Slide #262

TITLE: Asian Religious Mappaemundi
DATE: various
Only one major component of the overall picture remains to be described, namely, the existence of a tradition of religious cosmography in East Asia. This was not recognized by former sinologists but has been established in recent researches, especially by Professor Hirosi Nakamura. Essentially, this type of "world" map is centred on the legendary mountain in Central Asia or the north of Tibet, Mt Khun-Lun (a.k.a. Mt Meru, Mt Sineru); west of which are unknown regions, while to the east Korea, China, Indo-China and India form a series of promontory-continents extending into the eastern oceans. Around these outer seas there is a ring-continent, itself surrounded by further ocean. They are particularly common in Korea, where they occur as woodcuts, manuscripts and paintings on screens. Though they undoubtedly represent archaic tradition, paradoxically most of them are of late date (17th and 18th centuries). Nakamura showed, however, that many legendary names of countries taken from the Shan Hai Ching appear on them (110 out of 145), while others come from the Yü Kung chapter of the Shu Ching, the Mu Thien Tzu Chuan, Lieh Tzu, etc. Often there are annotations in the corners copied straight out of Huai Nan Tzu. No place-name later than the11th century has been seen on any of them.

There can be no reason for doubting that the Koreans received this tradition from China, though it seems never to have been so popular there. Herrmann reproduces a Chinese map of exactly this type dating from 1607 by Jên-ch'ao. There is another in the Thu Shu Pien [On Maps and Books] of 1562 by Chang Huang, entitled Ssu Hai Hua I Tsung T'u [Complete Map of China and the Barbarian Lands within the Four Seas]. This again has Mt Khun-Lun in the center. It also includes Ming place-names, however, as well as earlier ones which come mostly from the Ta Than Hsi Yu Chi of Hsüan-Chuang. The compiler says:

This map is copied from a Buddhist work. It represents Jambu-dvipa within the four oceans of the universe. . . Although this map, taken from the Fo-tsu-t'ung-ki, does not clearly represent the shape of the world, I give it here. Such Buddhist attempts are not as a rule convincing.

This note is of much interest, since it indicates the Buddhist-Taoist nature of the tradition, and the low value in which it was held by scholars who knew the scientific tradition of P'ei Hsiu. The Fo-tsu-t'ung-ki [Records of the Lineage of Buddha and the Patriarchs] still exists in the Tripitaka; it is due to a monk Chih-Phan and dates from 1270. Its cosmographical content seems to go back, through 7th century intermediates of which only the names are known, to the Ssu Hai Pai Chhüan Shui Yuan Chi [Record of the Sources of the Four Seas, and the Hundred Rivers] with maps, prepared by another monk, Tao-An, in 347. It is just about this time that we have a text which is perhaps the principal literary reference to this wheel-map tradition. It occurs in the Han Wu Ti Nei Chuan [Inside Story of the Emperor Han Wu Ti], a Taoist work of the Chin period attributed to Ko Hung. The emperor is on a visit to the legendary goddess of the West, Hsi Wang Mu.

On this occasion the emperor also saw certain baskets in which there were small books bound in purple silk. He asked whether these contained the techniques of the holy immortals, and if he could be allowed to glance at them. Hsi Wang Mu took them out, and showed one of them to him, saying, ' This is the Wu Yo Chen Hsing T'u [Map of the True Topography of the Five Sacred Mountains]. Only yesterday all the immortals from Chhing-Chhêng Mountain came to me asking for it, so I shall have to give it to them. It was made by San-Thien Thai-Shang Tao-Chun [The Highly Exalted Taoist Adept of the Three Heavens, the second person of the Taoist trinity] and it is extremely secret and important. How could so gross a person as you be fit to carry it ? I will bestow upon you instead the Mysteriously Shining Scripture of Life [Ling Kuang Sêng Ching] so that you may communicate with the spirits and strengthen your will [to immortality].'

But the emperor, prostrating himself, insisted on having the map. Hsi Wang Mu then said: ' Formerly, in the first year of the Shang-Huang Chhing-Hsu reign-period, San-Thien Thai-Shang Tao-Chün went down to see the wide world, and investigated the differences of lengths and breadths among the rivers and seas. He also observed the differences of height among the hills and mountains. Then he established the (position of the) Pillar of Heaven [Thien Chu]* and arranged the geographical features in their positions all around it. Then he placed (on the map) the five mountains (in a manner) imitated from the method of concentric zones (i.e. taking one as central and disposing the others symmetrically). The position of honor (the central position) was given to Mt Khun-Lun, where the dwellings of the immortals were depicted. An important position was also given to the island mountain of Phêng-Lai where the halls of the spirits were shown. The Water of An [the moon ?] was written in as very holy and sublime, since it is the origin of the extreme Yin influence. The Thai Ti [Great Emperor, i.e. the sun ?] was similarly placed in the wilderness where the Fu-Sang tree is. A little hill no bigger than ten feet square can be the place where human fates are decided. A small island in the stormy sea has enough room for the Nine Sages. Each continent has its proper name... AII of them are located in an orderly marmer in the huge environing ocean with its tremendous waves. The rivers are also seen flowing (on the map), some green and some black. The assembly of spirits is depicted cradled on the waves; all the immortals and jade girls are gathered there. Some of their names are hardly known, but their real features are distinctly displayed. Thus by the use of the compass and the square the rivers and their upper reaches were measured, and the mountains drawn with circularly curving lines. The mountain ranges bend back upon themselves and the smaller hills wander back and forth. The height of the mountains and the extent of their slopes are shown by lines turning and curving. Indeed, they look like written characters. Thus the written names of the mountains were determined by their respective natural shapes, and the reality of the mountains is enshrined in symbols. The diagrams of the shapes (of the mountains) were kept secret in Yuan-Thai, but nhen they were taken out they served as charms and talismans among the immortals. By their aid,Taoists can pass safely over all mountains and rivers. The hundred gods and the assembly of unmortals showed the greatest respect for this map and acclaimed it. Although you are not their equal, you have often visited mountains and marshes, you have the heart of an eager seeker after truth, and you have not lost sight of the Tao. I rejoice that this is so, and now therefore I will give you the map. Keep it, I enjoin you, with as much care as men show to their lords and their parents. For if the secrets of it should leak out to ordinary people, evil fortune will overtake you.'

* Another 4th century A.D. work, the Shen I Ching [Book of the Spiritual and the Strange] says (para. 9, Chung Huang Ching): 'On Mt Khun-Lun there is a bronze pillar which goes high up to the heavens. This is calledThien Ch. It has a circumference of 3000 li, as round and smooth as if cut with a knife. At the base of it are the dwellings of the immortals.' Again, the Shui Ching opens with the statement that Mt Khun-Lun is orbocentric.

There can thus be no doubt that we are dealing with a mixed Buddhist-Taoist tradition of religious cosmography, analogous to that of Europe, but with Mt Khun-Lun as orbocentric instead of Jerusalem.

The cosmography of Tsou Yen in the 4th century B.C. had an environing ocean, though not a central mountain. So also did that of the Kai Thien astronomical theorists. When Yung Hêng, therefore, in the Sung, says in his Sou Tshai I Wên Lu [Collection of Strange Things Heard] that 'in the east, north and south there are seas with different names, but actually it is all one sea', he was not necessarily influenced by the wheel-map tradition, though this may be a reference to it. However, the Ho Thu Wei Kua Ti Hsiang (one of the apocryphal or weft classics) says that Mt Khun-Lun is in the center of the earth, corresponding to heaven, and that the eighty regions are scattered all around it. China is in the southeast and occupies only one of these regions. If this was written towards the end of the later Han, one can perhaps see in it the combination of the indigenous world-picture deriving from Tsou Yen with that which came from India.

Obviously everything points to an Indian origin of these wheel-maps; and perhaps of all wheel-maps have a Babylonian origin. The current research work on the cosmography of the Indians shows that in ancient Buddhist and Jain ideas there were four continents centering upon Mt Meru:Uttarakuru in the north, Purvavideha in the east, Jambu-dvipa in the south, and Aparagoyana in the west. Brahmanic tradition also has the ring-continent, modified by what were perhaps echoes of Greek 'climates'. The classical literary exposition of this cosmography is that of the Abhidharma--kosa probably composed about 370 and crystallized in stone by the Khmer people, builders of the (9th century) Phnom Bakhen between Angkor Vat and Angkor Thom, and still one of the wonders of the world. One can hardly doubt that this cosmographic tradition came into China with Buddhism, perhaps joining earlier indigenous conceptions of Mt Khun-Lun as central. But never did it triumph over the tradition of scientific cartography there, as its analogue did in Europe.

