agree fairly well in the upper portions, the breast and sleeve coverings, they are considerably at variance in the middle and lower parts, though notwithstanding both evidently represent the same type of armor. The cut of the Wu pei chi is identical with the one figured by AMIOT; S and the quaint text supplied by him is found there also. It runs thus: "The lion-armor of the T'ang. First, five catties of the ' plant penetrating into the bones' 6 and three catties of radish- seeds are mixed into a pap which is placed in clear water to the quantity of a hundred catties, and boiled till it bubbles two hundred times. The residue is removed, and five scales of the pangolin 7 are added; further, 1 Several designations for other kinds of helmets are added, and it is further said that in the south also old cotton is used in their making. 1 GILES, Biographical Dictionary, p. 251. The passage alluded to above is con- tained in the biography of Han Shi-chung in the Annals of the Sung Dynasty (Sung ski, Ch. 364, p. 6 b). A "lion-armor" (shi-tse kia) is mentioned in the Annals of the Yuan Dynasty ( Yuan ski, Ch. 79, p. 24 b, K'ien-lung edition). 1 Ch. 105, p. 6. 4 L. c. t p. 15. L. c., Plate XXVIII. Tou ku ts'ao, identified with Mercurialis Itiocarpa, an euphorbiaceous plant (G. A. STUART, Chinese Materia Medica revised from P. Porter Smith's Work , p. 263, Shanghai, 1911). T Ch'uan shan kia, the scaly ant-eater (Manis tetra dactyla). The word ch'uan is here written with the character ' river ' (No. 2728) instead of No. 2739. This animal 34 CHINESE CLAY FIGURES rrmmmmti PIG. 50. 'Lion- Armor" of the Tang Period (from Wu pet cht). DEFENSIVE ARMOR OF THE T'ANG PERIOD 305 three catties of salt of Ta-t'ung, three catties of saltpetre, five ounces of stony nitre, and half a catty of sal-ammoniac. This mixture is tightly shut up in a kettle, and boiled for a day and night. Then the kettle is opened, and the mass is beaten with a leather ladle to secure various grades of thickness, and formed into the shapes of willow-leaves, fish- scales, square leaves, and rectangles. This armor has the advantage of being light in weight, and is much employed in the south." This is apparently an alchemical recipe intended to produce a cut- proof body protection. The ingredients like the scales of the pangolin rest on sympathetic notions. Of course, it should not be understood with Amiot that the armor was manufactured from this substance; the illustrations show that the question is that of a substantial metal plate armor, although in the text it is a question of scales, and that the metal plates were covered with this essence. The idea of rendering the wearer invulnerable was perhaps responsible for the title of "lion-armor;" and this name, which conveys the impression of a rendering of Sanskrit simhavarman, savors of Indian-Buddhist influence. Indeed, on ex- amining closely the two designs of this armor, we cannot fail to notice that it is identical with the one represented in the late Buddhist art of China during the Ming period, especially in the statues of Wei-t'o (Veda) and the Four Heavenly Kings, the guardians of the world and armed defensors of the Buddhist religion. Numerous specimens of these in all dimensions, carved from wood or cast in bronze, are in the Museum's collection; whatever their artistic and scientific interest may be, they have no value for the study of body armor which is mechanically copied in various conventional and stereotyped designs not properly understood by the artists. is an inhabitant of Fukien Province and Formosa, and has its trunk, limbs, and tail covered with large, horny, imbricated scales, which it elevates in rolling itself into a ball when defending itself against an enemy; the scales are medicinally employed (see J. H. EDWARDS, China Review, Vol. XXII, p. 714). Regarding the word "pan- golin" see YULE and BURNELL (Hobson-Jobson, p. 668), and A. MARRB (Petit Vocabulaire des mots malays que 1' usage a introduits dans les langues d'Europe, p. 1 1 , Rome, 1866). VII. HORSE ARMOR AND CLAY FIGURES OF HORSES Steeds shielded with armor are alluded to as early as the Shi king. It appears that horses harnessed to the war-chariots were sometimes covered at that period with a means of defence, 1 which, judging from the use of the word kiai (compare p. 195) in this connection, seems to have been of the type of scale armor, the scales being cut out of thin strips of hide or leather. During the Ch'un Ts'iu period, the horses of the war-chariots were likewise armored. 2 This horse armor of the archaic epoch was a plain caparison, and widely different from the com- plex and composite armor which, as we know with certainty, existed in the Mongol period. As to metal armor for horses (ma k'ai), we hear it mentioned for the first time toward the end of or shortly after the Han, in two small compositions of the famed usurper Ts'ao Ts'ao, who died in 220 A.D., and of his son Ts'ao Chi (192-232). The latter says that the ancient emperors bestowed on their servants certain kinds of armor styled "shining like ink" (mo kuang) and "brilliant lustre" (ming kuang), an armor with double seat in the trousers, an armor with rings and chains, and a set of horse metal armor (ma k'ai). This passage is very sus- picious because of its retrospective character: the metal armor (k'ai), while it existed at the author's time, had not yet appeared in the days of the early emperors; and the word is here used thrice consecutively with reference to them. The "ring and chain armor," as previously 1 LEGGE, Chinese Classics, Vol. IV, pp. 131, 194. LEGGE translates in the one case "the chariot with its team in mail," and in the other case "his mail-covered team," explaining that the mail for the horses was made of thin plates of metal, scale-like. This interpretation is erroneous. The same misconception occurs in S. COUVREUR'S translation of the Shi king (p. 136), "les quatre cheyaux munis de minces cuirasses de me'tal," and is adopted by GILES (No. 1734); while in the other passage COUVREUR (p. 90) is correct in translating "les quatre chevaux munis de cuirasses," provided cuirasses is taken in its literal sense of "hide armor." It is impossible to assume that during a period when metal armor for the protection of the human body was entirely unknown, it should have been utilized in guarding a horse. Man of that age could conceive and employ no other armor for his horse than for himself; and since he was acquainted only with plain hide armor and hide scale armor, these two types must have served likewise for the horse, the term kiai being in favor of scale armor. The translations of the two passages of Shi king have to be corrected accordingly. The frontlets on the foreheads of the horses (yang, No. 12,882) , once mentioned in Shi king (LEGGE, Chinese Classics, Vol. IV, p. 547) and once in Tso chuan, did not form part of an armor, but were metal ornaments which served for purely decorative purposes, and emitted pleasing sounds when the animal moved. * LEGGE, /. c., Vol. V, p. 345. 