CHAPTER XXXIX

CHAMPA[329]

THE kingdom of Champa, though a considerable power from about the third century until the end of the fifteenth, has attracted less attention than Camboja or Java. Its name is a thing of the past and known only to students: its monuments are inferior in size and artistic merit to those of the other Hindu kingdoms in the Far East and perhaps its chief interest is that it furnishes the oldest Sanskrit inscription yet known from these regions.

Champa occupied the south-eastern corner of Asia beyond the Malay Peninsula, if the word corner can be properly applied to such rounded outlines. Its extent varied at different epochs, but it may be roughly defined in the language of modern geography as the southern portion of Annam, comprising the provinces of Quãng-nam in the north and Bînh-Thuan in the south with the intervening country. It was divided into three provinces, which respectively became the seat of empire at different periods. They were (i) in the north Amarâvatî (the modern Quãng-nam) with the towns of Indrapura and Sinhapura; (ii) in the middle Vijaya (the modern Bing-Dinh) with the town of Vijaya and the port of Śrî-Vinaya; (iii) in the south Pâṇḍurânga or Panran (the modern provinces of Phanrang and Binh-Thuan) with the town of Vîrapura or Râjapura. A section of Pâṇḍurânga called Kauthâra (the modern Kanh hoa) was a separate province at certain times. Like the modern Annam, Champa appears to have been mainly a littoral kingdom and not to have extended far into the mountains of the interior.

Champa was the ancient name of a town in western Bengal near Bhagalpur, but its application to these regions does not seem due to any connection with north-eastern India. The conquerors of the country, who were called Chams, had a certain amount of Indian culture and considered the classical name Champa as an elegant expression for the land of the Chams. Judging by their language these Chams belonged to the Malay-Polynesian group and their distribution along the littoral suggests that they were invaders from the sea like the Malay pirates from whom they themselves subsequently suffered. The earliest inscription in the Cham language dates from the beginning of the ninth century but it is preceded by a long series of Sanskrit inscriptions the oldest of which, that of Vo-can[330], is attributed at latest to the third century, and refers to an earlier king. It therefore seems probable that the Hindu dynasty of Chaṃpa was founded between 150 and 200 A.D. but there is no evidence to show whether a Malay race already settled in Champa was conquered and hinduized by Indian invaders, or whether the Chams were already hinduized when they arrived, possibly from Java.

The inferiority of the Chams to the Khmers in civilization was the result of their more troubled history. Both countries had to contend against the same difficulty--a powerful and aggressive neighbour on either side. Camboja between Siam and Annam in 1800 was in very much the same position as Champa had been between Camboja and Annam five hundred years earlier. But between 950 and 1150 A.D. when Champa by no means enjoyed stability and peace, the history of Camboja, if not altogether tranquil, at least records several long reigns of powerful kings who were able to embellish their capital and assure its security. The Chams were exposed to attacks not only from Annam but also from the more formidable if distant Chinese and their capital, instead of remaining stationary through several centuries like Angkor Thom, was frequently moved as one or other of the three provinces became more important.

The inscription of Vo-can is in correct Sanskrit prose and contains a fragmentary address from a king who seems to have been a Buddhist and writes somewhat in the style of Asoka. He boasts that he is of the family of Śrîmârarâja. The letters closely resemble those of Rudradaman's inscription at Girnar and contemporary inscriptions at Kanheri. The text is much mutilated so that we know neither the name of the writer nor his relationship to Śrîmâra. But the latter was evidently the founder of the dynasty and may have been separated from his descendant by several generations. It is noticeable that his name does not end in Varman, like those of later kings. If he lived at the end of the second century this would harmonize with the oldest Chinese notices which fix the rise of Lin-I (their name for Champa) about 192 A.D.[331] Agreeably to this we also hear that Hun T'ien founded an Indian kingdom in Fu-nan considerably before 265 A.D. and that some time between 220 and 280 a king of Fu-nan sent an embassy to India. The name Fu-nan may include Champa. But though we hear of Hindu kingdoms in these districts at an early date we know nothing of their civilization or history, nor do we obtain much information from those Cham legends which represent the dynasties of Champa as descended from two clans, those of the cabbage palm (aréquier) and cocoanut.