Our habitable world, according to the Buddhist cosmographical view, is a continent called Jambu-dvipa lying to the south of Mt. Meru (Mt Khun-Lun or Sineru), which is commonly identified either with the Pamir Mountains of Central Asia or with the sacred Mount Kailasa (Kailas) in Tebet. Jambu-dvipa, or "southern continent", is derived from Sanskrit and was a term used to describe India, and by extension, the entire world, in ancient Indian cosmology. The continent is said to be wide in the north and pointed in south, having in its center Lake Anavatapta from which flow four large rivers. This is to be considered as a topographical reflection of the Indian peninsula, Lake Anavatapta symbolizing Lake Manasarovar in the Himalayas and the four rivers representing the Ganges, the Indus, the Oxus and the Tarim. Together with Buddhism, this geographical conception of the world was introduced into China and Korea and thence into Japan, where it found expression in various types of Buddhist maps. These maps, therefore, have some characteristics analogous to the Christian world maps of the Middle Ages, for they both express their religious view of the world rather than a scientific, empirical view of the earth. But while the latter gradually went out of use in Europe from the beginning of the Renaissance, the former claimed, with popular support, a right to existence until the middle of the l9th century, in opposition to the newer world maps of European origin which had been gradually gaining ground from the 16th century. This may be contributed largely to the extraordinary circumstances of Japan during her national isolation from 1639 until 1854.

KOREA: In old Korea, geographical atlases in the form of ordinary books, stitched with leather, were comparatively rare, but, in another form, mounted on screens, were very much in everyday use. Such an atlas was certainly very convenient, yet it was not commonly found among either the Chinese or the Japanese.

The most surprising thing about these old Korean atlases is that, except for very modern examples of the 19th century which show how very completely foreign knowledge had at last invaded the country, they invariably contain the same material, as if they were indifferent to the progress of geographical knowledge. They usually include the following maps, and in the order given: mappamundi [world map], China, Japan, Liu-Kiu, Korea and the eight provinces of Korea (one map to each province). Occasionally, however, the order is slightly modified, Japan and the Liu-Kiu, or the mappamundi, being put at the end of the atlas. These atlases have survived from ancient times, either as wood-cuts or as manuscripts. They reveal very great differences in size, in technical skill displayed, in color and even in title, but none in their basic content.

To begin with one is struck by the quaintness of the mappamundi at the beginning of the volume. As illustrated here, this style of mappamundi could be described as the medallion portrait of the right profile of a man, with carefully dressed hair and with his mouth open. The head forms the "central world continent", in which the Celestial Empire occupies almost the middle of the face. The front, or eastern part, of the coiffure represents Korea, the chin and neck, to the south, Annam and India; the back or western part of the coiffure the Western countries. This central continent is surrounded by a great ocean, in which there are numerous islands. Outside the ocean is a second continent in the form of a ring, which, in its turn, is surrounded by an immense and "endless ocean". In the latter are two islands, one to the east and the other to the west. In some manuscripts there is a third island to the south. According to Hulbert, the map is considered "complete", that is, nothing is conceded as being unknown, it was considered unwise to allow the public to suspect that the cartographer did not know of the existence of any land. Reproduced on the following page from the British Museum manuscript Atlas is an example of the more usual form that this type of map takes, although, as will be shown throughout this monograph, there were slight variations in the renderings of these world maps.

According to Hulbert, the central continent represents Asia alone, India being at the southwest corner. The only rivers definitely given are the Yellow River (#122), the Yang-tse (#129), and the Mekong (#130), all of which are represented fairly well. The general course of these rivers is only approximated here. It is interesting to note that China proper is represented as lying between the Yellow River and the Yang-tse, which, as history proves, was the real birthplace of the Chinese people. Korea (#119) and the Yellow Sea are also well represented, but when we go farther afield than this we find confusion. Annam (#132) is not far out of the way; but the land to the west of it, which must represent Siam [Thailand] and Burma, is highly fanciful. The water to the east of India must be the Bay of Bengal, and the two forks at the head of it may represent the Irrawaddi and Salween Rivers. If this is so, the Kwen-lun Mountains (#123) are not far enough out of their proper place to call for severe criticism. Using a liberal imagination, the unnamed lake to the north of these mountains may represent the Caspian Sea. The name of the map is placed at the two upper corners (#146 & #147) and means Under the Heavens and The Whole Map. Around the edge are arranged some statements of a general nature, namely, that the distance around the world is some 30 million miles; that the world contains 84,000 different countries; that it is surrounded by a limitless ocean; that the stars vary in width from 12 to 30 miles; that the sun and moon are 900 miles wide; and that the heavens and the earth are separated by a distance of 120,000,600 miles [English].

In the far north Hulbert notes the that the area labeled as the Covered Lake (#4) might well refer to the great ice-covered Arctic Ocean; and the Land Without Sunshine (#6) is not a bad description of the sub-arctic winter. The Land Where Political Divisions Were First Made (#8) demonstrates the ego-centric view of the Chinese cartographer.

One of the most important of these names is that of Pusang (#17), to the extreme right of the map. In Korean and Chinese legend this land lay 70,000 li, [or 21,000 miles] to the east of China. In that country grew enormous trees, 400 feet in height. Some scholars think that this area refers to America, which, of course matches the description and location. The distance is exaggerated, but the fact that it lies far to the east, and that it grows such phenomenal trees, would indicate that the land mentionwed is the Pacific coast of America.; but it should be noted also there there are two other places named Pusang as well (#67 & #76), so that it appears that the Chinese were somewhat unsure of its location (or the interpretation of the term Pusang needs to be clarified). Hulbert continues the interpretation of place-names on this map by noting somewhat north of Pusang lies Heaven Balance Mountains (#10), which reminds him of the tale of Atlas. He also speculates that the Land of Superior Men (#22) could refer to the Aztec civilization, which was experiencing its zenith at the time this map was made. In the Land of Women (#24) Hulbert finds a counterpart to the Amazons of Western mythology. To the south there is the Land Where People Do Not Die (#33), evidently Sheol, Happy Hunting-grounds, Valhalla of the West. The Land Where People Have Animal's Heads (#35) and the Land of Giants (#9) are a Brobdingnagian conceit. In the far west we find the Cloud-governed Land (#46) which may be a reference to the British Isles.

Passing to the northeast, again we find the Land of Hairy People (#62) which may be a refernce to the Ainus. If #64 represents the southern island of Japan, Kyshu, and #68 the middle island, Honshu, then #62 might easily represent the northern island of Japan, Hokkaido, where the Ainus live. The land numbered #65 is right beside Japan, and is called Land of Asking Trousers by Hulbert, an example of the often "forced translations". Glass Bead Land, or Loo-choo Islands (#72) seems to lie too far away from Japan; and no determination has been made as to the identification of the large island containing #68, #69, and #70, which seems to lie partly between Japan and Loo-choo. Siam [Thialand] (#74) is mistakenly made out to be an island, and is placed north of Annam. The Land Where Wood is Eaten (#82) may refer to the fact that so many of the people of Polynesia live on fruits almost exclusively. The Land of Fire-haters (#89) lying near the Equator is well-named, though it is hard to recognize it as any specific place. Passing toward the west, we find the four contiguous peoples called respectively, Extraordinary Land, Land of Good Agriculture, Land of Musical People, and Land of Vain People. These are located approximately where Europe ought to be, given the relative scheme of the map. Norway might be the area called the Deep Sun Land (#106), as the sun does not rise high in the heavens even in summer.

In reality it is nearly impossible today to accurately identify more than a few of the place-names on these mappaemundi. All of these people and places, however, are mentioned in one place or another in Chinese and Korean folklore and literature. When seen for the first time these Korean atlases appear to be very old, for the dominating idea in the mappaemundi is comparable to the mediaeval cosmographical conceptions of Cosmas Indicopleustes (Slide #202) and the imago mundi. The very beautiful, solid paper, sometimes waxed, upon which they are drawn is like old parchment. When badly preserved, as so often happens, they appear still more ancient.

In 1895 Professor Maurice Courant, of Lyons, reproduced one of these 18th century manuscript Korean mappaemundi in the second volume of his Korean Bibliography, though without giving any explanation of its origin or any description of it. In the following year M. Henri Cordier published a faithful and costly heliogravure reproduction of an 18th century manuscript atlas found in the Map Room in the British Museum. It is an atlas of the Kingdom of Korea and its provinces, with some other maps, the whole in Chinese characters, bought by the Museum in the previous year from an American traveller. Cordier has, in his collection, transcribed these Chinese characters, but he does not speak of the origin, source or date of this mappamundi, which he believes to be certainly of very great age. In this he was probably influenced by the opinion of Yi Ik-Seup, who says in his memoirs that the Korean mappaemundi existed from time immemorial. Yi Ik-Seup also says that, after long and patient search, he succeeded in finding two old Korean manuscripts, the one of purely Korean origin and the other from a Chinese source, containing Europe, Asia, Africa and even America. The former is referred to above and the latter, about which he makes no definite statement is possibly one of the maps made by the Jesuits in China and very badly copied by the Koreans (an example of this kind of map, Ye-chi-zien-do [Ancient World], was published in: "Chosön, the land of the morning calm. A sketch of Korea" by Percival Lowell, Boston 1886. The heliogravure reproduction of the map is entitled: "Korean map of the world, reduced to about one-fifth the area of the original. In the original the mountains are a vivid green and the sea a lilac"). Reproductions by Yi Ik-Seup, Courant and Cordier are all of manuscript maps, and no reproduction of a Korean woodcut mappamundi was made until that by H. B. Hulbert in 1904. In 1905 Carlo Rosetti also reproduced a manuscript mappamundi, translating its topographic names in the same way as previous editors, without giving any further description or information about them.

