Pramāṇa · Search · About · API

On Yuan Chwang's Travels in India, 629–645 A.D. v1.c8: Chapter V. Chuan II: General Description of India

On Yuan Chwang's Travels in India, 629–645 A.D. v1.c8

Language
English

Contents

Chapter V. Chuan II: General Description of India

v1.c8.1

CHUAN II.

Its Names

v1.c8.s1.1

The pilgrim having now arrived at the frontiers of the great country which he calls *Yin-tu* (India) gives his readers a "Pisgah-sight" of the land before taking them through its various kingdoms. And first he tells them of its name and its meaning and probable origin. His statements about the name may be roughly rendered as follows—

v1.c8.s1.2

We find that different counsels have confused the designations of *T'ien-chu* (India); the old names were *Shên-tu* and *Sien* (or *Hien*)-*tou*; now we must conform to the correct pronunciation and call it *Yin-tu*. The people of *Yin-tu* use local appellations for their respective countries; the various districts having different customs; adopting a general designation, and one which the people like, we call the country *Yin-tu* which means the "Moon".

v1.c8.s1.3

This rendering differs in some respects from that given by Julien which is neither very clear nor correct. Here, however, as in several other passages of the Records, it is not easy to make out the precise meaning of the author's statements. It is plain, however, that he is not dealing with names given to India generally but only with those used in Chinese books. Then his words would seem to indicate that he regarded *T'ien-chu*, *Shên-tu*, and *Sien-tou* as only dialectical varieties or mistaken transcriptions of *Yin-tu*, which was the standard pronunciation. Further his language does not seem to intimate, as Julien understood it to intimate, that Yin-tu was the name for all India used by the inhabitants of the country. In some other works we find it stated that Yin-tu was the native name for the whole country, and Indu-deśa given as the original Sanskrit term. Our author may have had this opinion but this does not seem to be the meaning of his statements here. On the contrary he apparently wishes us to understand that the natives of India had only designations of their own States, such as Magadha and Kausambhi, and that they were without a general name under which these could be included. It was the peoples beyond, as for example the Turks, who gave the name Yin-tu, and the Hu who gave Sin-tu, to a great territory of uncertain limits. Then the Buddhist writers of Kashmir, Gandhāra, and other countries beyond India proper, seem also to have sometimes used the name Yin-tu. But, as I-ching tells us, although this word may mean "moon" yet it was not the current name for India. In Buddhist literature India is called Jambudvipa, and portions of it Āryadeśa and Madhyadeśa.¹ One of the other names for India to be found in Buddhist literature is Indravardhana. But in the Chinese accounts of letters or missions sent by Indian rajahs to the court of China the rajahs are only represented as styling themselves kings of special countries in India. Thus the great Śīlāditya, who treated our pilgrim with great honour, is made in Chinese history to call himself king of Magadha.

v1.c8.s1.4

¹ Nan-hai-ch'i-kuei, ch. 25; Hunter's Ind. Emp. p. 33.

v1.c8.s1.5

Let us now examine in detail Yuan-chuang's statements about the terms he quotes as used in China to denote India and the history of these terms. The old name, as he tells us, is that which he, following precedent, writes Shên-tu (身毒) as the characters are now pronounced. This word emerges in Chinese history in the account which the famous envoy Chang Ch'ien (Kien) gives of his experiences in the Ta-hsia country (Bactria). In that we read that when Chang returned from his mission to the West he reported to Han Wu Ti (apparently about B. C. 123) that when in Ta-hsia he had seen bamboo poles and cloth from a district which is now comprised in the Province of Ssŭchuan. He had been told, he relates, that these commodities had been obtained at Shên-tu, as the name of the place is given in the ordinary texts of his report to the Emperor. Now Chinese writers tell us, and Western scholars have adopted and repeated the statements, that the Shên-tu of this story was India, and that all the other designations for that country in Chinese books such as Hsien-tou, Hsien-tu, Kan-tu, Küan (or Yuan)-tu, T'ien-chu, T'ien-tu, and Yin-tu are only phonetic corruptions of Shên-tu. These opinions seem to have been lightly formed and heedlessly followed, and it may be useful for us to enquire whether they have a good basis.

v1.c8.s1.6

In the first place then we find that there is doubt as to what was the precise form of the name of the country in Chang's statement. So instead of the character for Shên in Shên-tu given above we meet with several various readings. Such are 乾 and 犉 which probably represent one sound, something like Get or K'at. Now a foreign name like K'atu or Gachu as a name for India seems to have been in use. Then a third various reading for the Shên of Shên-tu is K'ien or Kan (乾) which may have been originally a copyist's slip for one of the characters read K'at.¹ We find also a fourth various reading for the syllable Shên of Shên-tu, viz—Küan or Yun (捐).² But the country described in Chinese literature under the name Yun-tu was evidently one to the east or north-east of all that has been called India.³ Then accepting the character now read Shên as the genuine text of Chang's report we are told that in this name it is to be pronounced like *in* or *yin*. This does not seem very improbable. But an etymological authority tells us that the character in question has, in this name, the sound *T'ien*.¹ There may be some truth in this statement. But it is not supported by authority, and seems rather fanciful.

v1.c8.s1.7

¹ Shih-chi (史記), ch. 123, Commentary. In the T'ung-chien-kang-mu, ch. 4, Yuan-shou (元狩) 1st y., this passage of the Shih-chi is quoted with the reading Kan-tu (乾篤) instead of Shên-tu. See also Kanghsi Dict. s. v. 身.

v1.c8.s1.8

² Han-Shu, ch. 96.

v1.c8.s1.9

³ T'ung-chih-liao, the Tu-yi-liao, ch. I; Han-Shu, l. c.

v1.c8.s1.10

¹ Wên-ch'i-tien-chu (問奇典註), *ch.* 2. p. 22. The change of Shên-tu into T'ien-tu may point to a Burmese pronunciation of Sindu as Thindu.