Chinese sources also state that a king called Fan-yi sent an embassy to China in 284 and give the names of several kings who reigned between 336 and 440. One of these, Fan-hu-ta, is apparently the Bhadravarman who has left some Sanskrit inscriptions dating from about 400 and who built the first temple at Mĩ-so'n. This became the national sanctuary of Champa: it was burnt down about 575 A.D. but rebuilt. Bhadravarman's son Gangarâja appears to have abdicated and to have gone on a pilgrimage to the Ganges[332]--another instance of the intercourse prevailing between these regions and India.

It would be useless to follow in detail the long chronicle of the kings of Champa but a few events merit mention. In 446 and again in 605 the Chinese invaded the country and severely chastised the inhabitants. But the second invasion was followed by a period of peace and prosperity. Śambhuvarman (†629) restored the temples of Mi-so'n and two of his successors, both called Vikrântavarman, were also great builders. The kings who reigned from 758 to 859, reckoned as the fifth dynasty, belonged to the south and had their capital at Vîrapura. The change seems to have been important, for the Chinese who had previously called the country Lin-I, henceforth call it Huan-wang. The natives continued to use the name Champa but Satyavarman and the other kings of the dynasty do not mention Mi-so'n though they adorned and endowed Po-nagar and other sanctuaries in the south. It was during this period (A.D. 774 and 787) that the province of Kauthâra was invaded by pirates, described as thin black barbarians and cannibals, and also as the armies of Java[333]. They pillaged the temples but were eventually expelled. They were probably Malays but it is difficult to believe that the Javanese could be seriously accused of cannibalism at this period[334].

The capital continued to be transferred under subsequent dynasties. Under the sixth (860-900) it was at Indrapura in the north: under the seventh (900-986) it returned to the south: under the eighth (989-1044) it was in Vijaya, the central province. These internal changes were accompanied by foreign attacks. The Khmers invaded the southern province in 945. On the north an Annamite Prince founded the kingdom of Dai-côviêt, which became a thorn in the side of Champa. In 982 its armies destroyed Indrapura, and in 1044 they captured Vijaya. In 1069 King Rudravarman was taken prisoner but was released in return for the cession of the three northernmost provinces. Indrapura however was rebuilt and for a time successful wars were waged against Camboja, but though the kings of Champa did not acquiesce in the loss of the northern provinces, and though Harivarman III (1074-80) was temporarily victorious, no real progress was made in the contest with Annam, whither the Chams had to send embassies practically admitting that they were a vassal state. In the next century further disastrous quarrels with Camboja ensued and in 1192 Champa was split into two kingdoms, Vijaya in the north under a Cambojan prince and Panran in the south governed by a Cham prince but under the suzerainty of Camboja. This arrangement was not successful and after much fighting Champa became a Khmer province though a very unruly one from 1203 till 1220. Subsequently the aggressive vigour of the Khmers was tempered by their own wars with Siam. But it was not the fate of Champa to be left in peace. The invasion of Khubilai lasted from 1278 to 1285 and in 1306 the provinces of O and Ly were ceded to Annam.

Champa now became for practical purposes an Annamite province and in 1318 the king fled to Java for refuge. This connection with Java is interesting and there are other instances of it. King Jaya Simhavarman III († 1307) of Champa married a Javanese princess called Tapasi. Later we hear in Javanese records that in the fifteenth century the princess Darawati of Champa married the king of Madjapahit and her sister married Raden Radmat, a prominent Moslim teacher in Java[335].

The power of the Chams was crushed by Annam in 1470. After this date they had little political importance but continued to exist as a nationality under their own rulers. In 1650 they revolted against Annam without success and the king was captured. But his widow was accorded a titular position and the Cham chronicle[336] continues the list of nominal kings down to 1822.