Caution is warranted in reading the works of Yi and Hulbert for they attempt to identify topographical names by the literal and often forced translation of the Chinese characters with names in use today (see examples on previous pages). When they did not know how to properly interpret them, they simply transcribed the pronunciation. The literal translation of proper names into place-names has, of course, no meaning, for every Chinese character can have, at the same time, many significations and, moreover, Chinese characters are symbolic as well as phonetic. Still other difficulties are encountered in place-names, for they change from era to era and from one chronicler to another. Yi says in one place that the mappamundi dates from time immemorial, and in another he is trying to identify America in the "ring" continent! Hulbert does not go so far, but, tired by a long and singularly forced interpretation, he says, "Human nature is the same all the world over. The conceptions shown in the Chinese and Korean mappaemundi are evidence of this, for they differ only in degree from the imago mundi of the Middle Ages." The Korean mappamundi has no value whatever from the geographical point of view, though it has a certain interest from that of folklore, for all place-names recorded in one or another example, and found here and there in Chinese literature, will one day be identified.

Before discussing the Korean mappaemundi from the point of view of ancient cartography, existing materials and their dates must be examined. There are two groups of materials, woodcuts and manuscripts. The latter, though incomparably more numerous than the former, are no more authentic, rather the contrary. Circumstances in Korea and Europe were not the same. In China and Korea prints from wood blocks were costly and limited to a small number, therefore printed books were often copied by hand. While this procedure often gave rise to unexpected mistakes, it was a simple and primitive method of reproduction which was practised until quite recently in the Far East. Remembering this we must focus our attention upon the printed atlases which, without exception, bear no date at all or only that of the copy. For this reason the examination of different editions is of supreme importance. Twelve different editions will be described here which were examined by Hirosi Nakamura.

I. (a) Tchien-ha-tchong-do [Mappamundi], anonymous (this map is anonymous, but bears a pseudonym
"Bu-un-muk-kaik," meaning "Penman Floating Cloud," often used to denote a Buddhist monk, one of
whom must therefore have been the copyist. It is little different from the others in the explanatory notice
above the drawing. This seems to me to be made at random and without bibliographical authority.
The third continent, outside the Great Desert, (below the main drawing), and its nine place-names, are
worth no more than this reference.), no place of publication, undated.
(b) Cho-sen-chi-do [Maps of Korea], anonymous, no place of publication, undated.
Library of the Imperial University of Keijo
(c) Without title, anonymous, no place of publication, undated.
(d) Tchien-ha-chi-do [Mappamundi], anonymous no place of publication, undated. All these atlases
are of the same edition, with no definite title. They are large quarto volumes, about 32 X 43 cm.,
containing ten maps, as follows: Mappamundi, China and the eight provinces of Korea. The
mappamundi measures 43 X 63 cm., with a printed surface of 42.5 X 50.5 cm. The sea and the
rivers are engraved in black and the land in white. The technique of the wood-engraving is primitive
but attractive. The last two of the four Chinese characters showing the title are enclosed in a rectangle.

II. (a) Without title, anonymous, no place of publication, undated.
Atlas, the Mappamundi of which is reproduced by H. B. Hulbert.
(b) Cho-sen-chi-do [Maps of Korea], anonymous no place of publication, undated.
Library of the Imperial University of Keijo, Ancient 4709/38.
This Atlas contains 13 maps, as follows: Mappamundi, China, Japan, Liu-Kiu, Korea and its eight provinces.
The wood engraving is like that of the preceding examples. Format, large octavo, 13.5 X 33 cm.
The mappamundi measures - paper, 37 X 33 cm., printed surface, 31.5 X 31.5 cm.. The four characters
showing the title are enclosed two by two in rectangles. In Mr. Leo Bagrow's collection there is a MS. copy
of this atlas entitled Tchien-ha-tchang-do [mappamundi].

III. (a) Tchien-ha-tchong-do, or Tchien-ha-do [Mappamundi], anonymous, no place of publication,
undated. Ecole des Langues Orientales Vivantes, Paris, D. 7989.
(b) Chio-do-bo [Map Treasures], anonymous, no place of publication, undated.
Mr. Hidetaka Nakamura's collection.
(c) Without title, anon., no place of publication, undated.

Contains 13 maps: Mappamundi, China, Japan, Liu-Kiu, Korea and its eight provinces. Engraved in a
fashion rather different from that of preceding atlases. Sea-coast and river-banks shown by fine
lines. Mappamundi bears title Tchien-ha-do [Mappamundi], in top right-hand corner of map and
measures - paper, 38 X 35 cm., engraved surface 31.7 X 36.5 cm. In the copy in L'Ecole des
Langues Orientales there is no map of the province of Keishodo.

IV. Without title, anonymous, no place of publication, undated.
Contains same 13 maps as III, which it greatly resembles, but is easily recognized as a separate edition
by, for example, the shape of the trees in the mappamundi.

V. Ye-chi-do [Geographical Atlas], by Kum-hos-an-in (Ai-kieng-chai), no place of publication,
zodical date, year of the old earth and of the cock.
Library of Baron Mitsui.
Library of the Imperial University of Keiyo, Ancient 4709/58.

VI. (a) Ye-chi-do [Geographical Atlas], by Kum-ho-san-in (Lo-on-gun at Honan), zodical date, year of the
old earth and of the cock. Professor T. Okudaera's collection.
(b) Tong-kuk-ye-chi-do [Korean Atlas, title printed on cover], by Kum-ho-san-in (Lo-on-gun at
Honan), zodical date, year of the old earth and of the cock.

VII. Chi-do-pien [Geographical Atlas, MS. title], by Lo-on-gun at Honan, zodical date, year of the
old earth and of the cock.

These three atlases, Nos. V, VI and VII, are different editions of the same original. They contain 13 maps: Mappamundi, China, Korea, Liu-Kiu, Japan and the eight Korean provinces. The technique of the engraving is like II above, but touch and design are quite different. Format, largo octavo, 20 X 31 cm., printed surface of mappamundi, 32.5 X 27 cm., bearing title Tchien-ha-do [mappamundi].

VIII. Chi-do [Geographical Maps], anonymous, no place of publication, undated.
This resembles No. IV, but is of quite a different edition. The map of Japan, for example, in
No. IV bears no legend, but this one does, taken from the Hai-dong-tsu-kuk-ki, edition of
1471. Format, large octavo, 18 X 30 cm. The mappamundi measures, paper, c. 37 X 30 cm.,
engraved surface, 32 X 27 cm.

IX. Chi-kwal [Summary of Geography], anonymous, no place of publication, undated.
Very much like the preceding, but clearly of a different edition. Shows prefectures, cantons,
fortresses, posting houses, military posts and so on, in legend at the top of the map of each
of the eight provinces (wanting in No. VIII). Differences also in design. Format, large octavo,
20 X 34.5 cm. Mappamundi measures, paper 39 X 34.5 cm., engraved surface, 32 X 27 cm.
Contains 13 maps, Mappamundi, China, Korea and its eight provinces, Liu-Kiu and Japan.

X Ye-chi-do [Geographical Atlas], anonymous, no place of publication, undated.
Library of the Imperial University of Keijo, Ancient 4709/37.
Contains only 10 maps: Mappamundi, China and the eight provinces of Korea. Japan, Liu-Kiu
and the general map of Korea missing. Except China, which is clearly different, the maps are
the same as those in No. IX.

XI. Chi-do [Geographical Maps] anonymous, no place of publication, undated.
Very like Nos. II and V in general appearance. Sea and rivers engraved in black, land in
white. Design is coarse. Format, large Mappamundi, China, Japan; Liu-Kiu, Korea and its
eight provinces.
mappamundi measures - paper 40 X33 cm, engraved surface 35.5 X 30 cm., bearing title Tchien-ha-do.
Contents archaic, yet it shows latitudes and longitudes, an extraordinary anachronism due to the
influence of the maps drawn by the Jesuits in China.

XII. Without title, anonymous, no place of publication, undated.
Contains only mappamundi. Paper, 38.5 X 35 cm, engraved surface elliptical in form major axis,
31.5, minor, 28.5 cm. The paper is folded in 4 X 7 cm. On the verso and in the center are two small
pieces of thick paper decorated with vignettes in low relief, forming the cover.
The mappamundi itself does not differ from others in essential features, but it shows the sea in wave-form,
and has many more place-names than usual in its western parts.

It is difficult to understand the adoption of so small an atlas format for such a mappamundi has no practical advantage. Nevertheless manuscript maps thus folded are often found, and this one has therefore no special significance; it is simply a question of habit and taste. Hulbert believes that two or three other different editions are in existence, but cannot be described here before having more carefully examined them. As to the manuscript atlases, they are so numerous that one can not venture to describe them all. However, it must be emphasized that the printed examples are more authentic than the manuscript, considering that the latter are invariably copied from the former.

Manuscript copies are of all sizes, according to the whim of the copyist. Ordinarily they were needed when there was a great expanse of paper had to be covered with decorative colored material, with writing or lettering to hide the empty spaces. They are mostly copies of one or other of the atlases listed above. Nevertheless we meet from time to time a particular type which does not occur in a printed form. It is not mounted on a screen but is stitched like an ordinary book. Its maps are small, drawn with great care and pleasantly colored. This is a six-volume atlas, with, as usual, no special title containing very many maps.