v1.c8.s1.11

The district or region which the envoy Chang reported as named, let us continue to say, Shên-tu, is briefly described by him and others of the Han period. It was several thousand *li* south-east from Bactria, near a river (or sea); its inhabitants used elephants in fighting. Some writers describe them as Buddhists; and they were in many respects like the people of Bactria, or like the Geti (Yue-ti) according to another account. Their country was about 2000 *li* south-west from what is now the Ch'êng-tu and Ning-yuan districts in Ssŭchuan, and it had a regular trade with the merchants of the Ch'êng-tu district, some of whom seem to have settled in it. Further, this country was not far from the western border of the Chinese empire in the Han time, and it was on the way from China to Bactria. So though the name Shên-tu came to be afterwards given to India yet in its first use it apparently denoted a small region in what is now Yunnan and Burmah.²

v1.c8.s1.12

² Han-Shu, *ch.* 95; Hou Han-Shu, *ch.* 88; Ma T. 1. *ch.* 338 gives much information about India compiled not very carefully from previous authorities; his account is translated in Julien's Mélanges p. 147.

v1.c8.s1.13

The name Hsien-tou was apparently applied to a region different from that designated Shên-tu.³ Like Hsien-tu (縣度), of which term it is perhaps only a variety, this name was probably used first by the Chinese for the Indus, called Sindhu in Sanskrit. The name was afterwards extended by them to a mountainous region, perhaps Ladak, through which the Indus flows. We find the Hsien-tu country mentioned in the same passages of the Han History with Shên-tu.

v1.c8.s1.14

³ But Hsien (Hien)-tou (縣度) came to be used as a name for India, and we find it described as a native designation for the whole country properly called Indravardhana. Su-kao-sêng-chuan, *ch.* 2. See also Fang-chih, *ch.* 1.

v1.c8.s1.15

We next come to T'ien-chu (天竺) and T'ien-tu (天篤) said to represent only one name pronounced something like Tendu or Tintok. We are told by one Chinese writer that the name T'ien-chu was first applied to India in the Han Ho-Ti period (A.D. 89 to 106) but the authority for the statement is not given. Another account makes Mêng K'an (about A.D. 230) the first to identify T'ien-chu with Shên-tu, but this likewise is unsupported by authority. We are also told that the chu (竺) of T'ien-chu is a short way of writing tu (篤), a statement which is open to very serious doubt.¹ This word tu occurs in the ancient classical literature, and native students declare that it represents an earlier chu. This is specially noted with reference to the occurrence of tu in a wellknown passage of the "Lun-Yü". Then as to the first part of the name there seems to have been an old and perhaps dialectical pronunciation of the character as Hien or Hin. This pronunciation is found at present in the dialect of Shao-wu foo in the Province of Fuhkeen in which 天竺 is read Hien-tu.²

v1.c8.s1.16

¹ Shih-chi, l. c.

v1.c8.s1.17

² The Chinese Recorder for September 1891, p. 408.

v1.c8.s1.18

But what was the sound originally represented by the character now read Chu in the compound T'ien-chu? It seems that no satisfactory and decisive answer can be given at present to this question. We find that in the Han period the character represented several sounds which cannot be said to be very like each other. The upper part chu meaning bamboo is not significant here, we are told, but only phonetic; and the lower part is significant, and refers the word to the category earth. The character might then be read something like du, but this account of the syllable may be doubted, as we learn also that the character was read like *tek*, an old and still current pronunciation of the word for bamboo. Then this same character was also read as *chuh*, *tuh*, *kat*, and *ko* or *gou*.¹ Something like the last was perhaps the earliest pronunciation of the character, and this is probably a corruption or abbreviation of a form like *kao* (竹) or *kung* (笁). This last form, unknown to the dictionaries apparently, occurs often in Japanese texts of Buddhist books instead of the character for *chu*. Now in the fact that *ko* or *gou* was an old sound of this character we have an explanation of a proper name found in the Tibetan version of the Buddhist "Sūtra in Forty-two Sections". One of the two Indian monks who came to China in the time of Han Ming Ti, and translated or drew up the above scripture, is styled in Chinese text *Chu Fa-lan*. These words apparently represented an Indian name like Dharma-pushpa, that is, Flower of Buddhism. Now the Tibetans transcribing the sounds of the characters for Chu Falan according to their own language wrote apparently *Go-ba-ran* and this became in the modern transcription *Gobharana*. This last word is neither Sanskrit nor Tibetan, but it has been adopted by Feer who has been followed by Beal and Eitel. That Chu in such expressions as Chu-Fa-lan (竺法蘭), is not part of the name, but means "India" or "Indian" we know from its occurrence in other expressions of a similar kind. We may also infer it, in this case, from the fact that it does not occur in some old editions of the above-mentioned scripture, which have only Fa-lan as the name of the Indian monk. So also in another Tibetan work we find him described as "Bhāraṇa Paṇḍita".² There is also another word in which we may perhaps recognize the *ko* pronunciation of our character *chu*. This word is the old "Tangut", more correctly Tan-ku, which was the Turkish-Persian designation for the country now called Tibet.¹ It is not improbable that, as some have supposed, this Tan-ku is simply the *T'ien-chu* of Chinese writers. And so this last may have been originally a Turkish term, used to denote a country immediately to the west of China, and between that country and Bactria.

v1.c8.s1.19

¹ Shuo-wên, ed. Kuei Fu-hsio, s. v. 竹. In the Fo-kuo-chi this character must be pronounced like *Tuh* or *Tak* as it forms the first syllable of the name Takshaśila.

v1.c8.s1.20

² Feer's Le Sūtra en 42 Articles p. 47; Ssŭ-shih-êrh-chang-ching, and Bun. No. 678 and Appx II. col. I; Journal Bengal A. S. No. LI. p. 89; Huth, Geschichte d. Bud. in d. Mongolei, tr. from Tibetan, S. 101.

v1.c8.s1.21

¹ Georgi's Alph. Tib. p. 10. In the Hai-kuo-t'u-chih it is expressly stated that *T'ien-chu* has been identified with the modern Hsi-Tsang or Tibet.

v1.c8.s1.22

*T'ien-tu*, on the other hand was the name of a place in the Eastern Sea mentioned in the "Shan-hai-ching" along with Chao-hsien or Korea. This place was afterwards identified wrongly with the *T'ien-chu* of writers on India and Buddhism.² But we find mention also of another *T'ien-tu* (written in the same way), a small country to the west of China, which has been supposed by some to be the Shên-tu of Chang Ch'ien.

v1.c8.s1.23

² Shan-hai-ching, ch. 18.