306 HORSE ARMOR 37 pointed out (p. 174), is an isolated instance in this period, and smacks of anachronism. For this reason also the metal horse mail must be looked upon with diffidence, and I am not inclined to attribute much importance to this text. PIG. 51. Armored Cavalier on Caparisoned Horse. Clay Figure in Collection of Mr. G. Eumorfopoulos. London (after Burlington Pine Arts Club, Exhibition of Early Chines* Pottery, Plate iv). In 519 A.D., A-na-kuai, the King of the Juan-juan, 1 presented to the Emperor Su-tsung of the Wei dynasty one set of fine and brilliant 1 mail complete for man and horse (j&n ma k'ai), and six sets of iron mail for man and horse. 8 Caparisoned war-horses are repeatedly mentioned in the History of 1 He committed suicide in 552, after having been vanquished by the Turks (HIRTH, Nachworte zur Inschrift des Tonjukuk, p. no). * This attribute is invariably used with reference to iron armor with varnished or polished plates. 1 Pet shi, Ch. 98, p. 6. 308 CHINESE CLAY FIGURES the T'ang Dynasty. The rebel Kao K'ai-tao, who conquered Yu-yang in 618 and styled himself Prince of Yen, for example, was in possession of several thousand mail-clad horses and ten thousand men. l Among the types of armor officially established by the T'ang dynasty we find also "horse cuirasses" (ma kia); and a charger caparisoned in this manner appears in a contemporaneous clay figure (Fig. 51) coated with a yellow glaze. The armor covers the war-horse almost down to its knees; and as it appears as a solid mass without any divisions, it may be one of hide (also the rider apparently wears a hide armor) ; it is possible, however, that the hide is merely the exterior cover, and is placed over an armor of solid plate mail indicated by the row of laminae along the lower edge. 2 Under the Sung dynasty the horses received facial masks of copper. 3 According to Ts'e fu yuan kuei, Chang Yen-tse", Governor of King- chou, 4 presented in 942, on his arrival at the capital, in order to show his gratitude for favors received, nine horses, and again fifty horses to- gether with silver saddles and bridles, and iron armor for the protection of the faces of horses and men; at a later date he presented fifty horses with gold saddles and bridles, with complete armor for the horses and men. The furniture of the horses of the Mongols is described by the Franciscan Piano Carpini in 1246* It was of two kinds, iron plate mail, as described in Chapter V, and leather scale armor. The latter consisted of five parts, the body armor in two halves extending from the head to the tail, and fastened to the saddle, a protection for the croup, a neck-guard, a breastplate reaching down to the knees, and an iron lamina on the forehead (being the chanfrin). In another passage the same writer says that many of the horses of Kuyuk had bits, breastplates, saddles, and cruppers, quite twenty marks' worth of gold. 6 The Armenian historian Haithon states that the horses of the Mongols, like their riders, were clothed with leather armor. 7 Interesting illustrations depicting the single pieces making the com- plete furniture of the horse are preserved in the Wu pei chi (Figs. 52-54) 1 Tang shu, Ch. 86, p. 4 b. 1 Also among the Moghuls the horses were first covered with mail, over which was put a decorated quilt (see H. BLOCHMANN, Ain I Akbari, Vol. I, Plate XIV, and the explanation on p. xi). * Sung shi, Ch. 197, p. 2. 4 In Kan-su Province (PLAYFAIR, Cities and Towns of China, 2d ed., No. 1112). 8 Edition of G. PULL, p. 87 (Studi italiani di filologia indo-iranica, Vol. IX, Firenze, 1913). This passage is lacking in the former editions of Carpini. * W. W. ROCKHILL, The Journey of William of Rubruck, p. 20. 7 G. ALTUNIAN, Die Mongolen und ihre Eroberungen, p. 81 (Berlin, 1911). HORSE ARMOR FIG. 52. Chanfrin and Armor for the Croup of a Horse (from W* pei chf). 3 io CHINESE CLAY FIGURES FIG. 53. Neck-Guard and Breastplate of Horse (from Wu pet chf). HORSE ARMOR ?mgms> r .r*\ ' _j-u_j >-_jj- i^-j->- /'o 7 o,^v.o;o.o:p : o; bfe^/JLaJ-JoCH FIG. 64. Half-Chanfrin and Trunk Mail of Horse (from Wu pei chi). 312 CHINESE CLAY FIGURES of 1 6 2 1 , where no description of them , however, is given . The armor parts for the croup, neck, breast, and trunk, consist of plate mail ; they represent the tradition of the Ming period, and may be identical with those of the Yuan. It is not known to me whether horse armature was still employed under the Manchu dynasty. Fig. 55 is here inserted after CIBOT; from what Chinese source this illustration is derived I do not know. It is FIG. 55. Chinese Sketch of Caparisoned Horse (from L. P. Cibot, Lettre sur les caracteres chinois, Brussels, 1773). interesting as showing a horse with complete equipment, a facial mask or frontal with chanfrin of scale armor, neck and shoulder guards of plate mail, and a chabraque enveloping the trunk. From what has been set forth above in regard to the relations be- tween Iran and China, it appears also that Chinese horse mail might have been influenced from the same direction. This influence is very probable ; but the discussion of this matter may be left for the present, as it is preferable to wait until a thorough investigation of Iranian horse mail has been made by a competent specialist; ample material for such study is particularly furnished by the Persian miniatures. 