I. Ye-chi-do [Geographical Atlas], anonymous, no place of compilation, undated.
Library of His Highness Prince Yi, Historical Division, Geographical Section, No. 42.
II. Ye-do [Geographical Atlas], anonymous, no place of compilation, undated.
Museum of the Central Government of Korea.
III. Chi-do [Geographical Maps], anonymous, no place of compilation, undated.
Museum of the Central Government of Korea, Cho, 61/27

These are copies made one from the other. The contents of the mappamundi in this atlas, as well as the manuscript copies, large and small, are always the same. Unfortunately they are almost all anonymous and undated. Atlases I and II contain maps of different epochs; for example, the general map of Korea, which has probably been copied from the "Treatise on Korean Geography", compiled in 1481 by the King of Korea; that of Japan, copied from a very old map which must have been introduced into Korea in the 15th century; that of the Liu-Kiu, semi-imaginary, produced before the Yuen Dynasty; and that of China, according to some maps of the Ming Dynasty. These atlases are, in fact, a compilation of anachronistic maps. The maps of the Korean provinces, however, contain information more recent than that in the royal treatise. The names of certain garrison towns, for example, make it clear that this form of atlas was not originally compiled before the latter half of the 15th century.

At the end of the preface in Atlases V and VI is the date, "The year of the oldest earth and of the cock," but this does not enable us to fix the zodiacal cycle for that year. In the preface also the author calls China "Great Ming." Can we therefore find the year in question in the Ming Dynasty? Unfortunately not, for in the geographical description at the end of Atlas VI, a description not found in Atlas V, there is a passage about the thermal spring Onyo, which says, "His Majesty King Hyen-zong and King Shuk-zong visited it and the temporary royal villa is there also." The later king occupied the throne between 1675 and 1720, and the Ming Dynasty came to an end in 1661.

Atlas V bears the names of the compiler Kum-ho-san-in and Ai-kieng-chai, Atlas VI the name of Kum-ho-san-in and Lo-on-gun, while Atlas VII, with a preface modified and shortened from that of previous editions, bears only the name of Lo-on-gun, and the date no longer appears in the preface, but at the end of the volume, where we read "To Zo-lieng, newly edited in the year of the young earth and of the cock, in the fourth cycle after the foundation of the new era of Tchoung-tcheng" (1628) which corresponds to 1849 in the Christian era. Are we then, according to this, to consider Atlas V the oldest, VII the most recent, and Kum-ho-san-in the first compiler?

The explanatory notice on the map of Japan says that after the new treaty between the authorities and the chief of the island of Tusima, Japanese merchants were authorized to trade in Korea. This treaty between the clan So of Tusima and Korea having been concluded in 1609, it is possible that this work may have been compiled in that year, and if we look for a scribe with nom-de-plume Kum-ho- san-in we can name Liu-hieng-su, who had been condemned to death in 1547, accused of plotting against the royal family.

The notice on the map of the province of Kankyo-do in Atlases VIII and IX gives the names of the Kings Se-jong, Soung-jong and Jung-jong. The last-named occupied the throne from 1506 to 1544, which shows that these atlases are later than this date, and very probably before the end of the Ming Dynasty, 1661. All enquiries by Nakamura, therefore, lead him to give the date of the compilation of these printed atlases as not before the 16th century, while the dates of their successive editions will probably be still later.

It is curious that we do not meet this form of mappamundi in Japan, for it is incredible that a map form so widespread in Korea should not have been introduced there at all. The bibliography of the subject mentions at least one Japanese who alludes to a similar mappamundi. He is named Yosino Zingoemon, an officer in the army of the Daimyo of Matura. This man, in the introduction to his "Story of War", written on a war-vessel in the port of Fusan, at the time of the invasion of Korea by Hideyosi, reflects, while looking at a mappamundi, on the extent of the world. His story refers, apparently, to a mappamundi of the type Tchien-ha-tchong-do.

One example of this Korean form of mappamundi published in Japan (map measuring 31.5 X 23.5 cm/paper - 48.2 X 33 cm) was once in the collection of Mr. T. Magami and is now owned by Mr. H. Ikenaga. It is an undated print, made at Nagasaki by Kassai-do, and has a manuscript geographical description around the mappamundi. It is entitled Sekai San-koku-ki ["three kingdoms", that is "Description of the Whole World" - India, China and Japan, which passed for the whole world] and though much altered, is easily recognizable as having the essential configuration of the Korean mappamundi, by the shape of the central continent, the numerous ocean islands, and the second ring-shaped continent, which has here an indented, rectangular form. It is, in its content, a real catastrophe in geographical knowledge, for it contains every kind of anachronism. To the northeast of the central continent Yezo takes on the form of Kamchatka, embodying Japanese knowledge as it existed at the end of the 18th century; there is a sandbank, curiously drawn in double-stippled lines, a method often adopted in Japanese portolans of the 16th and 17th centuries; three islands in black in the north, found only in Japanese editions of the mappamundi of Father M. Ricci, and so on. This map has a special interest in that it shows the transition from the Tchien-ha-tchong-do type of mappamundi to the more modern type to be considered later.

With regards to the content of the mappamundi, taking as example Atlas III or IV, which are those now most often seen. Among the 145 names marked are 40 mountains, 5 rivers, 4 lakes, 3 trees, 2 large regions and 91 names of countries or peoples, counting separately five names, each of which is used twice in different localities. Hardly a quarter of them can be accurately identified in the central continent and the adjacent islands. The remainder are considered by scholars as fictitious and fantastic, most of them legendary names from the Chan-hai-king, one of the oldest books of Chinese geography. No less than 110 are from the canons Hai-wei, Hai-nei and Ta-fang, the most fabulous parts of that book. In his enumeration, Nakamura has introduced some perceived corrections into the reading of 12 mutilated names among the 110, probable mistakes presumably made by the copyist of the map. The few other mythical names come from such other works as the chapter on Yu-kong in the Chang-chou, the Mou-hien-tseu-tchouan, Lie-tseu, Hoi-nan-tseu, Hai-nei-chih-tcheou-ki and so on, all works of different periods before the Hans. Some of these names have appeared in one book after another ever since that time, others appear only in more recent works.

Both the historical and geographical value, and the date of the Chan-hai-king have been long discussed. Modern Chinese scholars are agreed that it was not the compilation of a single person, but that heterogeneous materials were added from time to time to the initial collection. The scholar Lacouperie believes that the oldest and most authentic part of it is the Canon of the Five Chan-king, put together in the Tcheou Dynasty from very old sources, and that the canons Hai-wei-king and Hai-nei-kin,compilations of the Tcheou Dynasty, were added by Lou-hsiang between 80 and 89 years before our era, and finally that the the canon Ta-houang was added by Liu-siu, who died in 57 A.D. Mr. T. Ogawa believes that the second addition was made during the Ts'in and Han Dynasties, but in any case before the chronicler Se-ma-t'sien.

The identifiable geographical names are found in official chronicles from the Han Dynasty to that of the Thang. Nine are in the Han-chou (1st century), three in the Soui-chou (7th century), one in the Thong-tien (9th century), and also in the Sin-thang-chou (11th century), one in the Thang-chou (10th century). The solitary 11th century name marks the upward limit of the date of the geographical names in this mappamundi . Thus its content is made up of materials of very mixed date, gathered together through a long period of time, from the second or third century before the Christian era up to the 11th century A.D.

Certain atlases, as, for example, VIII and IX, have a notice in the right-hand margin of the mappamundi. "The distance across the Universe from the eastern extremity to the western is 223,500 li and 72 paces, the distance of the two extremities, from east to west, is 28,000 li." Another note in the left-hand margin reads, "the distance from the northern extremity of the Universe to the southern is 203,500 li 75 paces, the distance of the two extremities, from north to south, is 26,000 li." This is a quotation from a chapter of the Hoai-nan-tseu [Lecture on Geography], which says "The distance across the earth, this side from east to west is 28,000 li, and from north to south 26,000 li . . . Yi made Ta-chang walk from east to west and reckoned the distance across as 233,500 li and 75 paces. He then made Chou-hai walk from north to south and reckoned the distance as 233,500 li and 75 paces." Thus in the Hoai-nan-tseu distances are equal, while in the mappamundi they are slightly different. This must be due to a copyist's error, for in a manuscript Atlas (Nakamura's collection, No. 416), there is a phrase taken from the Hoai-nan-tseu, which begins "Yi made Tac-hang walk." These figures are also found in the Chan-hai-king at the end of the Tchong-chang-king chapter. The copyist of this mappamundi therefore found materials not only in the Chan-hai-king but in other works also.

There is a group of Korean mappaemundi which differ somewhat from those of the atlases enumerated in terms of the number of place-names in the western regions. They are all, in other respects, exactly alike. Examples of this group are to be found in: Atlas XII, described here ; MS. Atlas, reproduced by H. Cordier; mappamundi reproduced by M. Courant; MS. Atlas, Nakamura's collection, No.416; MS. Atlas, Library of the Imperial University of Keijo, Ancient 4709/1, and others. These mappaemundi contain 51 names more in the western regions than do others, but four of these are common to both groups so that in reality there are only 47 more. Of the latter 14 come from the Chan-hai-king, the remainder from the Han-chou. Only one is found in both documents.