v1.c8.s1.24

Whatever the name *T'ien-chu* may have signified originally, however, it came to be given by the Chinese in their literature to the great extent of territory between the Bay of Bengal and the Arabian Sea, and reaching from the Kapis country in the north to Ceylon in the south. Thus used it supplanted the old Shên-tu, and all other names for India among the Chinese; and it continued to be the general literary designation for that country down to the T'ang period when the new name *Yin-tu* was brought into fashion. We even find the term *T'ien-chu* used with a wider application, and it is employed as a synonym for "Buddhist countries", for example, in a title given to the "Fo-kuo-chi" of Fa-hsien. Nor has the term been quite put out of use by Yuan-chuang's correct name *Yintu*, and Yuan-chuang himself continues to use it occasionally. We find also each of its component parts sometimes made to do duty for the whole. This *Chung-T'ien* and *Hsi-T'ien* are respectively Middle and West India, while *Chu* in the Han and Ch'in periods and later was commonly used for India or Indian, a way in which *Hsi-T'ien* is also used.

v1.c8.s1.25

Leaving *T'ien-chu* to continue as a Chinese name for India, Yuan-chuang puts aside what he considers to be the corruptions of the term Yin-tu, and proceeds to use that form as the correct designation of the country. He goes on to suggest a reason for this word, meaning "moon", having come to be so employed. His explanation is apparently as follows—

v1.c8.s1.26

The unceasing revolutions of mortals' existences are a dark long night; were there not a warden of the dawn they would be like the night with its lights which succeeds the setting of the sun; although the night have the light of the stars that is not to be compared to the light of the clear moon. Hence probably India was likened to the moon as [since the sun of the Buddha set] it has had a succession of holy and wise men to teach the people and exercise rule as the moon sheds its bright influences,—on this account the country has been called Yin-tu.

v1.c8.s1.27

The comparison and explanation of our author, it must be admitted, are sorry things; and they are not improved in any of the translations. But the passage has probably some copyist's mistakes, and we must at least supply a clause which apparently has dropt out of the text. This clause is the important phrase *Fo-jih-chi-yin* (佛日既隱) which means "when the sun of the Buddha set". I have restored these words within square brackets in the body of the pilgrim's explanation, but it is probable that they occurred at the head of it also. The "long night" of the text is the interminable succession of renewed existences to non-Buddhists, and to the Buddhists the period between the death of one Buddha and the advent of another, but it is rather a state of affairs than a tract of time. It denotes a condition of spiritual darkness to mankind, an endless repetition of mortal life in many varieties; each life ignorant of the one before, and without any hint of the one to follow. There is no Buddha in the world; and so there is no one to end the night, and bring in the dawn of Nirvāna. The Buddha is the *Ssŭ-ch'ên* (司 晨) or Warden of the Dawn, the officer in charge of daybreak who ushers in the light of intelligence and the perfect way.

v1.c8.s1.28

Now on earth, when the "lights of night" succeed the setting of the sun, there are stars, and there is the moon. The stars, however, have only a shining, the brightness of a glow. But the moon has a light which illuminates and influences the world, and which transcends in brightness all other lights of the night. So other lands have had sporadic sages who made a glory for themselves revolving each in his own peculiar eccentric orbit. But India had a regular succession of great Sages who 'followed the great wheel' of ancient authority, each successor only expounding, renewing, or developing the wise teachings of his divine or human predecessors; thus keeping the light of primitive revelation shining among mortals. In Buddhist writings the Buddha is often compared to the moon, while the stars are sometimes the rival teachers of his time, and occasionally his own great disciples.

v1.c8.s1.29

A later Chinese writer, apparently under the impression that he had the authority of Yuan-chuang for the statement, tells us that *T'ien-chu* means moon. But he, like several other authors, explains the giving of this name to India in a different way from that described by the pilgrim. He says that the country was called *T'ien-chu* or Moon because it was as great and distinguished above the other countries of the world, as the moon is great among the stars of night—"velut inter ignes Luna minores". Other writers, like I-ching for example, are more discreetly wise, and refrain from proposing any explanation of the names for India. Admitting, they say, *Yin-tu* to be a Sanskrit term denoting the moon, yet it was not for that reason that the Chinese gave it as a name to the country, nor is the name the universal one. *Yin-tu* is the Chinese name for India as *Chi-na* and *Chen-tan* are terms used in that country to denote China, and apart from such use these names have no signification.¹ This is going too far, and the word India at least has a satisfactory explanation. When our pilgrim enquired about the size and form of the country, he was told that it was shaped like a crescent or, as it is in the text, a half-moon. The term used was apparently Indu-kalā, transcribed Yin-t'ê-ka-lo (印特伽羅).² This word means a digit of the moon or a crescent, but it is rendered in Chinese simply by yueh or moon. It was perhaps this fact which led to the absurd comparison and explanation of our text.

v1.c8.s1.30

¹ Nan-hai-ch'i-kuei l. c.

v1.c8.s1.31

² Supplement to I-ch'ie-ching-yin-yi, ch. 3. This of course is not the origin of the name for India, but it may account for the Chinese use of Yin-tu as a designation for the country.

v1.c8.s1.32

Our author in this passage mentions another general name for India, viz—Country of the brahmins (P'o-lo-mên-kuo).

v1.c8.s1.33

Among the various castes and clans of the country the brahmins, he says, were purest and in most esteem. So from their excellent reputation the name "Brāhmaṇa-country" had come to be a popular one for India.

v1.c8.s1.34

Now this is also a foreign designation, and one used by the Chinese especially. It does not seem to have been ever known, or at least current, in India. In Chinese literature we find it employed during the Sui period (A.D. 589 to 618) but it is rather a literary than a popular designation. In the shortened form Fan kuo (梵國), however, the name has long been in common use in all kinds of Chinese literature.

v1.c8.s1.35

The territory which Yuan-chuang calls Yin-tu was mapped off by him, as by others, into five great divisions called respectively North, East, West, Central, and South Yin-tu. The whole territory, he tells us,

v1.c8.s1.36

was above 90 000 li in circuit, with the Snowy Mountains (the Hindu Kush) on the north and the sea on its three other sides. It was politically divided into above seventy kingdoms; the heat of summer was very great, and the land was to a large extent marshy. The northern region was hilly with a brackish soil;

v1.c8.s1.37

the east was a rich fertile plain; the southern division had a luxuriant vegetation; and the west had a soil coarse and gravelly.