In Champa, as in Camboja, no books dating from the Hindu period have been preserved and probably there were not many. The Cham language appears not to have been used for literary purposes and whatever culture existed was exclusively Sanskrit. The kings are credited with an extensive knowledge of Sanskrit literature. An inscription at Po-nagar[337] (918 A.D.) says that Śrî Indravarman was acquainted with the Mîmâṃsâ and other systems of philosophy, Jinendra, and grammar together with the Kâśikâ (vṛitti) and the Śaivottara-Kalpa. Again an inscription of Mi-son[338] ascribes to Jaya Indravarmadeva (_c._ 1175 A.D.) proficiency in all the sciences as well as a knowledge of the Mahâyâna and the Dharmaśâstras, particularly the Nâradîya and Bhârgavîya. To some extent original compositions in Sanskrit must have been produced, for several of the inscriptions are of considerable length and one[339] gives a quotation from a work called the Purâṇârtha or Arthapurâṇaśâstra which appears to have been a chronicle of Champa. But the language of the inscriptions is often careless and incorrect and indicates that the study of Sanskrit was less flourishing than in Camboja.

2

The monuments of Champa, though considerable in size and number, are inferior to those of Camboja. The individual buildings are smaller and simpler and the groups into which they are combined lack unity. Brick was the chief material, stone being used only when brick would not serve, as for statues and lintels. The commonest type of edifice is a square pyramidal structure called by the Chams Kalan. A Kalan is as a rule erected on a hill or rising ground: its lowest storey has on the east a porch and vestibule, on the other three sides false doors. The same shape is repeated in four upper storeys of decreasing size which however serve merely for external decoration and correspond to nothing in the interior. This is a single windowless pyramidal cell lighted by the door and probably also by lamps placed in niches on the inner walls. In the centre stood a pedestal for a linga or an image, with a channel to carry off libations, leading to a spout in the wall. The outline of the tower is often varied by projecting figures or ornaments, but the sculpture is less lavish than in Camboja and Java.

In the greater religious sites several structures are grouped together. A square wall surrounds an enclosure entered by a gateway and containing one or more Kalans, as well as smaller buildings, probably for the use of priests. Before the gateway there is frequently a hall supported by columns but open at the sides.

All known specimens of Cham architecture are temples; palaces and other secular buildings were made of wood and have disappeared. Of the many sanctuaries which have been discovered, the most remarkable are those of Mi-son, and Dong Duong, both in the neighbourhood of Tourane, and Po Nagar close to Nhatrang.

Mi-son[340] is an undulating amphitheatre among mountains and contains eight or nine groups of temples, founded at different times. The earliest structures, erected by Bhadravarman I about 400, have disappeared[341] and were probably of wood, since we hear that they were burnt (apparently by an accident) in 575 A.D. New temples were constructed by Śambhuvarman about twenty-five years later and were dedicated to Śambhu-bhadreśvara, in which title the names of the founder, restorer and the deity are combined. These buildings, of which portions remain, represent the oldest and best period of Cham art. Another style begins under Vikrântavarman I between 657 and 679 A.D. This reign marks a period of decadence and though several buildings were erected at Mi-son during the eighth and ninth centuries, the locality was comparatively neglected[342] until the reign of Harivarman III (1074-1080). The temples had been ravaged by the Annamites but this king, being a successful warrior, was able to restore them and dedicated to them the booty which he had captured. Though his reign marks a period of temporary prosperity in the annals of Champa, the style which he inaugurated in architecture has little originality. It reverts to the ancient forms but shows conscious archaism rather than fresh vigour. The position of Mi-son, however, did not decline and about 1155 Jaya Harivarman I repaired the buildings, dedicated the booty taken in battle and erected a new temple in fulfilment of a vow. But after this period the princes of Champa had no authority in the district of Mi-son, and the Annamites, who seem to have disliked the religion of the Chams, plundered the temples.