This increase in the number of place-names cannot indicate a later addition, seeing that none of them are earlier than the first century. They ought rather to be considered as purposely omitted in the one group, because of lack of space. The following phrases are, in fact, often found in the commentaries: "These are not the whole of the names of countries," "the rest of the countries," "innumerable countries", indicating omissions. That these are intentional is the more clear when we read such phrases as, "twelve countries in the regions to the south and west," "numerous countries in the west" and so on.

The topographical names on the Korean mappamundi, therefore, were current in widely separated epochs. This might tempt us to doubt the authenticity of the map, or to suspect that it was drawn by amateurs, comparatively recently, in accordance with ancient writings, and that therefore it cannot be considered otherwise than as a kind of historical map. But if it were the product of such a caprice there should be many inconsistencies among the place-names, at least in some among the many editions, and above all, in some manuscript atlases made at different periods by different hands. Not only, however, are there no such inconsistencies, but, on the contrary, the same mistakes are slavishly repeated. This fact makes it the more certain that the mappamundi is the traditional reproduction of the original, and not a modern creation.

NOTE: Yi or li today means 100 millions, but not in the classics. Toan Yu-tshai (1735-1815), in his Commentaries compiled under the Hans, says that yi means 10,000 in all the canonical books. If this is so, the second figure in 223,500 must be redundant. However, it is possible that the word meaning 20,000, in Chinese two-10,000, was put into the text in the course of repeated copying. This would appear to be logical if we compare the figures for the distance across the universe from east to west with those from north to south, 203,500, which would then read 23,500. In Buddhist books yi may mean either 100 millions or 100 thousand.

Since place-names of the period from the 6th to the 11th century are few and sporadic in these Korean mappaemundi they must be considered as later additions. The great body of names are taken from the Chan-hai-king and the Han-chou, but above all from the former. This assumption is made the more convincing by the fact that overseas countries [Hai-wei], and those of the Great Desert [Ta-houang] are distinguished by colors according to the orientation of the Chan-hai-king. The colors of the four cardinal points and the center are in accordance with the philosophical idea of the five elements: in the east, blue; the west, white; the north, black (a very dark indigo, almost black); the south, red; and in the center, the Celestial Empire, yellow. The idea of the five elements wood, fire, earth, metal and water, is a dominating one in China. All natural and moral phenomena are interpreted by its principles. An explanation of the connection between the five elements and colors is found in the Sou-oen.

Now the Chan-hai-king, with the exception of the first five chapters, originally had accompanying pictures, and the text existed only to explain them. The original pictures were lost in a far-distant age, and those we see today, representing fantastic monsters, were remade in the 6th century in accordance with the ancient text. Many have therefore believed that the Korean mappamundi, or another very like it, must have been part of the Chan-hai-king as an explanatory picture of its cosmography.

Any supposition is feasible, but without solid foundation it remains only supposition. This hypothesis, that the Korean mappamundi had its origin in the Chan-hai-ing, must be admitted only with the greatest reserve. It may be very old, but it need not be as old as that. It cannot have taken on its peculiar form, nor have been given its traditional content, before the 11th century. None of the research scholars allow us to fix its date earlier than the 16th century. Most of the examples are of the 17th and 18th centuries. On the other hand, the closest examination of its content will not allow scholars to put its date any later than the 11th century. To soften this contradiction appropriate material must be found to bridge the gap between these two epochs. Such material will come, it is to be hoped, from Chinese sources rather than Korean, for this mappamundi is purely Chinese. It bears no trace of anything specially Korean, which is understandable when we consider that the sciences and the arts of Korea were almost always slavishly modelled upon those of China. In the following section we shall consider a group of mappaemundi which link the two epochs mentioned in this paragraph.

In the Museum of the Central Korean Government there is an eight-leaved screen bearing the number 9734 made up of maps of the world, China, Japan, Korea and other regions. It contains two mappaemundi, one of the Tchien-ha-tchong-do type, described above, the other of quite a different kind, entitled Ssu-hai-hoa-i-tsung-t'u-chi- hyong [The Complete Map of China and Barbarian Lands within the Four Seas], which measures 44 X 43 cm. The fact that it was mounted in screen-form shows that this type also must have been well known in Korea.

It is obvious that this map bears a close relationship to both the Korean mappamundi and the Go Tenjiku Zu, though only the central portion of the latter is shown, i.e, the central continent, with the neighboring islands of the surrounding ocean. A part of the second ring-continent is just visible in the east and south. These maps should also be compared with the Sakei San-koku Ki, mentioned above, to appreciate the transition from one form to another.

This type of mappamundi, like the others, must be of Chinese origin. There is, in fact, one exactly like it in a well-known work, the Thou-chou-pien (26.5 X 20.7 cm), an illustrated encyclopedia compiled by Tchang Hoang (1527-1608), where it appears in the 27th volume, entitled Ssu-hai-hoa-i-tsung-t'u,the same title as that of the screen map, but without the last two characters, -chi-hyong, meaning "shape of." The two maps are exactly alike in form and content, though they differ in color and dimensions

There is a third example of the same kind in Leo Bagrow's collection, entitled Tchien-chi-nam-tchem-bu-chu- chi- heung, and forming the first part of a Korean atlas, Tchien-chi-nam-tchem-chi-do, containing 29 maps. It is a copy, not very faithful, of the Thou-chou-pien map, with changed title, and altered also by the addition of explanatory phrases in the margin and on the map of Japan, in the shape of the islands, especially those of the west, and in other ways Its very existence leads us to believe that this form of mappamundi was common enough in Korea.

As to the dates of these three specimens, the two Korean examples are undated, so that we know only that of the Thou-chou-pien map, whose compilation was begun in 1562 and finished in 1577. But the map itself is not of this period, in spite of the addition of 13 names of the provinces of the Ming Empire. The date of the encyclopedia is only the date of the reproduction of this mappamundi, for the compiler himself twice clearly says that he has copied it from a Buddhist work, once in a note above and to the right of the map: "This map is found in a Buddhist work. It represents Jambud-vipa [the habitable world] within the four oceans of the Universe," and again in the text: "Since the map of the book Fo-sou-t'ong-ki does not represent clearly the shape of the world, I show here the world according to a Buddhist work, although Buddhist attempts are not as a rule convincing." The original map is therefore much older than the Thou-chou-pien, but unfortunately the compiler nowhere gives any clue as to the source from which he obtained it.

The two Korean maps must have been copied from the Thou-chou-pien map, and, since most of the Korean maps met with today are productions of the 17th and 18th centuries, these also are very probably of the same period. The compiler of the encyclopedia quotes "the map of the western regions and of the five Indies," from the Fo-tsou-t'ong-ki [Records of the Lineage of Buddha and the Patriarchs], 1269, by Tche-p'an, itself doubtless taken from the Si-yü-ki, as is stated in the text, and as we can see from the place-names on the map. Chavannes states that "This map has no geographical value, and is little more than a list of the countries given by Hiuen tchoang in the Si-yü-ki. These three maps, although later than those reproduced here from the prints made by Pei-lin, are, however, earlier than that of the Mongolian epoch which Bretschneider (Mediaeval Researches, vol. 2, pp. 45) believed to be the oldest Chinese map surviving to our days. The last-named map came from the King-che-ta-tien, edited in 1329, not 1331, as Bretschneider wrongly states. The map of the King-che-tat-ien does not represent the cartography of the Mongolian epoch, as Bretschneider says, nor does that of the Tche-p'an, as suggested by Chavannes. There is a well-authenticated manuscript Chinese atlas of the Mongolian epoch, the Kuang Yü T'u, the work of Chu Ssu-Pen compiled by Lo Hung-hsien under the Mings, in the library of the Imperial Cabinet of Tokyo (Slide #227). It is quite different from the two mentioned above and, according to Nakamura, was misunderstood by both Bretschneider and Chavannes.

The content of mappaemundi of this group is very different from those of the Tchien-ha-tchong-do (Slides #262I/J) type. The names of the 13 provinces of the Ming Dynasty are given, but for China as well as other countries there are topographic names of all periods. Most of these names are from the Si-yü-ki of Hiuen- tchoang, 602-664, a famous Chinese Buddhist traveller who went to India in the first half of the 7th century. Naturally, in so small a map, the names of all the countries he described cannot be given, 110 visited by him personally and 28 others of which he had heard. There are, in the map, many omissions and mistakes, due to repeated transcription from one copy to another, which shows that it was not made directly from the text of the Si-yü-ki. A certain number of names were drawn from other sources. The following remark, for example, is made about an island in the southeastern ocean: "Ye-pho-ti [Java] . . Fa-hien was thrown upon this island by a storm," and there is another small oval island, named Ho-ling [Kalinga in Java), shown as a separate island, described in I-tsing's voyages (635-713). Most of the topographical names concerning the islands in the ocean surrounding the central continent, and those on the portion of the ring-continent of the Tchien-ha-tchong-do, are fabulous, taken from the Chan-hai-king . This map is therefore a hybrid, made up from the Tchien-ha-tchong-do and another, probably the work of a Buddhist and based upon the Si-yü-ki, a source rich in details and topographical descriptions of the western regions and the Indies.