1 1 In an illuminated manuscript of the Shah-nameh preserved in the Royal Li- brary of Munich, and representing the costume and arms of the Persians in the seventeenth century, according to Egerton, the combatants generally wear conical helmets with solid guards over the neck and ears. The horses as well as their riders have a complete covering of mail with alternate rows of gold and silver scales (W. EGERTON, 111. Handbook of Indian Arms, p. 142). In ancient India, elephants and horses were protected by armor (G. OPPERT, On the Weapons, Army Organization, and Political Maxims of the Ancient Hindus, p. 8, Madras, 1880). The Chinese CLAY FIGURES OF HORSES 313 Numerous clay figures of horses and cavaliers have been unearthed in recent years from the graves of Shen-si and Ho-nan, and a brief description of these may find a suitable place here. Particulars in regard to the history of the burial of such clay figures and their signifi- cance will be given in Part II. The observation of the local differentia- tions is an essential point of view to be pursued in the study of these clay figures. The divergence between the grave-finds of Ho-nan and Shen-si is peculiarly manifest in the horses. Those of Shen-si usually represent the bare horse in a sober and mechanical conception ; l those of Ho-nan illustrate more realistic types, always harnessed, in a variety of poses effected particularly by manifold turns of the neck. Most of the horses are posed on a flat rectangular clay base. Among seven clay horses of miniature size acquired by the writer at Si-ngan fu, six are almost identical, while the seventh is differentiated only in that the mane is coarsely fashioned. The horse on Plate LXII is an exception, being somewhat better shaped, and coated with soft lead glazes in three colors, a deep brown, a light yellow, and a plant green; also saddle and sad- dle-cloth are represented (but not the stirrups); the saddle is padded with a textile material gracefully draped on both sides. The horse shown on Plate LXII I excels by its massive dimensions, but is other- wise the outcome of the routine work of an ordinary craftsman. The Ho-nan horses, on the other hand, appeal to us by the gracefulness of their motions, and the variety of actions in which they are represented (Plates LXIV, LXV) ; also the details of the harness are better and more efficiently worked out. In the horse on Plate LXVI, the trappings with their ornaments in metal, the tinkling bells on the breastband, as well as the lotus-flower designs on the crupper, are neatly moulded in relief. The clay figure of the horse on Plate LXVII, found in fragmentary condition north of the city of Ho-nan fu in 1910, is notable for its un- usual dimensions and its perfect glazing. 2 The natural coloration of the animal is reproduced by a light-yellow soft lead glaze; the saddle, of the pilgrim Huan Tsang reports that the Indian war-elephants were covered with strong armature (S. BEAL, Buddhist Records of the Western World, Vol. I, p. 82). In Tibet the high officers sometimes clothe their horse with armor, and a set was cap- tured by the British expedition under Colonel Younghusband. A Tibetan cavalry- man whose horse is clad with chanfrin, neck and breast guard, is pictured in WAD- DELL'S Lhasa and its Mysteries (Plate opp. p. 168). 1 Sometimes a mere saddle is represented without any other trappings; such a horse will be figured in Part II as forming part of a complete set of finds from the same grave. 1 The technique and colors of these glazes are identical with those on the statue of the Arhat recently acquired by the British Museum, and ably described by R. L. HOBSON (Burlington Magazine, Vol. XXV, 1914, pp. 69-73). The excellent colored plate accompanying this article affords a good view of the T'ang potter's glazes. 314 CHINESE CLAY FIGURES same form as the one in use at present, is glazed a plant green; the double saddle-cloth underneath it, dark brown intermingled with green. The seat of the saddle is padded with a material arranged in graceful drapery. The mane is brown; the ornamental metal pieces attached to the head- stall, the breastband, and crupper are glazed green. The design which is brought out on these is characteristic of the T'ang period, and found also as relief decoration on coeval pottery vases. 1 The horses on which human figures are mounted occupy a special place. Their significance in relation to the dead may be ascertained from their position in the grave: they were found either as preceding or as following the coffin. This seems to allude to the fact that they were regarded as the mounted escorts of the occupant of the grave, in the same manner as the living one, when on an official visit riding in a cart or in a sedan-chair, is accompanied by outriders in front and in the rear. As only persons of rank were granted this privilege, it seems certain that the same rule was observed in the grave, and that the clay statuettes of cavaliers appertain to dignitaries. From Shen-si only figures of male riders are known to me (Plates LXVIII-LXX). The Shen-si horses are of somewhat stronger build, taller, and with more developed chests, than the Ho-nan breed. In the former, the curly hair on the forehead is parted and combed toward the sides, while in the latter it hangs straight downward. The men wear a pompon in the front of their round caps, and are strangely clad in long gowns. The cavalier on Plate LXVIII makes a poor figure as a horse- man, and shows that the Chinese of the T'ang period had as poor a knowledge of the art of riding as at present. The women of Ho-nan are better seated in the saddle than the men of Shen-si. The rider in question has his left foot pushed forward and his right foot backward; his hands come too near to the horse's neck, and seem to be in motion. 1 An illustration of such a vase will be found in Part II. Chinese horse- trappings of the T'ang period may be viewed in TOyei Shuko, Vol. Ill, Plates 196, 197. In none of the clay figures which have come to my notice is the saddle-girth represented. Judging from the clay figures, saddlery must have been almost the same in the T'ang period as at present. The frame of the modern saddle is carved from wood, frequently covered with shagreen and edged with metal-work, usually iron incrusted with silver wire forming geometric or floral designs. The seat is padded with a blue or red satin or velvet cover. There are, as a rule, two saddle-cloths, the lower one of wadded cotton cloth, the upper either of leather, ornamented with designs in color or appliqu6 patterns, or of wool or silk carpeting. A single bridle of cotton webbing is used. Headpiece, breastband, and crupper are usually decorated with brass work, or sometimes with silver gilt. A neckcollar fitted with small brass bells is occa- sionally added. Two tassels of red-dyed horse-hair are suspended, the one from the breastband, the other from the band under the chin. The stirrups are large and heavy with solid bases ellipsoid in shape, usually of iron damaskeened with silver, more rarely of brass. In Kan-su and north-eastern Tibet, wooden stirrups were also observed and