Indian Measures of Space

v1.c8.s2.1

Our author now proceeds to give the names of measures of space and time which were in use among the people of India or were taught in their standard books of learning and religion.

v1.c8.s2.2

He begins at the top of the gradation with the Yojana which, he says, had always represented a day's journey for a royal army. The old Chinese equivalent for it, he says, was 40 *li*, the people of India counted it as thirty *li*, while the Buddhist books treated it as equal to only sixteen *li*.

v1.c8.s2.3

We are not told, however, that in India the Yojana varied in different places and at different times.

v1.c8.s2.4

Then the Yojana, he states, was divided into eight Kroṣa, the Kroṣa into 500 Bows, the Bow into four Cubits, and the Cubit into twenty-four Fingers. Forgetting, apparently, to mention the division of the Finger into three Joints Yuan-chuang proceeds to state the division of the Finger into seven Wheat (properly Barley)-grains. Thence the subdivision by sevens is carried on through the Louse, the Nit, Crevice-dust, Ox-hair [Dust], Sheep-wool [Dust], Hare-hair [Dust], Copper [Dust], Water [Dust], and Fine Dust to Extremely Fine Dust. This last is the ultimate monad of matter and is indivisible.

v1.c8.s2.5

This enumeration of Indian measures of space was apparently written down from memory, and it does not quite agree with any of the other accounts we have. In the Abhidharmamahāvibhāsha-lun,¹ compiled by the 500 Arhats and translated by Yuan-chuang, we find a similar enumeration, leaving it undecided, however, whether "seven copper-dusts" made one "Water-dust", or seven of the latter made one of the former. In this, and in the other books in which we find the measures of space given, the word for dust is added to each of the terms Ox-hair, Sheep-wool, Hare's-hair, Copper, and Water, and I have accordingly inserted it in the version here given of Yuan-chuang's account. Instead of *t'ung*, copper, the D text has *chin*, gold, perhaps used in the sense of metal, and this is the reading of Yuan-chuang's "Abhidharma-tsang-hsien-tsung-lun". Then the "Abhidharma-kośa-lun", which also has *chin* instead of *t'ung*, makes seven "metal-dusts" equivalent to one "water-dust" thus reversing Yuan-chuang's arrangement.¹ The word dust here should perhaps be replaced by atom or particle.

v1.c8.s2.6

¹ Abhidharma-ta-vibhāsha-lun, *cls* 136 (Bun. No. 1263).

v1.c8.s2.7

¹ Abhidharma-tsang-hsien-tsung-lun, *ch.* 17 (No. 1266); Abhidharma-kośa-lun, *ch.* 12 (No. 1267).

v1.c8.s2.8

Another enumeration of Indian measures of space is given in the Lalitavistara and its translations Tibetan and Chinese, and another in the Avadāna XXXIII of the Divyāvadāna of Messrs Cowell and Neil.² The latter is represented in the Chinese collection of Buddhist books by four treatises. In none of all these works is there anything corresponding to the words "copper" and "water" of our author's list. Moreover each of them makes the Window-Dust or Sunbeam-mote—the "Crevice-Dust" of our author—to be one seventh of a Hare (or Moon)-Dust and equal to seven particles of Fine Dust. Julien took the "copper water" of our text to be one term and translated it by "l'eau de cuivre (Tamrāpa?)", but this is undoubtedly wrong.³ In this gradation of measures the "Extremely Fine Dust" is a monad of thought, a logical necessity, and has no separate existence in matter. The lowest actual unit of matter is the *aṇu* of the Divyāvadāna, which is the "Fine Dust" of our author. This too, however, though visible to the deva-sight, is invisible to the human sight and impalpable to the other human senses. But it is a material substance, the most minute of all material sizes and quantities, and the ultimate atom into which dust or metal or water can be analysed. It takes seven of these, according to some, to equal one Atom (truṭi or tuṭi), and seven of these to make one Sunbeam-mote. If we omit the two words "Copper" and "Water" from our text, and remove the term "Crevice-Dust" to its place, we have an enumeration of measures which agrees substantially with that of the Divyāvadāna up to the Krośa. Some of the Chinese texts represent the Krośa, translated by shêng (聲) a sound, to be 2000 Bows, and in some the Barley-grain is subdivided, not as by Yuan-chuang, but into seven Mustard-seeds.

v1.c8.s2.9

² Lalitavistara *ch.* 12: Foucaux's Rgya-cher-rol-pa, p. 142 and note; Fang-kuang-ta-chuang-yen-ching, *ch.* 4 (No. 159); Divyāv. p. 644; Matanga-sūtra, *ch.* 2 (No. 645).

v1.c8.s2.10

³ See also the Tsa-abhidharma-hsien-lun, *ch.* 2 (No. 1288); Alberuni, *ch.* XXXIV and XXXVII; Abhidharma-shun-chêng-li-lun, *ch.* 32. In this treatise we have all the measures of space given by Yuan-chuang but the "Metal-dust" is one-seventh of the "Water-dust". It gives also the division of the Finger into three Finger-joints.

Measures of Time

v1.c8.s3.1

Our author next goes on to describe the measures of time in India, beginning with the divisions of the Day-night period. Here also he mainly follows Sanghabhadra's treatise¹, and differs from most other writers, Buddhist and orthodox.

v1.c8.s3.2

¹ Abhidharma-shun-chêng-li-lun, l. c. For the measures of Time generally see the references in the above note 3 on p. 142.

v1.c8.s3.3

He calls the Kṣaṇa the shortest space of time and makes 120 of it equal to one Tatkṣaṇa. Then 60 Tatkṣaṇas make one Lava, 30 Lavas make one Muhurta, five of these make one "time" (時), and six 'times' make one Day-night. The six 'times' of this last are, we are told, distributed equally between the day and the night. But the non-Buddhist people of India, Yuan-chuang tells us, divided the day and night each into four "times".

v1.c8.s3.4

It will be seen that Yuan-chuang here puts the Kṣaṇa below the Tatkṣaṇa, in this agreeing with the Abhidharma treatises of Sanghabhadra and Dharmatara. The Divyāvadāna, on the other hand, makes 120 Tatkṣaṇas equal to one Kṣaṇa, and 60 Kṣaṇas equal to one Lava. In some Chinese versions of the sacred books the tatkṣaṇa is not mentioned. The kṣaṇa is defined as the time occupied by a woman in spinning one hsün (尋) of thread, but the word is generally used by Buddhist writers in the sense of an instant, the twinkling of an eye, the very shortest measurable space of time.