The best known map of this kind is that of the Tenjiku koku [the Indies] inserted in the third volume of the Syûgaisyô (size of the original 34 X 25 cm.), an encyclopedia compiled about the middle of the 14th century, and handed down to posterity in a hand copy whose content has been augmented at different epochs. The oldest edition is of the Keicho period, 1596-1615, but it does not contain the map of the Indies, which must therefore have been added in a more recent edition, unless it always existed in certain other MS. examples. There is no real proof that it existed in the 14th century. Its central continent has almost the same shape as that of the Thou-chou-pien map, with smooth coastal lines forming a shield-shaped, almost geometrical figure Its content owes much to the Si-yü-ki, as is evident not only from the place-names but also because the draughtman continually quotes phrases from that work. Its author was not an inhabitant of the fabulous world of the Tchien-ha-tchong-do, seeing that he calls it a map of the Indies and not a mappamundi.

There is another interesting map of this kind in the Hosyoin temple in the Siba Park at Tokyo, reproduced as frontispiece in the second volume of the travels of Yu-ho Den, in the collection of Buddhist books, 1917, and entitled Sei-eki Zu [Map of the Western Regions]. It is accompanied by two documents, also reproduced in the book. They give the source of the map and collate and correct the inscriptions it contains.

According to these documents Hiuen-tchoang, while on his travels in the Indies, found the original of the map, and, after having marked on it in vermilion ink the route he had followed, he placed it in the T'sing-loung-tseu [Blue Dragon] Temple at Tch'ang-ngan, then the capital of China. Huei-kuo (746- 805), chief priest of the temple, presented it to Kukai (774-835), his best Japanese disciple, when he returned to his own country. He placed it in the Tozi Temple, which he founded at Kyoto. Much later a priest named Komatudani presented it to the 41st chief priest of the temple Zozyozi at Yedo. The Iatter, delighted with this wonderful Buddhist geographical treasure, and deeming it too rare and important to keep to himself, caused another copy of it to be made, for what he had received was only a rough sketch. Ninkai, Ryoseki and Dyogetu drew the map, while Ryogetu, Keigi and Zyunsin added the place-names, collating and correcting them according to the Si-yü-ki and other works. The copy was finished in 1736.

These editors made history by quoting many Chinese books on the title of the map, which was originally Go Tenjuku Zu [Map of the Five Indies"] (Slide #263) and which, in their opinion, was not correct, since it dealt not only with the Indies but with the western regions also. They called it, therefore, the Map of Western Regions.

This information is not easy to accept, since it often confuses the copy and the original, Moreover it is impossible that the map should have been of Indian origin, and it is doubtful that it was constructed by Hiuen-tchoang, seeing that it shows an area greater even than the Indies. According to Nakamura, it was drawn by someone else, after his return, to illustrate the story of his travels, and that the map given to Kukai by his master was a copy, and not the original. But whatever the authorship of the map there is no doubt that it was brought from China by a religious student, whether Kukai or another. To decide this question other more ancient and more authentic materials are necessary.

From the point of view of date alone the Hosyoin copy is not interesting, for we know of many other similar specimens much older. An example of Japanese world maps representing Buddhist cosmology can be seen in the earliest map of this type, the Nansenbushu Bankoku Shoka no Zu [Outline Map of All the Countries of the Jambu-dvipa] (Slide #263.4). Nansen Bushu is a Buddhist word derived from the Sanskrit, Jambu-dvipa, or the southern continent. In 1710 (Hoei 7th year) this large (1 m 45 cm X 1 m 15 cm) woodcut map, supplied with a detailed explanatory text in Chinese, was published at Kyoto by a priest named Rokashi Hotan, under the pseudonym of Zuda Rokwa Si (died in 1738 at age 85). It is not a faithful copy of the preceding, but has many corrections and additions. The most striking innovation is an indication of Europe to the northwest of the central continent, and of the New World in an island to the southeast.

In the preface, in the upper margin of the sheet, are listed the titles of no less than 102 works, Buddhist writings, Chinese annals, etc., both religious and profane, which the author consulted when making his map. Hotan says in the Santshai-thou-hoei, the Thou-chou-pien, the Ming-yi-thong-tchi and others, both the situation and the extent of countries are often false because of the lack of knowledge of Sanskrit. After careful study of the Fo-tsou-t'ong-ki map, and of the mistakes he had found in it, he had concluded that the draughtsman had drawn solely upon theSi-yü-ki, and not upon other sources, such as, for example, the Chi-kia-fang-chi and the Tshan'entchoan, for his nomenclature. Hence the mistakes. Even in those remote times, he says, there existed manuscript maps, named Go-Tenziku-Zu [Map of the Five Indies] (Slide #263) in the most famous temples, but they were even worse than that of the Fo-tsou-t'ong-ki. In spite of this author's boasting his map is, in reality, according to Nakamura nothing but a mutilated copy of the "Map of the Five Indies," made up from a confusion of heterogeneous and anachronistic materials, and including topographic names from the time of the Chan-hai-king up to modern times, some betraying European influence. This map became the prototype of Buddhist world maps, the Nan-en-budai Shokoku Shuran no Zu (a world map), the date of which is still uncertain, and the Sekai Daiso Zu (a world map), En-bu-dai-Zu (a world map), Tenjuku Yochi Zu (a map of India), a trilogy by Sonto, a Buddhist, are derived from Hotan's world map. The special features of these maps are the representation of an imaginary India, where Buddha was born, and the illustration of the religious world as expounded by Buddhists. In the Buddhist world maps or Shumi World, the space for the regions called Nan-sen-bu and Nan-en-budai, etc. (equivalent to the continents on the real earth) is very small. It has been maintained that their artists could not have closed their eyes, as far as these parts are concerned, to the objective world maps made in Europe. But these maps treat India as the heart of their world, and consequently we can say that they never recognized the European world maps. In their maps the Buddhists connected the Five Continents with the Spiritual World where the spirit of human beings must go after death. In these maps we find confusion of the visionary world and the real. Hotan's world map was so popular that it is not difficult to find today.

This map can be described as a characteristic specimen of the type of early East Asiatic maps whose main distinctive feature is their completely unscientific character. They are based not on objective geographical knowledge or surveys but only on the more or less legendary statements in the Buddhist literature and Chinese works of the most diverse types, which are moreover represented in an anachronistic mixture. Thus, on this map we see, in addition to the mythical Anukodatchi- pond which represents the center of the universe and from which flow four rivers in the four cardinal directions, in the left-hand upper corner of the map a region designated as Euroba, around which are grouped, clockwise, the following countries: Umukari (Hungary ?),Oranda, Barantan, Komo (Holland, the country of the redhaired), Aruhaniya (obviously Albania), Itaryia, Suransa (obviously France) and Inkeresu (England). Hiroshi Nakamura regarded this map, therefore, simply a mutilated copy of the map of the five Indies which is said to have come to Japan about 835, and a copy of which, dating from the 14th century, is preserved in the Horyuji Temple of Nara, Japan.

The copy in question is a reprint published by the bookdealer Bundaiken Uhei. The map has been reproduced in Mototsugu Kurita's atlas Nihon Kohan Chizu Shusei, Tokyo and Osaka, 1932 and in color in Hugh Cortazzi's Isle of Gold, 1992. Another, almost analogous edition of the same map, also dated 1710, but published by Chobei Nagata in Kyoto, is reproduced in George H. Bean's, A List of Japanese Maps of the Tokugawa Era, Jenkintown 1951 .

From every point of view, however, the map preserved among the precious objects in the North Pavilion of the Horyuzi Temple at Nara is the most interesting. It is a very large map, 1.67 m. X 1.78 m., magnificently drawn, brightly and agreeably colored, the sea in ultramarine, the mountains in a greyish-green, and Hiuen-tchoang's itinerary in vermilion. On the verso is the title, Go Tenjiku-Zu, but neither author's name nor date. The inscriptions are only partly completed. That in the right-hand margin bears an important notice, which says: "This map was made by the priest Zyukwai, in the fifth month of the third year Teizi. It is preserved in the convent of the Horyuzi temple. Thus this note gives us the author's name and the date of the first map of Horyuzi. The date corresponds to the month of May 1365.

The nomenclature of this map, as well as that of the Hosyoin map, is taken from the Si-yü-ki . They are just the itineraries of Hiuen-tchoang's travels. Among several maps of this kind that of Horyuzi is certainly the oldest and the most authentic in existence, though even it is not quite free from alterations. The others are very far from being in their original condition. Japan and Korea, northeast of the central continent, having no connection with the Si-yü-ki, must be later additions. In the Horyuzi and Hoyuzi and Hosyoin maps the Japanese Archipelago is shown by a bird's eye view, in Hotan's map and in its reduced reproduction it is shown in plan, adopted from a modern map, and in the Syûgaisyô map Korea has been placed in a rectangle to the northeast of the central continent in place of Japan. Is this because the map is of peninsular origin? All these heterogeneous elements, so different from the other parts of the map are clearly the result of retouching at a later date, so that, in its most authentic form, the map of Hiuen-tchong's itinerary had doubtless the shape of a shield, with a few small, rocky islands near the south and southeast coasts of the central continent, but without showing either Japan or Korea. The shape of the central continent is just the same as in the Korean mappamundi, as was demonstrated also by the hybrid map.