collected by the writer; these are made as substitutes only when iron is lacking. Compare also Plate XXII. CLAY FIGURES OF HORSES 315 Whoever has observed Chinese riding will have witnessed such perform- ances; and in this case the potter must be granted all credit for his power of observation. There is another type of mounted soldier from Shen-si, whose left hand appears as if seizing the bridles, while he is pressing his right hand against his chest (Plate LXIX, Fig. 2). The figure on Plate LXX is curious in exhibiting a helmeted soldier rising in the saddle in an upright position, in order to salute by lifting his folded hands to the height of his face. The headstall of the horse is decorated with floral ornaments, probably chased in metal. In the Ho-nan types, the horses prick up their ears; their necks are elegantly curved; the manes are either upright, or falling down to the right side, and are carefully modelled. In all Ho-nan figures of riders known to me, the stirrups are represented. 1 Fig. i of Plate LXIX illustrates a female rider very well seated; the body of the clay is coated with a yellowish-green glaze, and the mane of the animal is well treated; but the form of the head is bad. In the figure on Plate LXXI the mane of the steed is painted vermilion. The woman 2 wears male attire, a girdled coat with triangular lapels (as in our man's clothing), trousers, and boots; she is sitting straight and with arms crossed, the short sleeves rendering the hands visible. The saddle-cloth is painted with small circles in black ink, and thus is presumably intended for a panther's skin. The reins and crupper likewise are so decorated, and there are a few black circles on the neck of the animal. The stirrups are repre- sented. The horse illustrated on Plate LXXII is fairly well modelled. The neck is painted red, and overstrewn with white spots. Headstall and bridle are painted in black outlines, while the crupper is brought out in relief. The muscles of the head, the nostrils, the jaws (agape), teeth, and tongue are carefully modelled. The woman, almost Japanese in expression, wears a flat cap, from which a long ribbon is floating down her back. Her dress is painted a brown-red. Her right arm is hanging down, her left hand is raised to seize the bridles. The saddle-cloth seems to be a cotton quilt. 1 As has already been shown by F. HIRTH (Zeitschrift fur Ethnologic, 1890, Verhandlungen, p. 209), stirrups were in vogue during the T'ang period; the people availed themselves of iron stirrups, those of the dignitaries were made from the metal alloy called t'ou-shi. 1 Horseback-riding was a common exercise for women in the T'ang period. Female equestrians were represented by pictorial art. Yang Kuei-fei was painted in the act of mounting on horseback (GILES, Introduction to the History of Chinese Pictorial Art, p. 50). In the Gallery of the Sung Emperors there was a picture by Chang Suan, representing a Japanese woman on horseback (Suan ho hua p'u, Ch. 5, p. 6). London.) I 1 1 1 a ;r. 3 i ^c L p B I g" c 1 c - * 00 3 s 1 a c 1 _5 "rt d ,: -3 '~-i : 'o c o +3 a rj ^ 'u c 9 X LU O 1 w n 0) 5 "o s 1 t 2 c c. g H CL i z T< I *3 1 "c S Q i ft X M Q 2 3 v, ' C i 1^ BWg'i ui M rt C la If 1-1 'd _, H? 5 5,3 B "M W o 'S. FIELD MUSEUM OF NATURAL HISTORV. nisd aig^fil owi sriJ .3 .3 .iM IX 3TAJ9 1 .q 993) HOMflA sarH owi io nbla-saoom baon&t .fcnsri mcnl .iigailT srii moiH .rarftegoJ bdassiq ^Udgii TLINGIT HIDE ARMOR. PLATE XI. AMERICAN HIDE ARMOR (see p. 183). Made from hard, tanned moose-skin of two thicknesses, the two layers being tightly pressed together. From the Tlingit, Alaska. Presented by Mr. E. E. Ayer. Cat. No. 18165. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. XI. TLINGIT HIDE ARMOR. - r. o 5 s -d a> o o- O ctf ** G W "^ 1-1 S 5 "In T3 fij jjT & O "S) G '5 'S --^ *" c8 HH . ^^ 4i r _: ei s i uJ 04 < a! ^ 1O 8 213 .fo PQ !_, la S| OT 'S O M M A3 > *5 i .3 -5 ffi a> ^ O .2 -^ 'O H veils UK Jo sn3m \B .slbnfiri wollori ,fni;s bio .oVI .ieO .mo $ ,ri Io dorri bnfi bi^uO .Jaeo .o>I .iD .mo 17 , ia 9319V9T sri i no angiaab amsa sitt worfa 1 BRONZE SWOROS of THE HAN Ptatoo PLATE XX. TWO-EDGED BRONZE SWORDS OF THE HAN PERIOD (see p. 215). Fig. I. Much-worn blade, highly polished by means of an alloy of mercury and tin (such as is employed for metal mirrors) , rhomboid guard, hollow handle. Length 45.6 cm. Cat. No. 116754. Fig. 2. Unpolished blade, solid handle. Length, 45 cm. Cat. No. 116757. Fig. 3. Blade, guard, and handle, made in one cast. Guard and knob of hilt show the same designs on the reverse side. Length, 71 cm. Cat. No. 116756. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. XX. 1 2 3 BRONZE SWORDS OF THE HAN PERIOD. FIELD MUSEUM OF NATURAL ANTHROPOU III. PL. XXI. .(dis .q 393) .mo QQ brie mo S.ssi , 3TAJ9 |3HT ^O |lo aJnBnnl .s-i .aat .ao VI .abiBug-biowa 9 toid biodnprh xttiw IT ba no (AN PtRiOD. PLATE XXI. CAST-IRON WEAPONS OF THE HAN PERIOD (see p. 216). Figs. 1-2. Remnants of cast-iron spears. Length, 122.8 cm and 99 cm. Cat. Nos. 120995, 120996. Figs- 3~4- Cast-iron swords with rhomboid bronze sword-guards. Length, 117.6 cm and 114.3 c m - Cat. Nos. 120993, 120994. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. XXI. 1234 CAST-IRON WEAPONS OF THE HAN PERIOD. X X LU p a IDEALIZED PORTRAIT OF A SOVEREIGN OF THE HUNS. M 0) TO l| "S o .0 C *I MJ3 S H 0) J3 *> J3 ** O i-i IH S- It A&> % 5 *t |.s 81 il c '3 |l si C l-i S ^ S3 ll ^ g; S J ll ^^5 . o ^1 8 * 8.S ai :i rt a +> KH tXO TO *2 H "5 S o a D O K w 41 n 11 ** C oJ 8 c J5 G s 4-> .2 a I c S w 3 & s -u 1 03 | 8 s 8 OJ 'a & ^3 g 5 _G C c3 -u t t :d to give an idea of the Chinese conception of a sovereign of the Huns (see p. 224). t nted on the hunt. His keen eyes have spied a bird in the branches of a tree, and he ix the arrow to the bow-string. The string passes through the sleeve of his left arm, s aes not have to hold the bow while trotting or galloping. He wears a turban, large ea high boots. His long under-garment displays a checkered design, and may be compose llternately black and white squares, such as we still find among the tribes of eastern ar beria. His cloak is sleeveless, buttoned in front, and with girdle. The horse is furnish* >uble saddle-cloth, an ornamented rug, and a leather (or felt) cover. The saddle S 1 1 I 43 41 O c .22 w * d, a I 41 $ 'J3 a 3 a 1 *8 8 9 t-> *j GO C TO O 43 a C/] VH I a ^ cu irger reproduction of the portrait. The entire composition may be viewed in L. Binyo in the Far East, 2d ed., Plate VIII (London, 1913). ; collection of Sir William van Home, Montreal, Canada. Secured through Mr. Stephs i, to whom I am indebted for a photograph of the painting. ii I FIELD MUSEUM OFNATU ANTMftQPCX. iliXX ..