v1.c8.s3.5

The word kṣaṇa is commonly transcribed in Chinese books as in our text, and it is rarely translated. The lava is sometimes rendered by shih (時), time, and sometimes by fēn (分), a division. So also Muhurta is sometimes translated by shih, time, but more frequently by hsü-yü (須臾), an instant or moment, such being also the original meaning of muhurta. But hsü-yü when used as a translation of this word does not denote an instant but a period of 48 minutes, the thirtieth part of a Day-night. The day is divided into three "times", viz. forenoon, noon, and afternoon, and hence it is called Trisandhya. In like manner the night is divided into three "times" or watches and hence it is called Triyāmā.

v1.c8.s3.6

Our author next goes on to enumerate the divisions, natural and artificial, of the month and the year in India. He distinguishes between the common four-fold division of the seasons, and the three-fold one used by Buddhists. The latter division was into a hot season (Grīshma) followed by a rainy season (Varsha), and then a cold season (Hemanta). We have next the names of the months of the year in their order beginning with Chaitra. Then comes an interesting passage which, as it appears in our texts, presents some difficulty. The meaning seems to be something like this—

v1.c8.s3.7

"Hence the professed Buddhists of India, complying with the sacred instructions of the Buddha, observe (lit. sit) two periods of Retreat, either the early or the later three months. The former period begins on our 16th day of the 5th month, and the latter on the 16th of the 6th month. Previous translators of the Sūtras and Vinaya use "Observe the summer" or "Observe the end of the winter". These mistranslations are due to the people of outlying lands not understanding the standard language, or to the non-harmonizing of provincialisms".

v1.c8.s3.8

The first sentence of this passage evidently means that the Buddhist monks of India could make either the former or the later three months of summer their period of Retreat. My interpretation of the passage differs a little from that of Julien who substitutes *yü* (雨) *rain*, for the *liang* (兩) *two*, of the text, supporting his change of reading by a quotation of the present passage in a Buddhist Cyclopedia. But one of two copies of this Cyclopedia in my possession gives *liang* and the other has *huo* (或). Moreover all texts of the "Hsi-yü-chi" seem to agree in having *liang* here: and we read in other books of two and even three periods of Retreat. For the monks of India, however, these were all included within the Rain-season, the four months which began with the 16th of their fourth month and ended on the 15th of the 8th month. The full period of Retreat was three months; and Buddha ordained that this period might be counted either from the middle of the fourth or the middle of the fifth month. The conjecture may be hazarded that Yuan-chuang originally wrote *liang-yü*(兩雨)-*an-chü* that is "two Rain-Retreats" and that a copyist thinking there was a mistake left out the second character. This restoration does not make good style but something of the kind is apparently needed as Yuan-chuang's expression for the Retreat was *yü-an-chü*.

v1.c8.s3.9

The Sanskrit term for the Retreat is Varshā (in Pali Vassa) which means simply rains, the *rainy season*, from *varsha* which denotes, along with other things, *rain* and *a year*. The usual expression for "keeping Retreat" is *varshām vas* (in Pali, *vassam vasati*) or *varshām sthā*, meaning respectively to reside, and to rest, during the rainy season. For these terms the Chinese give various equivalents such as the *Tso-hsia* and *Tso-la* of some, and the *Tso-an-chü* or *Tso-yü-an-chü* of Yuan-chuang and others. For the Buddhists of India as for the other people of that country the "rainy season" began on the 16th of the month Ashādha (the fourth of their year), and continued for four months. This was chiefly for religious purposes, but to the non-Buddhists of India three months of this period formed also their summer. This may help to explain the use of the phrase *Tso-hsia* which is a short form for the full expression *Tso-hsia-yü-an-chü* meaning "to observe the Summer Rain Retreat". Then *Tso-hsia* and Tso-la mean also to pass a year as an ordained monk, the precedence of a brother being settled by his "years in religion". The phrase Tso-la or Tso-la-an-chü is used specially of the strict anchorite who observed two Retreats, one in the summer and one in the winter. It might be also applied to brethren in strange lands, Tokhara for example, whose Rainy season occurred at the end of the winter. Yuan-chuang seems to think that the terms Tso-Hsia and Tso-La are not correct renderings from the Sanskrit and they certainly are not literal translations. He supposes the mistakes to have arisen either from the translators having been natives of countries remote from Mid-India, and so ignorant of the correct term and its proper pronunciation, or from the use of an expression which had only local application and currency. But the "non-harmonizing of provincialisms" denotes not only the misuse of local terms, but also ignorance of the idioms in one language which should be used to represent the corresponding idioms of another. Thus a Chinese or Indian scholar translating a Sanskrit book into Chinese without a thorough knowledge of the Sanskrit and Chinese idioms would not harmonize the countries' languages. Julien takes "Mid-kingdom" here to mean China but it certainly denotes Mid-India. In that region people called the Rainy season Varshā, but in other places the word was pronounced vasso, or barh, or barkh, or barsh. So translators, Yuan-chuang thinks, may have in some cases mistaken the word, or they may have misunderstood either the original, or the Chinese term they were using in translation. Thus the important fact that the Retreat was ordained on account of the Rains is put out of view by the renderings Tso-Hsia and Tso-La. There was not, however, any ignorance of Sanskrit or Chinese in the use of these terms, and good scholars in the two languages such as Fa-hsien and I-ching use Tso-hsia and An-chü indifferently. In countries in which there was no long regular Rainy season the Retreat became of importance as a time for spiritual improvement by study of the sacred books and prolonged meditation, and as giving a year's seniority to the brother among his brethren.