There must certainly have been formerly in China some examples of Hiuen-tchoang's itinerary, now long lost, while that taken from China to Japan must have been preserved there through long centuries. This very precious geographical relic, though it may have undergone some slight alteration, remains nevertheless to shed much light on the development of the Korean mappamundi.

From Nakamura's examination of a many manuscript and printed Korean mappaemundi he has come to the conclusion that they cannot in themselves be dated earlier than the 16th century. Their content however, their nomenclature, being for the most part that of the Han period, and taking into consideration the periods at which the few additions to that nomenclature have been made, is of a date not Iatter than the 11th century.

A hybrid map, forming a link between the Korean mappamundi and Hiuent-choang's itinerary, which dates from the 8th century, causes one to seek further the connection between these two groups of maps to establish the parentage of the Korean mappamundi.

The first portions of this monograph conveys the results of Nakamura's researches into the nomenclature of these maps from the historical point of view, and into their dates, but it is absolutely necessary to carry out alongside these examinations an investigation of their relationships in configuration, that is to say, to study their morphology. That is what is proposed for this final portion.

We do not know if Hiuen-tchoang's itinerary map was known from the commencement as The Map of the Five Indies, although it may commonly have been so called. The editors of the map preserved in the Hosyoin temple insisted that such a title was wrong and renamed it The Map of the Western Regions. Which is, in Nakamura's opinion, still not correct, for it contains not only the Western Regions but India and China also, so that Terazima is right in calling it The Map of the Western Regions and of the Indies, the title he gives to his small reproduction of part of Hiuen-thoang's map.

But by whatever title it may be called, its content is always that of Hiuen-tchoang's itinerary. Moreover, its central continent being always entirely surrounded by an immense ocean, it would seem that for its draughtsman this continent represented the whole of the then known world. This idea does not arise simply from the imagination, for the Hindus and the Chinese, like the Greeks, believed that an immense, unnavigable ocean entirely surrounded the habitable world. There is the same idea in the hybrid map and in Hotan's. In putting forward the idea that the Chinese of the 7th and 8th centuries believed that the geographical knowledge acquired by Hiuen-tchoang in his travels was concerned with the whole of the then known world we are saying that their world had become smaller than that known in the Han period, which is impossible. We shall presently see what really was the extent of geographical and cartographical knowledge in the Thang period.

Even as far back as the Tcheou Dynasty, Chinese cartography was well established where the court then had regular officials whose duties were concerned with the production and preservation of maps. Nevertheless certain scholars are doubtful about the positive knowledge of scientific cartography possessed by the Chinese of that period. Yet in face of the number of early references to maps it cannot be doubted that they had, under the Ts'in Dynasty, maps sufficiently exact for their own purposes. Under the Han Dynasty maps played a very important part in political and military affairs, and many of them covered a vast area. The invention of paper at the beginning of the second century, by the eunuch Ts'ai Louen, made possible a great step forward in cartography, thanks to its handiness, and its cheapness compared with wooden or bamboo tablets, or the linen or silk stuffs in use up to that time. At the same time the expansion of geographical knowledge under the Han Dynasty brought out the importance of the invention of paper. But it was not until the middle of the 3rd century that the correct scientific principles for the production of a good map were enunciated. It was P'ei Hsiu, 234-271, who formulated them so that he may rightly be called the father of Chinese cartography. Under the Thangs China had attained to a high degree of civilization, perhaps the highest it has ever reached, and cartography then made remarkable progress. This is the time when the famous Hai-wei-hoa-yi-thou [Map of the Celestial Empire and the Barbarian countries within the seas] was produced and presented to the Emperor by the celebrated cartographer Kia Tan, 730-805. It was an exact and detailed map 33 feet high and 30 wide, on the scale of one tsoun [a Chinese inch], to a 100 li [a Chinese league] that is, about 1:1,500,000, taking a li as 300 pou [a Chinese pace], and a pou as 5 tchih [a Chinese foot]. It must have covered a vast extent, almost the whole of Asia. To our great regret is has disappeared without any trace, but the author of the stela map at Si-gna-fou, engraved in 1137, of which the original drawing was made before the middle of the 11th century, said that he had consulted it, and that it contained many hundreds of kingdoms.

Therefore, so far from the area described in Hiuen-tchoang's travels being the whole of the world known to the Chinese at the Thang period, the world they actually did know was very much greater. Kia Tan's map is perhaps not a proof of this, since it no longer exists, and was always, in any case, kept secret at the Imperial Court, and so was not easily consulted even by those of the generation in which it was produced. But there is another extremely important map of this period known to us and existing today. It is of the whole of Asia, and in the Thang period would have been regarded as a mappamundi. It was originally in scroll form and in three parts, whose dimensions were respectively: 31.5 X 994 cm, 30.7 X 932 cm, and 29.5 X 217 cm. The first and third parts each contain two modern leaves, those of the first part being the preface by Onson Kosugi, 1834-1910, a well-known classicist, and those of the third a Government certificate. The document itself is therefore a scroll about 30 cm. high and 18 meters long.

This Sino-Tibetan map, like the imago mundi of the European Middle Ages, is simple but grotesque. A score of rectangles clumsily arranged, represent the various countries. The names of the countries, written in the rectangles are in Chinese and Tibetan characters. No one succeeded in deciphering these until Professor Teramoto did so, publishing his findings at the end of 1931 in an article entitled "Relations between Japan and Tibet in the history of Japan" (Slide #208) .

The map covers almost the whole of Asia, from the extreme east to Persia and the Byzantine Empire in the west, from the countries of the Uigours, the Kirghis and the Turks in the north to the Indies in the south, an area incomparably wider than that covered by Hiuen-tchoang's travels. It proves finally, therefore, that Hiuen-tchoang's map does not represent the whole of the world known to the Chinese at the time of the Thang Dynasty. Both maps came from the same temple so that they must have been co-existent there, and moreover, there is reason to believe that both had their origin at almost the same period. It is not very probable that the originators of these two maps tried deliberately to represent the world in two different ways. Moreover, the geometrical form of the Sino-Tibetan map was foreign to China, so that it must owe something to western influence. For it is a historical fact that, under the Thangs, the Chinese, having conquered the eastern Turks, annexed an immense territory, stretching from Tarbagatai in the north to the Indus in the south, and their national prestige was then at its zenith. They were constantly in touch diplomatically and commercially with Tibet, Persia, Arabia, India and other countries both by land and sea.
This alone is sufficient to prove that the Chinese at this period well knew that the Map of the Five Indies was not of the whole world, but that it extended into the west and north beyond the areas indicated by the Si-yü-ki. Why, then, should the area described by Hiuen-tchoang have been represented in the form of a shield surrounded by an immense ocean? Is it due to the influence of Buddhist or Brahmin cosmography? Did the draughtsman of this Buddhist pilgrim's itinerary exaggerate, making these travels represent the whole world? Could such an adaptation of the Buddhist cosmography afterwards have become traditional? This could not have been so, for Hiuen's travels were not always mapped in the form of a shield. There is one map of his journeys, drawn as an itinerary, preserved in the Tongdosa Temple in the province of Kei-shodo (south) in Korea. It is a scroll, 31 to 35.5 cm. high and 6.37 meters long, of which 1.49 m. on the left is text and the remainder the kind of drawing seen in Latin itineraries. The scroll begins on the extreme right with the Touen-houang region and finishes with Ceylon. It is dated April 1652, lunar calendar, is roughly drawn and has no special interest from the geographical point of view, but it serves to prove that the map of the travels of Hiuen-tchoang was not always drawn in shield form.

The fact that the Chinese represented the geographical content of the Si-yü-ki in world form, without taking the slightest trouble to show the true shape of countries they knew well, not even China itself shows that, at the period of the Thang Dynasty, 7th to 8th century, they had a mappamundi of the Tchien-ha-tchong-do (Slides #262I/J) type in common use that they adopted it just as it was, without modifying it at all, and entered on it the Si-yü-ki nomenclature.

Therefore it may be said that the Korean mappamundi had its beginnings at a period rather earlier that than of the Thang Dynasty; that a map very like it in form was introduced into Korea from China at a certain period and, thanks to the development of printing in the 16th century.