<_ .q il/. vd aifirT . ,-noD .8 .;< ;, .o PERSIAN CHAIN MAIL. FRONT View PLATE XXlll. PERSIAN CHAIN MAIL (see p. 244). Made of twisted iron wire, with helmet. Obtained at Tiflis by Mr. Charles R. Crane, Chicago, and now in the possession of Dr. Charles B. Cory, Chicago. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. XXIII. PERSIAN CHAIN MAIL. FRONT VIEW FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. XXIV. PERSIAN CHAIN MAIL, BACK VIEW. ;AL HISTORY i .VXX 3TAJS I no nworia iisM nisdO neiaig*! at nignobd .( -q 338) astelq gnibsogiq ow* sd) no stngmamp srfT .9lbnBri noii bn 9 Jbfiid ba^a riJiw biowa bggbg-owT . i .giq I .isvlia rfiiw .olbrtBrf oriJ ; : bio rlctiw bgJatnoni 9iB gbsld 9dJ .91^*3 'j- t -yd bennol 3} slDBd ariT Jifirn to JoI^nti/BO .^ .gi 1 ? aifido Io i9B ?iania B Io adaianoo mleq gdJ gnitostoiq Item 9riT .mo 8 1 (dJansJ .9irw noii moil 2-4 OUTFIT BELONGING TO PERSIAN CHAIN MAIL PLATE XXV. Outfit belonging to Persian Chain Mail shown on the two preceding plates (see p. 244). Fig. I. Two-edged sword with steel blade and iron handle. The ornaments on the blade are incrusted with gold; those on the handle, with silver. Fig. 2. Iron arm-guard, with representations of four scenes in Persian style. Fig. 3. Hauberk, consisting of a coif of mail, suspended from a wadded cotton quilt. Width, 26 cm. Fig. 4. Gauntlet of mail. The back is formed by. red cotton stuff, lined with chamois leather. The mail protecting the palm consists of a single layer of chain twisted from iron wire. Length, 18 cm. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. XXV. 1 2-4 OUTFIT BELONGING TO PERSIAN CHAIN MAIL S" 1 3 ? w 06 j 25* 1=1 313 SgS B-a "tt T3 B ) Hi > s SJ C H fe "8 S ^^ -43 ^3 .2 o t/7 -*j G O, . C O 5 -2 o Z <-> '3 . O * r3 a rt ^ ^ r ^ a g s d '9 rt . c 03 C -so-^o * a o - s o w | w .c 2 M S PJ v ^ n ( &i 3 M-M .i ?n -> V FIELD MUSEUM OF NA POLOGY, VOL. XIII., PL. XXVII. IIVXX 3TAJR J^^ .q 998) 2CIJ3IH8 VJAT3UIT mol ditvr batasoiBmo has .abirf-aoisoommiojiro itso bteiite xavnoO oini bohoqmi bns aibal ni beiirtOBkrnBm sisw bnbl sidt io abbiriS 8J airiT .alioo icluoib lo .1 - .oVI .isO .antsdiT \o bbid8 io bbirfa .s TIBETAN S^ PLATE XXVI I. TIBETAN SHIELDS (see p. 257). Fig. I. Convex shield cut out of rhinoceros-hide, and ornamented with four brass bosses. Shields of this kind were manufactured in India and imported into Tibet. Cat. No. 122178. Fig. 2. Shield of rattan, plaited in the basketry style of circular coils. This is the national shield of the Tibetans. Cat. No. 122179. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. XXVII. TIBETAN SHIELDS. FIELD MUSEUM OF NATURAL HISTORY. ANTHROKH II, PL. XXVIII. .IIIVXX .q 993) xawjaH viAxaaiT gninnol ,9iiw lavlia bns blog iltrw botarnoni .aJaarie ba^a lo bsaoqmoO -aaon A .aqBn arlJ io noboaioiq artt 10^ *i oi barioB^Jfi si Item to iioo A .a ni jaaon aril to noitostoiq 9ff^ tol 89'naa ,nwob bfte qv %aibilz ,inoil ni (Isasn) bns sibnl ni baiutocfejfHjrn siaw aqyi airl^ to eJsmisH .rrv/ob si Ho .*9dtT o^ni J-PERSIAN STYLE. PLATE XXVIM. TIBETAN HELMET (see p. 257). Composed of steel sheets, incrusted with gold and silver wire, forming floral designs. A coif of mail is attached to it for the protection of the nape. A nose- guard (nasal) in front, sliding up and down, serves for the protection of the nose; in the illustration it is down. Helmets of this type were manufactured in India and imported into Tibet. Cat. No. 122180. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. XXVIII. TIBETAN HELMET OF INDO-PERSIAN STYLE. * FIELD MUSEUM OF NATURAL HISTORY. jftat&rt. VOL. xiii, PL. xxx. noil - ~v-t.- i > r jtM .JsO .tno^.ot ,Jrfi9H ' rfJiw CLAY FIGURE OF SOLDIER, FRONT AND BACK VIEWS. PLATE XXX. CLAY FIGURE OF SOLDIER (see p. 277). Both front and back views are shown. He is clad with armor in combination with costume. T'ang period (618-906). Height, 20.5 cm. Cat. No. 117916. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. XXX. CLAY FIGURE OF SOLDIER, FRONT AND BACK VIEWS. . ,^J * S s .3 p rt O, a 4) C -K/l X vT _i 2 5 U FIELD MUSEUM OF NATURAL HISTORY. ANTHROK CLAY FIGURES OF PLATE XXXli. ARMORED KNIGHTS (see p. 278). Clay figures from Ho-nan Province, of T'ang period (618-906). Height, 35 cm. Cat. Nos. 118063, 118068. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. XXXII. CLAY FIGURES OF ARMORED GUARDIANS OF THE GRAVE. FIELD MUSEUM OF NA ! ANTHROPOLOUY, VOL W- IIIXXX 3TAJ9. vq<*' ;> >j / J.<. rfq Tifihfsoa ov/r .o-jfrr/oil is-r bnn ,li bohoq s n 'T ;mo p x .| PLATE xxxni. MARBLE MOCK-GATE (see p. 279). This formed the entrance to a tomb, and was dug up in the environment of the city of Hien-yang, Shen-si Province. Two soaring phenixes are carved in flat relief on the lintel. The gate is marked by lines and kept closed by means of a bolt, brought out in high relief. In each of the two wings is finely traced the figure of a guardian completely armored with plate mail, and handling a sword. Height, 52.5 cm; width, 34.5 cm; T'ang period (618-906). Thickness, 8.2 can. Cat. No. 121623. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. XXXIII. MARBLE MOCK-GATE OF A TOMB. FIELD MUSEUM OF NATURAL HISTORY. XXXIV. HJiw baisvoo ei bate grii , s&d omulq sriT e no baouboiqai ad 3 .gniniliotni HB BB boiiaani iw ,inoil adi no bavBigna srf^ no bsinsasiqat nsad ion J3 IB bonratdO . HORSEMAN'S SUIT OF ARMC* PLATE XXXIV. CHINESE PLATE ARMOR (see p. 284). Horseman's uniform, of K'ien-lung period (1736-95). The skirt is covered with four parallel rows of light and elastic steel laminae. In the coat, the steel plates are inserted as an interlining. Steel helmet, surmounted by velvet plume, dragons being engraved on the front, with silk covers for neck, ears, and occiput. The plume has not been represented on the Plate, in order that the suit might be reproduced on a larger scale. Obtained at Si-ngan. Cat. No. 118344. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. XXXIV. HORSEMAN'S SUIT OF ARMOR. FIELD MUSEUM OF NATURAL HISTORY. -91 91 -blog brie faanil temforf I39i3 .abu t bamsft orfj lo ii loiq arii 10! ylno banir.i xftiw baaerio bne ,batlq bssario bas ,iliup ARTILLERY-MAN'S SUIT OF ARMOR. PLATE XXXV. CHINESE PLATE ARMOR (see p. 285). Artillery-man's uniform, of K'ien-lung period (1736-95). The plates are re- tained only for the protection of the shoulders. Each lamina is of steel and gold- plated, and chased with a four-clawed dragon soaring in clouds. Steel helmet lined with quilt, and chased with gilt figures of dragons in pursuit of the flamed jewel. Obtained at Si-ngan. Cat. No. 118346. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. XXXV. ARTILLERY-MAN'S SUiT OF ARMOR. FIELD MUSEUM OF NATURAL HISTORY. dJrw bsiavoo zi io JB bdnifiJdO . bfioid .nBgn-i8 AUCHER'S SUIT OF ARMOR, FRONT VIEW. PLATE XXXVI. CHINESE PLATE ARMOR (see p. 286). Archer's uniform, of K'ien-lung period (1736-95). The interior is covered with broad steel plates, and the shoulders are protected by brass plates. Obtained at Si-ngan. Cat. No. 118345. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. XXXVI. ARCHER'S SUIT OF ARMOR, FRONT VIEW. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. XXXVII. ARCHER'S SUIT OF ARMOR. BACK VIEW. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY uJ >.i ymuJEo.o a no ion ^-rrr/-,^ UNIFORM OF PALACE GUARD-OFFICER, FRONT AND BACK VIEWS. PLATE xxxvni. CEREMONIAL UNIFORM (see p. 286). Belonging to guard-officer of the first rank detailed on duty in the Imperial Palace. The costume is magnificently embroidered with heavy gold thread, and studded with gilt bosses. The shoulder-plates are arranged in the same manner as in the suit of armor on Plate XXXV. Cat. No. 32853. Helmet, bow-case, and quiver belonging to this uniform are represented on the following two Plates. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. XXXVIII. 1 2 UNIFORM OF PALACE GUARD-OFFICER, FRONT AND BACK VIEWS. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. XXXIX HELMET OF PALACE OFFICER. FIELD MUSEUM OF PLATE'XLI. KOREAN PLATE ARMOR (see p. 288). Front and back of the coat are strengthened in the interior by seven parallel rows of rectangular steel plates, coated on both sides with a black varnish. Length, 81 cm. Cat. No. 33281. The following Plate illustrates the interior of this suit, with the iron casque. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. XLI. KOREAN STEEL PLATE ARMOR, EXTERIOR. FIELD MUSEUM OF NATURAL HISTORY. -Biooab HB SB batosani SIB sstelq isqqoo ni .temferf adi at on as inimal 10 slfivrvme evil ni ,m i iuodfi . KOREAN COURT COSTUME OF HIGH 0? PLATE XLlll. KOREAN PSEUDO-PLATE ARMOR (see p. 289). It has no plates, but the rows of brass bosses on the surface of the coat are decora- tive survivals or reminiscences of plate armor. Thin copper plates are inserted as an interlining in the ear-muffs and nape-guard attached to the helmet. Length, about I m. Cat. No. 33263. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. XLIII. KOREAN COURT COSTUME OF HIGH OFFICIAL. FIELD Ml ANTHROPOLOGY, VOL. XIII. PL. XUV v ORPHIC FORM OF THE Goo OF DEATH. PLATE XLIV. YAMA, THE GOD OF DEATH (see p. 294). He stands on the body of a sow, and is represented with the head of a horned bull, and with eagle-claws on his hands and feet. His head is surrounded by flames. Clay figure from Shen-si, of mediaeval times, probably T'ang period (618-906). Traces of red pigment; eyeballs painted black. Height, 60 cm. Cat. No. 117987. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. XLIV. ZOOMORPHIC FORM OF THE GOD OF DEATH. j.tno^T ; bntroic griibniv/ ZOOMORPHIC FORM OF THE GOD OF D 1 PLATE XLV. YAMA, THE GOD OF DEATH (see p. 295). He stands on the body of a bull, and is represented with the head of a horned bull, surrounded by flames. A snake is winding around his left arm. Clay figure from Shen-si, of mediaeval times, probably T'ang period (618-906). Height, 34 cm. Cat. No. 117985. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. XLV. ZOOMORPHIC FORM OF THE GOD OF DEATH. FIELD MUSFUM OF NATURAL HISTORY. .IVJX 3TAJ9 ilJT ,A rtT .^3 oiil-Ifimin bn .1 ri^w b* ^ bri ,Rlld3Y') ,morf ;l- bohaq ZOOMORPHIC FORM OF THE Goo OF DEATH PLATE XLVI. YAMA, THE GOD OF DEATH (see p. 295). In the same pose as the two preceding figures, but without a bull. Demoniacal face with human traits and animal-like ears. The head is surmounted by a long, twisted horn. He is clad with a leopard-skin, indicated by rows of black and red circular spots. Face painted red; horn, eyeballs, and beard, black. Clay figure from Shen-si. T'ang period (618-906). Height, 47 cm. Cat. No. 117988. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. XLVI. ZOOMORPHIC FORM OF THE GOD OF DEATH. FIELD MUSEUM OF NATURAL HISTORY. 1HOPOLOGY, VOL. XIM, >: frljjiml i 1 2 INTERMEDIARY AND HUMAN FORMS OF THE GOD OF OF*- PLATE XLVII. CLAY FIGURES FROM SHEN-SI AND HO-NAN (see p. 295). Fig. I. Intermediary form of the God of Death. His head is modelled in the style of the bull-faced Yama. as shown on Plates XLIV and XLV, but he is equipped with armor in the same manner as the human forms. He stands over the figure of a demon, and seems to have grasped a weapon in his right hand, which is perforated. Clay figure from Shen-si. T'ang period (618-906). Height, 45 cm. Cat. No. 117998. Fig. 2. Fragmentary clay figure from Ho-nan, of armored knight with plumed head-dress, a type evolved from Yama as triumphant warrior. Here inserted for comparison of the head-dress with that in Fig. I . Height, 31.5 cm. Cat. No. 117994. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. XLVII. 1 2 INTERMEDIARY AND HUMAN FORMS OF THE Goo OF DEATH. FIELD MUSEUM OF NATURA! HISTORY. ANTHROPOLOGY, VOL. XIII, PL. XLVIII. s to Jfiftt gi abol lormB Aitvr bsIO tfilq gniwollol ariJ i ,BOTBY to PLATE XLVlll. FORM OF THE GOD OF DEATH (see p. 295). Intermediary between the zoomorphic and anthropomorphic types. The atti- tude is that of a triumphant victor, standing over the figure of a crouching demon. Clad with armor and an elaborate head-dress, like the figures of knights shown on the following plates, he shares the two-horned bull-head with the purely animal forms of Yama, illustrated previously. Clay figure from Shen-si. T'ang period (618-906). Height, 68 cm. Cat. No. 117993. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. XLVIII. THE GOD OF DEATH. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII. PL. XLIX. HUMAN FORM OF THE Goo OF D PLATE XLIX. THE TRIUMPHANT GOD OF DEATH (see p. 295). He is represented as a knight with complete armor, standing on the figure of a demon. The figure is coated, except the head, with glaze in four colors, green, blue, brown, and yellowish white. Clay figure from Shen-si. T'ang period (618-906). Height, 52.6 cm. Cat. No. 118000. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. XLIX. HUMAN FORM OF THE GOD OF DEATH. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. L. OWi 91B 919rfT .1OJ bnuol O9ad gvsri oi lo uui ) MSG bn.fi ,Ilud g THE GOD OF DEATH. PLATE L. HUMAN FORM OF THE GOD OF DEATH (see p. 296). Posed on the back of a reclining bull, and clad with sheet armor. There are two identical specimens of this figure in the Museum collection, said to have been found in the same grave. Clay figure from Shen-si. T'ang period (618-906). Height, 67 cm. Cat. No. 118006. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. L. THE GOD OF DEATH. FIELD MUSEUM OF NATURAL HISTORY. "TOPOLOGY. VOL. XIII. PL. LI. J 3TAJS id) ixtnsq v HUMAN FORM OF THE Goo OF DEATH. PLATE LI. HUMAN FORM OF THE GOD OF DEATH (see p. 296). Posed on the back of a reclining bull, and clad with sheet armor. His left arm is akimbo, and his right hand is raised as though throwing a weapon. Clay figure from Ho-nan. T'ang period (618-906). Height, 39 cm. Cat. No. 117989. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. LI. HUMAN FORM OF THE GOD OF DEATH. FIELD MUSEUM OF NATURAL HISTORY. TOPOLOGY, VOL. XIII, PL. Lll. PLATE Lll. HUMAN FORM OF THE GOD OF DEATH (see p. 296). The figure of the bull is lost, but may be supplemented in accordance with the figure in the preceding Plate, with which it agrees in pose and general style. It is , however, much more artistic. The face is well modelled and very expressive. Note the mustache with turned-up tips. The clay piece, which appears dark on the Plate, is a recent supplement. The entire clay figure, with the exception of the head, is glazed in three colors, green, brown, and yellowish-white. From Ho-nan. T'ang period (618-906). Height, 68.8 cm. Cat. No. 118069. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. LM. GLAZED FIGURE OF THE GOD OF DEATH. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY :-H ft K) .mo 8; CLAY FIGURES OF THE GOD o PLATE Llll. THE GOD OF DEATH (see p. 296). Fig. I. Of the same type and style as the clay figure on Plate LI, only without helmet. His hair is parted and bound up in a chignon. Fig. 2. In this figure, the pose of hands and feet is reversed, the right arm being akimbo, and the left one being raised. He stands on the body of a demon. Clay figures from Ho-nan. T'ang period (618-906). Height, 40 and 38 cm. Cat. Nos. 117876, 117991. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. Llll. 1 2 CLAY FI3URES OF THE GOD or DEATH. FIELD MUSEUM OF NATURAL HISTORY. TOPOLOGY, VOL. XIII. PL. LIV. k> .nomab lo v ^bod osou/d B lo 3iv- L 3o bo) > ,3WOld ,29^9 too ariJ jgrbiti'-l .boO 9ftijfl ^pPRs oa ;>ald HI " i 9V39k arfl to noitioq 01 gnisd allfidav ngiq bst e rl^rw baiavoo ai ) bohaq gne'T .nsn- :u .o'/I JiO .mo f.o 1 2 CLAY FIGURES OF THE GOD OF DEATH. PLATE LIV. THE GOD OF DEATH (see p. 296). Fig. i. The God of Death trampling on the body of a demon, of the same style and pose as Fig. 2 on the preceding Plate. From Ho-nan. T'ang period (618-906). Height, 38.3 cm. Cat. No. 118065. Fig. 2. The God of Death trampling on the figure of a human body (probably child), coated with a thick layer df white pipe clay, eyes, brows, nose, and mouth being painted in black; so are also the boots of the God. Further, the outlines of his eyes are black (the eyeballs being red). The middle portion of the sleeve of his right arm is covered with a red pigment. From Ho-nan. T'ang period (618-906). Height, 29.3 cm. Cat. No. 117995. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. LIV. 1 2 CLAY FIGURES OF THE GOD OF DEATH. FIELD MUSE' AL HISTORY. . ton ai 9iufi 3 ^H^HH5fiSfe4iud ~fto amJb;t6J? ,J biiog ,9itrg^l ^Hft^ |oi ,b9l moil .(dop-8i LEAD FIGURE OF THE GOD OF DEATH. PLATE l_V. THE GOD OF DEATH (see p. 296). Represented as armored knight, standing on a bull or a demon (the figure is not sufficiently distinct to allow of positive identification). Miniature figure, solid cast from lead, in high relief; the back is flat. From Shen-si. T'ang period (618-906). Height, ii cm; width, 4.3 cm; thickness, 2.2 cm. Cat. No. 117091. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. LV. LEAD FIGURE OF THE GOD OF DEATH. AL HISTORY. u .; Jo t arfo no )uo frfgjjoid g SOB! zid lo bns J jft boh PLATE LVl. GUARDIAN OF THE GRAVE (see p. 297). Knight or warrior clad with sheet armor, animal-heads being brought out on the sleeves. In the point of armor, in the weird and demoniacal expression of his face (he is represented as shouting), and in the style of his chignon (compare Plates L and LIII, Fig. i), he reveals his affinity with Yama, the God of Death. Clay figure from Ho-nan, of unusual dimensions. T'ang period (618-906). Height, 79.7 cm. Cat. No. 118154. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. LVI. CLAY FIGURE OF ARMORED KNIGHT. FIELD MUSEUM OF NATURAL HISTORY. nvj .loans Jsads lo ^bnJa orf) ods oriJ "^IW snimun aofl&i aixfT :oo IB bns .albbim six! ni aoqsdw ibom-IbW boh id . ota adi ni bsni zblod sH CLAY FIGURE OF ARMORED QUAP PLATE LVII. GUARDIAN OF THE GRAVE (see p. 297). This figure affords a good example for the study of sheet armor. Plastron and dossiere are conspicuously represented, each consisting of two halves joined in the middle, and are connected by leather straps running over the shoulders. He holds a weapon in his right hand. Well-modelled clay figure from Ho-nan, with traces of red pigment. T'ang pe- riod (618-906). Height, 61.9 cm. Cat. No. 118008. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. LVII. CLAY FIGURE OF ARMORED GUARDIAN. ANTHROPOLOGY 302) ."IVAHxJ OTT ruin-op! .mo d. CLAY FIOURE OF ARMORED GUARDIAN. PLATE LVlll. GUARDIAN OF THE GRAVE (see p. 297). Warrior clad with sheet armor, cape, and hooded helmet. Clay figure from Ho-nan. T'ang period (618-906). Height, 33.6 cm. Cat. No. 118013. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. LVIII. CLAY FIGURE OF ARMORED GUARDIAN. FIELD MUSEUM OF NATURAL HISTORY di hns PAIR OF ARMORED GUARDIANS. PLATE LIX. GUARDIANS OF THE GRAVE (see p. 297). Warriors clad with sheet armor and hoods. They wear bushy mustaches with turned-up tips. Clay figures from Shen-si. T'ang period (618-906). Height, 45 and 46 cm. Cat. Nos. 118061, 118062. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. LIX. 1 2 PAIR OF ARMORED GUARDIANS. FIELD MUSEUM OF NATURAL HISTORY. ntinaaaiqdfr ,> SrTeiH .temlsd s> verf oJ arrraae 7B9q3 nofx>o T .dj^^H no8ii FRAGMENTARY CLAY FIGURE OF ARMORED GUARDIAN. PLATE LX. GUARDIAN OF THE GRAVE (see p. 297). Upper portion of clay figure, representing shouting warrior clad with sheet armor, shoulder-guards, and hood-like helmet. His right fist has an aperture (made by means of a drill), in which a wooden spear seems to have been inserted. From Ho-nan. T'ang period (618-906). Height, 34 cm. Cat. No. 118011. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. LX. FRAGMENTARY CLAY FIGURE OF ARMORED GUARDIAN. FIELD MUSE RAL HISTOv IXJ 3TAJ9 .(oo J 'to snjsF 1 7 gni ^b gaiDsctariq 9rlJ ai J MARBLE RELIEF OF GUARDIAN OF THE WORLD. PLATE LXl. GUARDIAN OF THE WORLD (see p. 300). One of the four Lokapala or Guardians of the World of Hindu mythology, who hold sway at the foot of the World-Mountain Sumeru. This is King Virupaksha residing on the western side of the mountain, holding in his left hand a miniature pagoda, and seizing a sword with his right. Here inserted to illustrate the identity of sheet armor in Buddhist stone sculpture with that in the preceding clay figures of the same epoch. Relief marble plaque, obtained from the temple King-ch'eng-se at Si-ngan, Shen-si. T'ang period (618-906). Height, 38 cm; width, 21 cm. Cat. No. 121555. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY. VOL. XIII. PL. LXI. *r Q A MARBLE RELIEF OF GUARDIAN OF THE WORLD. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLC IIXJ 5TAJ9 aajaaAS iq .uorfo ; GLAZED CLAY FIGURE 01 PLATE LXII. CLAY FIGURE OF SADDLED HORSE (see p. 313). Coated with soft lead glazes in three colors, a deep brown, a light yellow, and a plant green. Saddle-cloth and saddle are represented, the latter being padded with a gracefully draped textile material. Excavated in Lung chou, prefecture of Feng-siang, province of Shen-si. T'ang period (618-906). Height, 27.5 cm. Cat. No. 118039. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII. PL. LXII. GLAZED CLAY FIGURE OF HORSE. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. LXIII. jlnini 9vt33m srfT .anoisnamib ati ni aisoxa bablirom 21 li^f sriT .bsliabom Ilaw yhlBi SIB ,a)oorf bohoq gnr, M .^sO .coo CLAY FIGURE OF HORSE, FR< PLATE LXlll. CLAY FIGURE OF HORSE (see p. 313). This unglazed figure excels in its dimensions. The massive trunk and chest of the horse, its feet and hoofs, are fairly well modelled. The tail is moulded separately and stuck in. From Shen-si. T'ang period (618-906). Height, 52.8 cm. Cat. No. 118036. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. LXIII. CLAY FIGURE OF HORSE, FROM SHEN-SI. FIELD v NATURAL HISTORY. ANTHROPOLOGY. -VIXJ 3TAJR .q 39B) aesroH TO srIT . tj amtr^ aeiod si SIB slbi . . ri Job-alb: ) bt)h9q gnfi'T .aei PLATE LXIV. CLAY FIGURE OF HORSE (see p. 313). This horse turns its head sideways. The muscles are brought out in its head. Headstall, saddle-cloth, and padded saddle are represented. From Ho-nan. T'ang period (618-906). Height, 27.7 cm. Cat. No. 118038. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. LXIV CLAY FIGURE OF SADDLED HORSE, FROM HO-NAN. >F NATURAL HlSTORl OPOt-OGV, VOL. XIII, PL. LXV. .VXJ 3TAJ9 aefloH TO Jrighqu bna oJ amsae lottoq arii ,fiijg ied'ffai. OM HO-NAN PLATE LXV. CLAY FIGURE OF HORSE (see p. 313). Horse with complete harness and upright mane. The head is well modelled; and, though the pose is somewhat stiff, the potter seems to have attempted to repre- sent the animal as though mourning for its deceased master. From Ho-nan. T'ang period (618-906). Height, 30 cm. Cat. No. 118060. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. LXV. CLAY FIGURE OF SADDLED HORSE, FROM HO-NAN. PL. LXV PLATE LXVI. CLAY FIGURE OF HORSE (see p. 313). Horse with complete harness, mourning for its dead master. The trappings with their metal ornaments, the tinkling bells on the breastband, as well as the de- signs of lotuses on the crupper, are neatly moulded in relief. From Ho- nan. T'ang period (618-906). Height, 32 cm. Cat. No. 118037. FIELD MUSEUM Of NATURAL HISTORY. ANTHROPOLOGY. VOL. XIII, PL. LXVI. CLAY FIGURE OF SADDLED HORSE. FROM HO-NAN. ANTHROPOLOGY, VOL, XIII, PL LXVII. fans B ,00-ng-Jn, qz aril m bauoi bohsq gne'T GLAZED CLAY FIGURE OF HORSE. FROM ( PLATE LXVll. CLAY FIGURE OF HORSE (see p. 313). Fragmentary figure of horse, of unusual dimensions, and coated with lead glazes of light-yellow, plant-green, and brown tints. From Ho-nan; found in the spring of 1910 during the cuttings for a railroad north of the city of Ho-nan fu. T'ang period (618-906). Height, 80 cm. Cat. No. 118040. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL, XIII, PL. LXVII. GLAZED CLAY FIGURE OF HORSE, FROM HO-NAN. ANTHROPOLOGY, VOL. XIII, PL. LXVIII. IIIVXJ ? io d*tttni 9fw 10 ^*ic A adi lo bcsdsiol ad^ no lied ariT .nMoo fxjnaq goe'T .ia-aariS cnoii oVl .ieO .mo SMEN- PLATE LXVlll. CAVALIER (see p. 314). Horseman, escort of the inmate of the grave. Such figures were placed in front of, or behind, the coffin. The hair on the forehead of the horse is parted and combed toward the sides. Clay figure from Shen-si. T'ang period (618-906). Height, 33 cm. Cat. No. 118049. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. LXVIII. CLAY FIGURE OF CAVALIER, FROM SHEN-SI. u o 1 i M IP s PLATE en 4 j C W " J rt "M M 1 T3 ^ O 2 ^0 B e c 1 o u w: 3 ^ 1 B a ^ woman llowish- a 6 ffi E "rt O "o C 3 H "S CJ n r ~* * " tn >, i; ^ f^ O o cu s 00 0) at u O c ^ 4J 'J^ ^j* ^ T X X -I a -5 jj $ ^ S Ji i ra ,C C -^"^ LJ v._i- w -u \o 5 W ^ G jT CL HH -4~* i- 1 OO & < ^^1^? bl "U r3 c ^^ 5 1 1 1 2 .8 X S- li a FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, \. -IXXJ .Q 392) VIAMOW rrf rfJrw ieco (dOQ-8id) bohaq anfi'T .nen-oH moi .8?o8u .oM .^fiO .mo CLAY FIGURE OF EQUESTRIAN WOMAN. PLATE LXXl. HORSEWOMAN (see p. 315). Wearing male attire, a girdled coat with triangular lapels, trousers, and boots. The saddle-cloth is formed by a panther-skin. Clay figure from Ho-nan. T'ang period (618-906). Height, 30.2 cm. Cat. No. 118058. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. LXXI. CLAY FIGURE OF EQUESTRIAN WOMAN. M OF NATURAL HISTORY. ANTHROPOLOGY, nwob .stoqz HXXJ 3TAJ 1 i -q 993) rfoirlw mo'd qco tsfi bei l?-8id) boheq gae'T. .vSO^fi 10 srtt lo sa-oH moil .^sO ,mo CLAY FIGURE OF EQUESTRIAN WOMAN. PLATE LXXll. HORSEWOMAN (see p. 315). In brownish-red dress, with flat cap from which a long ribbon is floating down her back. The neck of the horse is painted red and overstrewn with white spots. The muscles of the head, the nostrils, jaws, teeth, and tongue are carefully modelled. Clay figure from Ho-nan. T'ang period (618-906). Height, 36 cm. Cat. No. 118057. FIELD MUSEUM OF NATURAL HISTORY. ANTHROPOLOGY, VOL. XIII, PL. LXXII. CLAY FIGURE OF EQUESTRIAN WOMAN. YORK cn GETTY CENTER LIBRARY 3 3125 00965 8325