Cities and Houses

v1.c8.s4.1

We have next a short description of the general characters of the cities and buildings of India. The passage is an interesting one and the meaning may be given somewhat as follows—

v1.c8.s4.2

"As to their inhabited towns and cities the quadrangular walls of the cities (or according to one text, of the various regions) are broad and high, while the thoroughfares are narrow tortuous passages. The shops are on the highways and booths (or, inns) line the roads. Butchers, fishermen, public performers, executioners, and scavengers have their habitations marked by a distinguishing sign. They are forced to live outside the city and they sneak along on the left when going about in the hamlets. As to the construction of houses and enclosing walls, the country being low and moist, most of the city-walls are built of bricks, while walls of houses and inclosures are wattled bamboo or wood. Their halls and terraced belvederes have wooden flat-roofed rooms, and are coated with chunam, and covered with tiles burnt or unburnt. They are of extraordinary height, and in style like those of China. The [houses] thatched with coarse or common grass are of bricks or boards; their walls are ornamented with chunam; the floor is purified with cow-dung and strewn with flowers of the season; in these matters they differ from us. But the Buddhist monasteries are of most remarkable architecture. They have a tower at each of the four corners of the quadrangle and three high halls in a tier. The rafters and roofbeams are carved with strange figures, and the doors, windows, and walls are painted in various colours. The houses of the laity are sumptuous inside and economical outside. The inner rooms and the central hall vary in their dimensions, and there is no rule for form or construction for the tiers of the terraces or the rows of high rooms. Their doors open to the east, and the throne faces east.

v1.c8.s4.3

For seats all use corded benches. The royal family, the grandees, officials and gentry adorn their benches in different ways, but all have the same style (or form) of seat. The sovereign's dais is exceedingly wide and high, and it is dotted with small pearls. What is called the "Lion's Seat" (that is, the actual throne) is covered with fine cloth, and is mounted by a jewelled footstool. The ordinary officials according to their

v1.c8.s4.4

fancy carve the frames of their seats in different ways, and adorn them with precious substances.

Dress and Personal Characteristics

v1.c8.s5.1

The inner clothing and outward attire of the people have no tailoring; as to colour a fresh white is esteemed and motley is of no account. The men wind a strip of cloth round the waist and up to the armpits and leave the right shoulder bare. The women wear a long robe which covers both shoulders and falls down loose. The hair on the crown of the head is made into a coil, all the rest of the hair hanging down. Some clip their mustaches or have other fantastic fashions. Garlands are worn on the head and necklaces on the body.

v1.c8.s5.2

The names for their clothing materials are *Kiao-shê-ye* (Kausheya) and muslin (*tieh*) and calico (*pu*), Kausheya being silk from a wild silk-worm; *Ch'ii* (or *Ch'u*)-*mo* (Kshauma), a kind of linen; *Han* (or *Kan*)-*po-lo* (Kambala) a texture of fine wool (sheep's wool or goat's hair), and *Ho-la-li* (Ral?) a texture made from the wool of a wild animal--this wool being fine and soft and easily spun and woven is prized as a material for clothing. In North India where the climate is very cold closely fitting jackets are worn somewhat like those of the Tartars (*Hu*).

v1.c8.s5.3

The garbs of the non-Buddhists (religieux) are varied and extraordinary. Some wear peacocks' tails; some adorn themselves with a necklace of skulls; some are quite naked; some cover the body with grass or boards; some pull out their hair and clip their moustaches; some mat their side-hair and make a top-knot coil. Their clothing is not fixed and the colour varies.

v1.c8.s5.4

In this passage, it will be noticed, the clothing materials used by the lay people of India are arranged in four groups. The first is called by the pilgrim "Kausheya clothing and muslin and cloth" (僑奢耶衣及氈布等). Now kausheya (or kaus'eya) is silk made from the cocoon of the Bombyx Mori, and *tieh-pu* is cotton-cloth or *tieh* and cotton cloth. It is perhaps better to regard *tieh* and *pu* as names of two materials, and in another treatise we find Kausheya, *tieh*, and *ts'ui* (毳) grouped together.¹ This *ts'ui* was apparently a kind of coarse cotton cloth, and we find a *ts'ui-ka-pei* or "rough cotton" used to stuff cushions. The term kausheya was applied not only to silk stuffs but also to mixtures of silk and linen or cotton.¹ Our pilgrim evidently makes one group of "silk clothing" and cottons. This is not to be much wondered at when we reflect that he, like the other Chinese of his time and district, knew nothing of the cotton plant and the cloths derived from it. Moreover we should probably regard this description of the dress of the natives of India as derived from an earlier account.

v1.c8.s5.5

¹ Ta-fang-têng-ta-chi-ching, *ch*. 11 (No. 61 tr. cir. A.D. 400).

v1.c8.s5.6

¹ Sar. Vin. vibhāsha, ch. 5 (No. 1135): Sêng-chi-lü, ch. 9.

v1.c8.s5.7

The second kind of clothing material here mentioned is the Kshauma or Linen. This term also is to be regarded as denoting a class. It comprehends, we must suppose, the fabrics made from the Kshumā or flax, the śanā or jute, and the bhangā or hemp. These three plants are mentioned in Chinese translations from the Sanskrit as yielding stuffs from which clothes were made. This word kshauma denotes not only linen but also silk textures.

v1.c8.s5.8

The third group is the kambala. This word, which denotes "woollen cloth" and "a blanket", is here evidently used in the sense of fine woollen cloth for making clothing. Like the kausheya and the kshauma the kambala clothing was allowed to the Buddhist Brethren.

v1.c8.s5.9

The fourth kind of stuff mentioned as used for clothing material is called by Yuan-chuang Ho-la-li (褐刺納). There does not seem to be any known Sanskrit word with which this can be identified. As Yuan-chuang spells foreign words the three characters may stand for Ral, a Tibetan word meaning "goat's hair", from Ra, a goat. This Ho-la-li or Ral is also probably the Lo-i (羅衣) or "Lo (Ra) clothes" of other Buddhist texts. In Sanskrit also we find rallaka which denotes a wild animal and a stuff made from its hair, and rallaka-kambala which is a fine woollen cloth.

v1.c8.s5.10

Our pilgrim's description proceeds—

v1.c8.s5.11

The clerical costume of the Sha-mên (Śramaṇas) is only the three robes and the Sêng-kio-ki and Ni-p'o-so-na. As to the three robes the Schools adhere to different styles having broad or narrow fringes and small or large folds. The sêng-kio-ki goes over the left shoulder covers the armpits, joined on the right and opening on the left side and in length reaching to below the waist. As to the Ni-p'o-so-na, since no belt is worn when it is put on, it is gathered into plaits and secured by one of these, the size and colour of the plaits vary in the different schools.