There is no doubt that the map by Jên-ch'ao, which also represented Jambu-dvipa as India-centric continent, utilized the Map of the Five Indies for reference, as we can see from the method of drawing both the boundary lines of the Five Indies and the courses of the four big rivers. But the distinctive feature of the map is that China, which had till then occupied a purely notional situation, was represented as a vast country in the eastern part of the continent with place-names of the period of the Ming Dynasty. Evidently the composition of this map owed much to the ideas of Chih-p'an and Jên-ch'ao's book contains another map entitled Tun-chên-tan-kuo-t'u [Map of the Eastern Region or China], which followed Chih-p'an's Topographical Map of the Eastern Region or China. Jên-ch'ao's map of Jambu-dvipa, however, is no mere combination of three maps in Fo-tsu-t'ung-ki. In his map we no longer see the juxtaposition of India on the west and China on the east, for India is placed in the center and China to the northeast of India in the eastern quarter of the continent. As may be read in the text of the Fa-chieh-an-li-t'u, there was a theory that, with the Pamir Plateau as its center, the world could be divided into four quarters, viz., the country of the elephants on the south (India), that of humans on the east (China), that of horses on the north (Mongolia, Central Asia and other regions of the nomadic peoples), and that of treasure on the west (Iran and other western regions). Jên-ch'ao based his map on this theory. Originating from an ancient Indian legend, it had been introduced into China as early as the 3rd century A.D., but it was not until the theory came to be taken up from the 7th century in Buddhist works, such as the Si-yü-ki, that it began to attract wide attention. Belief in this theory may explain why the Map of the Five Indies, in its composition of the world, arranges Chên-tan [China] on the east and Po-szu-kuo [Iran] on the west in opposition to Central Asia and India, though this no more than reflects their geographical position. To sum up, taking over Chih-p'an's view of China as equal with India, and on the basis of the theory of four divisions of the world implicit in the Map of the Five Indies, Jên-ch'ao tried to reunify the world which had been disintegrated into three regional maps in the Fo-tsu-t'ung-ki, and he thus succeeded in offering an image of the whole of Jambu-dvipa anew.

We cannot suppose, however, that Jên-ch'ao was the first to produce a map of this type. When we note that his map is not free from errors and omissions due to repeated transcription, and that almost contemporaneously there were maps of closely analogous type, such as the T'u-shu-pien map mentioned below, there is little doubt that maps of this kind had already been produced before by somebody. Possibly one such prototype of Jên-ch'ao's map, introduced into Japan and repeatedly copied, constituted the greatly transfigured Syûgaisyô map. The forerunner of these maps can therefore be traced back far beyond the end of the Ming Dynasty. Perhaps Jên-ch'ao, too, drew his map on the basis of a map of this type, simultaneously taking the Fo-tsou-t'ong-ki and other new maps for reference. In this map the Han-hai [Large Sea] is represented as a desert and the River Huang rises in Lake Oden-tala. This shows that Jên-ch'ao consulted some new maps of the Ming period.

At the beginning of the 17th century there appeared another map of Jambu-dvipa. This is the Ssu-hai-hua-i-tsung-t'u [Map of the Civilized World and its Outlying Barbarous Regions within the Four Seas] contained in the T'u-shu-pien, a Chinese encyclopedia compiled by Chang-huang and published in 1613. With the exception of many quaint islands lying scattered in the surrounding seas, this map has much in common with Jên-ch'ao's, the central continent consisting of India, with the place-names from the Si-yü-ki, and of China under the Ming Dynasty. There is no denying the close relationship of the two. In the text of the T'u-shu-pien is found a quotation from the prefatory note written by an unknown priest who drew this map. According to it, the map was newly compiled on the basis three maps in the Fo-tsu-t'ung-ki, as well as many other materials. This statement, however, is unbelievable, since the map is full of errors. It is more probable that, utilizing diverse information it was modelled mainly on the afore-mentioned prototype of Jên-ch'ao's map.

The map, however, has a distinctive feature of its own, namely, the outline of the continent in the center and a number of islands Iying scattered around it. The continent as a whole is wide in the north and pointed in the south, and still shows the traditional shape of Jambu-dvipa. But instead of the geometrical figure, its coastline is drawn more realistically and is indented. What attracts particular attention is that Fu-lin [the Byzantine Empire], west of the continent, is represented as a peninsula symmetrical with Korea in the east. Perhaps this indicates that people were no longer content with geographical representations in an unrealistic form, as the pilgrimage map of Holy India had been transformed into the map of the world, in which the intellectual interest requires a more objective and concrete image. And this growth of geographical consciousness led to a fresh demand for maps to comprise the whole of the known world. Among the islands shown on this map were Japan, Ryuku, P'u-kan [Pagan in Burma], Ta-ts'in [the Roman Empire], Yeh-mo-t'i [Java], and Ho-ing [Kalinga in Java], the last two recorded by Fa-hsien and I-tsing in their accounts of travel to India. But apart from them, most of the names given are fabulous and seem to have been taken from mystical, prophetic works such as the Hung-fan-wu-hsing-ch'uan written by Liu-hsiang in the period of the Han Dynasty. It may be because the Buddhist Holy Writings referred to some islands belonging to Jambu-dvipa that the compiler of this map arranged islands around the central continent so as to represent the various data obtained from sources other than the Si-yü-ki or the Fo-tsou-t'ong-ki.

Professor Nakamura treats this map in detail in his paper and states that the T'u-chu-pien (Slide #262K) map is a hybrid between the Korean mappamundi of Tchien-ha-tchong-do type and the map of Jambu-dvipa. He rests this statement on the ground that a map closely resembling the T'u-shu-pien map is to be found among the maps of the Tchien-ha-tchong-do type which were widely circulated in Korea. But with regard to their representation of both the central continent and outlying islands, these two maps have very little in common and there seems to be no need to seek any particular relation between them. It is true that among the Korean mappaemundi may be found one almost exactly like the T'u-shuh-pien map. But this can be explained on the supposition that the Koreans, happening to find a map of similar form in this widely circulated encyclopedia, adopted it as a substitute for their world map. Any close relationship, beyond this, can hardly be imagined between the Chinese Buddhist World Map and the Korean Tchien-ha-tchong-do. (Slides #262I/J)

But neither Jên-ch'ao's map nor thisT'u-shu-pien map could not go beyond Asia in their representation. Even so, it cannot be denied that a sense of a wide world is expressed in it. Certainly the map is full of inaccurate and even false information, but it gave the Chinese people a fresh conception of the world in contrast to their traditional view of their own country as the center of the world both culturally and geographically. Chang-huang, the compiler of the T'u-shu-pien, who was rather critical of Buddhist teachings, dared to insert this map in his book saying that "although this map is not altogether believable, it shows that this earth of ours extends infinitely". But so long as it remains a Buddhist map, dogmatism stands in the way of reality. In the T'u-shu-pien map (Slide #262K) the eastern coastline of the continent is represented with approximate accuracy, yet Fu-lin is drawn on the western side, is a fictitious peninsula similar to Korea on the east, so as to keep the symmetrical figure of the continent of Jambu-dvipa. As for the contents of the map, more respect was paid to the old classical authority than to the new geographical findings, so that many quaint, legendary place-names were mentioned to advertise the bigoted belief that the world of Buddhist teaching included even the farthest corners. The problem now was to introduce heterogeneous information without conflict with the Buddhist teachings and to give the dogma some apparent plausibility. So it was of course quite inconceivable that the compiler of this map should utilize the newly acquired information for the purpose of altering the dogmatic notion of the world. Therein lies the character and limitation of the Buddhist maps. The T'u-shu-pien map (Slide #262K)was rough and small, but as it appeared in a popular encyclopedia, it attracted wide attention, which in time came to affect even the maps produced in Japan.

Among the maps made by Japanese we find none which regard China as the center of the world. Only two world maps representing Chinese thought were published in Japan, the Daimin Kyuhen Bankoku Jinseki Rotei Zenzu [Complete Map of the Nine Large Regions of the Greater Ming, i.e., China, and all Countries of the World, with data on distances], a world map, date unknown; and the Choi Ichiran (a world map) published in 1835, and these are reprints of the Tenka Kyuhen Bankoku Jinseki Rotei Zenzu (a world map) published in China, 1663. There are not many of these Chinese world maps in Japan, but the remaining Daimin Kyuhen Bankoku Jinseki Rotei Zenzu, (Slide #263.5)of various sizes, are common. In this map illustrated on the following page, China is situated predominantly in the center, and the other countries are arranged round it, North and South America being drawn as separate small islands, and Europe and Africa being left uncompleted; this expresses contemporary Chinese ideas on world geography. Countries like Korea, Japan and Ryuku are only described in the text and are not represented on the map. The fact that such dogmatic world maps were widely favored in old Japan shows that some Japanese in the Age of National Isolation believed China to be the center of the world and all the other countries to be in subordination to her. The shhet under discussion is a reprint by the bookdealer Yahaku Umemura in Kyoto, which is tentatively dated 1700 by Professor Kurita.

*Ayusawa, S., "The types of world map made in Japan's Age of National Isolation", Imago Mundi #10, pp. 123-128, Figures 2 and 3.
*Bagrow, L., History of Cartography, pp. 197-205.
*Cortazzi, H., Isles of Gold, Antique Maps of Japan, pp. 6-38, Plates 48/49 (color).
*Hulbert, H.B., "An Ancient Map of the World", American Geographic Society Bulletin, pp. 600-605.
*Muroga, N. and Unno, K., "The Buddhist world map in Japan and its contact with European maps"
*Nakamura, H., "Old Chinese World Maps Preserved by the Koreans", Imago Mundi #4, pp. 3-22.
*Nakamura, H., East Asia in Old Maps, p.8, Figure 1.
*Needham, J., Science and Civilization in China, vol. 3, pp. 566-568, Figure 242.
*Raisz, E., General Cartography

* illustrated

Late Medieval Maps