v1.c8.s5.12

For the first part of this passage Julien has the following—"Les Cha-men (Çramanas) n'ont que trois sortes de vêtements, savoir le Sêng-kia-tchi (Saṅghāṭi) le Seng-kio-ki (Saṅkakchikā), et le Ni-po-sie-na (Nivāsana). La coupe et la façon de ces trois vêtements varient suivant les écoles. Les uns ont une bordure large ou étroite, les autres ont des pans petits ou grands". Here the translator spoils the description by interpolating the words "savoir le Sêng-kia-tchi (Saṅghāṭi)", leaving out the word "and", and inserting "ces" in the clause "La coupe et la façon de ces trois vêtements". The "Three robes" of the Buddhist monk are quite distinct from the two articles of his dress here mentioned by name. The "three robes" are always given as the Antaravāsaka, the Saṅghāṭi, and the Uttarāsaṅga. Of these we have already met with the second and third in our traveller's account of Balkh, and there we met also with the article of clothing called Seng-kio-ki. This last word is apparently for the original which is Samkachchhika in Pali and Julien's Sanskrit Saṅkakshikā. This is translated in a Chinese note to our text by "covering armpits". Professor Rhys Davids translates the Pali word by "vest", but the description given seems to suit a rude shirt or jacket with one sleeve which was buttoned or looped on the left shoulder. One name for the vestment as worn by monks in China is P'ien-shan (偏衫) or "one-sided jacket".¹ The other article of monk's costume mentioned by name here is the Ni-p'o-so-na or Nivāsana. This is rendered in Chinese by chün (裙) an old native term denoting a "skirt" on the lower part of a robe of ceremony. Nivāsana is a common term for an under-robe or lower garment, but it is here used in its restricted sense as designating the particular kind of skirt or under-robe worn by Buddhist monks. This was, according to regulation, four ells long by one and a half in width, and it reached from the waist to about three finger-breadths above the ankle. As Yuan-chuang here tells us the Schools were distinguished by differences in the wearing of the Nivāsana.¹ Thus, as I-ching tells us, the Sarvāstivādins wore the skirt with a pair of plaits turned out on both sides of it, and the Mahāsanghikas crossed the end of the right side to the top of the left side, tucking it in to keep the skirt in its place. This skirt or Nivāsa had no string or girdle and it was evidently something like the Malay Sarong which, as Colonel Yule tells us, is an old Indian form of dress. This garment also is self-securing, and is not in need of a belt or girdle. The two articles of dress here mentioned and described, viz. the Sañkakshika and the Nivāsana were in addition to the Three Robes which formed originally the full clerical costume of the bhikshu. They are often mentioned in the canonical books, having been allowed apparently as soon as Buddhism began to spread. The mode of wearing the Nivāsana and its colour and fashion caused much discussion and unpleasant feeling in the early church.

v1.c8.s5.13

¹ Shih-shih-yao-lan (釋氏要覽), ch. 1; Vinaya Vol. ii, p. 272 and Vinaya Texts Vol. iii, p. 351; Sêng-chi-lü l. c.

v1.c8.s5.14

¹ Ssŭ-fên-lü, ch. 19; Nan-hai-ch‘i-kuei, ch. 10, 11 and Takakusu; P‘i-ni-mu-ching, ch. 8.

v1.c8.s5.15

The pilgrim's description continues—

v1.c8.s5.16

The Kshatriyas and Brahmins are clean-handed and unostentatious, pure and simple in life and very frugal. The dress and ornaments of the kings and grandees are very extraordinary. Garlands and tiaras with precious stones are their head-adornments; and their bodies are adorned with rings, bracelets, and necklaces. Wealthy mercantile people have only bracelets. Most of the people go barefoot and shoes are rare. They stain their teeth red or black, wear their hair cut even, bore their ears, have long noses and large eyes; such are they in outward appearance.

v1.c8.s5.17

They are pure of themselves and not from compulsion. Before every meal they must have a wash; the fragments and remains are not served up again; the food utensils are not passed on; those utensils which are of pottery or wood must be thrown away after use, and those which are of gold, silver, copper, or iron get another polishing. As soon as a meal is over they chew the tooth-stick and make themselves clean; before they have finished ablutions they do not come into contact with each other; they always wash after urinating; they smear their bodies with scented unguents such as sandal and saffron. When the king goes to his bath there is the music of drums and stringed instruments and song; worship is performed and there are bathing and washing.

v1.c8.s5.18

The last sentence of this passage is in Julien's version—"Quand le roi se dispose à sortir, des musiciens battent le tambour et chantent aux sons de la guitare. Avant d'offrir un sacrifice, ou d'adresser des prières (aux dieux), ils se lavent et se baignent". Here Julien evidently had for the first clause the B reading *chün-wang-chiang-ts'ü*, meaning "when the king is about to go out". But in the A, C, and D texts the reading instead of *ts'ü* is *yii*, meaning "to bathe", and this is evidently the correct reading. Then Julien seems to change the author's meaning by making the second clause a new sentence and introducing the word "avant". The author's meaning seems to be that when the king took his bath there was the performance of certain acts of worship.¹

v1.c8.s5.19

¹ The text is—君王將浴 or 撾鼓奏弦歌祭祀拜訶沐浴盥洗.

Written and Spoken Language &c

v1.c8.s6.1

The description next proceeds to tell of the writing and learning of the Hindus.

v1.c8.s6.2

Their system of writing was invented, as is known, by the deva Brahmā who at the beginning instituted as patterns forty seven [written] words. These were combined and applied as objects arose and circumstances occurred; ramifying like streams they spread far and wide becoming modified a little by place and people. In language, speaking generally, they have not varied from the original source, but the people of "Mid India" are preeminently explicit and correct in speech, their expressions being harmonious and elegant, like those of the devas, and their intonation clear and distinct, serving as rule and pattern for others. The people of neighbouring territories and foreign countries repeating errors until these became the norm, and emulous for vulgarities, have lost the pure style.

v1.c8.s6.3

The statement here made to the effect that the Sanskrit alphabet was invented by the god Brahmā is repeated in several other books by Buddhist writers. Some tell us that Brahmā was once a rājah on earth, and that he then invented an alphabet of 72 letters called the “Kharu writing” (佉樓書). Disgusted with the bad treatment given to these letters he proceeded to swallow them all; but two, *a* and *au* escaped from his mouth and remained among men.¹ But we are also told that Brahmā invented the Brahma writing first, and that afterwards Kharoshtha produced the script which bears his name.² Another account represents the Brahman writing (or Devanagari) to have been the invention of a wise (kovida) Brahmin, and the Kharu writing to have been the work of a stupid (kharu) rishi. This Kharu writing is that mentioned in the Lalitavistara and other books under the name Kharoshtha (or Kharosta). This word is translated by “Ass-ear”, and is the name of an ancient rishi who was a great astronomer and astrologer. In some Buddhist treatises we find the invention of letters ascribed to the Buddha, and in some Siva, as in Indian tradition, is credited with the first teaching of spelling and writing.³ The “forty seven words” of our passage are the twelve symbols which represent the ten vowels, and anusvāra and visarga, and the thirty-five consonants; and so constitute the alphabet. The letters admit of endless combinations to make words as objects require names and circumstances need expression. Some authors give the number of the letters in the Sanskrit alphabet as less, and others as more, than the number here given; but this is generally regarded as the correct number. With the statements here made by Yuan-chuang about the Sanskrit alphabet and language we may compare the more detailed account given in the third chuan of the Life.¹

v1.c8.s6.4

¹ Pai-lun-su (百論疏).

v1.c8.s6.5

² Liu-shu-liao (六書略), ch. 5.

v1.c8.s6.6

³ See Ta-pan-nie-p'an-ching, ch. 8 (No. 113); Si-t'an-san-mi-ch'ao (悉曇三密鈔), ch. 1; Si-t'an-tsang, ch. 1; Si-t'an-tzŭ-chi (字記).

v1.c8.s6.7

¹ Julien I, p. 165; cf. Alberuni Vol. i, p. 170; Bühler's Ind. Palæographie pˢ. 1 and 19 to 30.

v1.c8.s6.8

It is evident that Yuan-chuang, like other non-Indian Buddhists, had been taught to regard the spoken and written language of "Mid-India" as at once the parent and the standard of all the dialects of "North-India". These latter had departed a little from the correct form in their writing, some of them, as in Gandhāra, having written alphabets so unlike the parent one that they had special names. In oral speech the border lands and outlying regions generally had come to differ much from the people of "Mid-India". They had lost the rich purity of the standard language, and had persisted in erroneous forms of expression until these had come to be taught as the rule.

v1.c8.s6.9

The description continues—

v1.c8.s6.10

As to their archives and records there are separate custodians of these. The official annals and state-papers are called collectively *ni-lo-pi-t'u* (or *ch'a*); in these good and bad are recorded, and instances of public calamity and good fortune are set forth in detail.

v1.c8.s6.11

The *Ni-lo-pi-t'u* of this passage has been rightly restored by Julien as Nilapiṭa, and the Chinese annotator tells us the word means "Dark-blue store". We find the word Nilapiṭa in our Sanskrit dictionaries, but the P. W. gives only one illustration of its use, and that is the passage before us.

v1.c8.s6.12

Proceeding to the education and learning of the people of India our author writes—

v1.c8.s6.13

In beginning the education of their children and winning them on to progress they follow the "Twelve Chapters". When the children are seven years of age the great treatises of the Five Sciences are gradually communicated to them. The first science is Grammar which teaches and explains words, and classifies their distinctions. The second is that of the skilled professions [concerned with] the principles of the mechanical arts, the dual processes, and astrology. The third is the science of medicine [embracing] exorcising charms, medicine, the use of the stone, the needle, moxa. The fourth is the science of reasoning, by which the orthodox and heterodox are ascertained, and the true and false are thoroughly sought out. The fifth is the science of the Internal which investigates and teaches the five degrees of religious attainments (lit. the “five vehicles”) and the subtle doctrine of karma.

v1.c8.s6.14

The “Twelve Chapters” of this passage is in the original *Shi-êrh-chang* (十二章) and Julien translates this by “un livre en douze sections”. In a note to this rendering he translates a short passage from a well-known Buddhist Dictionary about a book called the “Siddham-chang”. This is doubtless the sort of work to which the pilgrim refers as the first book which the children of India learned. The name is made up of *Siddham* which means, we are told, “Perfection” or “May good fortune be attained”, and *chang* the Chinese word for a “section” or “chapter”. But Julien makes the whole stand for a Sanskrit compound *Siddhavastu*, a term apparently known only from his use of it. From a passage in I-ching’s “Nan-hai-ch'i-kuei” and from other works we learn that the Siddham-chang was the name of a child’s primer ABC., the first chapter of which was headed by the word *siddham*.¹ This word forms an “auspicious invocation”, and the Buddhists used it alone or with “Namo Sarvajnāya, “Praise to the omniscient [Buddha]” prefixed, at the beginning of their primers. They used it in a similar way to head such documents as deeds of gift to religious establishments. In these places Bühler took the word to mean “Success”, i. e. may there be success, an interpretation which agrees with the accepted Chinese rendering. But Fleet thinks that *siddham* in these places is to be understood as meaning “Perfection has been attained by Buddha”, an interpretation which does not seem to be so good.¹ Instead of siddham the non-Buddhist teachers in India placed "siddhir-astu" meaning "May there be success (or accomplishment)" at the head of their ABCs. Thus these books came to be called Siddham or Siddhir-astu, the former being the name by which they became known to the Chinese. There are many varieties of them and the number of chapters or sections ranges from nine to eighteen, the latter being the number in the work which may be regarded as the standard one in China. This is the Si-tan-tsü-chi (悉曇字記) by the monk Chih-kuang (智廣) of the T'ang period taken from the Siddham of Prajnābodhi of South India. A Siddham gives the Sanskrit alphabet, beginning with the vowels and proceeding in the order in which the letters are given in our Sanskrit grammars, then the combinations made by single consonants and vowels, and then those made by two or more consonants with a vowel. In some of the Siddhams made for Chinese use we are told that this word denotes "the alphabet", while in others we are told that it is a designation for the twelve so-called vowels, but the statements are not borne out by any authority, and are evidently not correct. It may be interesting, however, in connection with subject to quote a statement from Alberuni. He relates—"The most generally known alphabet is called Siddha-mātṛkā, which is by some considered as originating from Kashmir, for the people of Kashmir use it. But it is also used in Vārāṇasi. This town and Kashmir are the high schools of Hindu science. The same writing is used in Madhyadeśa, i. e. the middle country, the country all around Kanauj, which is also called Aryāvarta".² According to I-ching a child began his primer when he was six years of age and learned it within six months.

First 94 of 195 passages. Read on

Linked texts (18)