Po-nagar[343] is near the port of Nha-trang and overlooks the sea. Being smaller that Mi-son it has more unity but still shows little attempt to combine in one architectural whole the buildings of which it is composed.

An inscription[344] states with curious precision that the shrine was first erected in the year 5911 of the Dvâpara age and this fantastic chronology shows that in our tenth century it was regarded as ancient. As at Mi-son, the original buildings were probably of wood for in 774 they were sacked and burnt by pirates who carried off the image[345]. Shortly afterwards they were rebuilt in brick by King Satyavarman and the existing southern tower probably dates from his reign, but the great central tower was built by Harivarman I (817 A.D.) and the other edifices are later.

Po Nagar or Yang Po Nagar means the Lady or Goddess of the city. She was commonly called Bhagavatî in Sanskrit[346] and appears to have been the chief object of worship at Nha-trang, although Śiva was associated with her under the name of Bhagavatîśvara. In 1050 an ardhanarî image representing Śiva and Bhagavatî combined in one figure was presented to the temple by King Parameśvara and a dedicatory inscription describes this double deity as the cosmic principle.

When Champa was finally conquered the temple was sold to the Annamites, who admitted that they could not acquire it except by some special and peaceful arrangement. Even now they still continue the worship of the goddess though they no longer know who she is[347].

Dong Duong, about twenty kilometres to the south of Mi-son, marks the site of the ancient capital Indrapura. The monument which has made its name known differs from those already described. Compared with them it has some pretensions to be a whole, laid out on a definite plan and it is Buddhist. It consists of three courts[348] surrounded by walls and entered by massive porticoes. In the third there are about twenty buildings and perhaps it did not escape the fault common to Cham architecture of presenting a collection of disconnected and unrelated edifices, but still there is clearly an attempt to lead up from the outermost portico through halls and gateways to the principal shrine. From an inscription dated 875 A.D. we learn that the ruins are those of a temple and vihâra erected by King Indravarman and dedicated to Avalokita under the name of Lakshmîndra Lokeśvara.

3

The religion of Champa was practically identical with that of Camboja. If the inscriptions of the former tell us more about mukhalingas and koshas and those of the latter have more allusions to the worship of the compound deity Hari-hara, this is probably a matter of chance. But even supposing that different cults were specially prominent at different places, it seems clear that all the gods and ceremonies known in Camboja were also known in Champa and _vice versa_. In both countries the national religion was Hinduism, mainly of the Śivaite type, accompanied by Mahayanist Buddhism which occasionally came to the front under royal patronage. In both any indigenous beliefs which may have existed did not form a separate system. It is probable however that the goddess known at Po-nagar as Bhagavatî was an ancient local deity worshipped before the Hindu immigration and an inscription found at Mi-son recommends those whose eyes are diseased to propitiate Kuvera and thus secure protection against Ekâkshapingalâ, "the tawny one-eyed (spirit)." Though this goddess or demon was probably a creation of local fancy, similar identifications of Kâlî with the spirits presiding over cholera, smallpox, etc., take place in India.

The social system was theoretically based on the four castes, but Chinese accounts indicate that in questions of marriage and inheritance older ideas connected with matriarchy and a division into clans still had weight. But the language of the inscriptions is most orthodox. King Vikrântavarman[349] quotes with approval the saying that the horse sacrifice is the best of good deeds and the murder of a Brahman the worst of sins. Brahmans, chaplains (purohita), pandits and ascetics are frequently mentioned as worthy of honour and gifts. The high priest or royal chaplain is styled Śrîparamapurohita but it does not appear that there was a sacerdotal family enjoying the unique position held by the Śivakaivalyas in Camboja. The frequent changes of capital and dynasty in Champa were unfavourable to continuity in either Church or State.

Śivaism, without any hostility to Vishṇuism or Buddhism, was the dominant creed. The earliest known inscription, that of Vo-can, contains indications of Buddhism, but three others believed to date from about 400 A.D. invoke Śiva under some such title as Bhadreśvara, indicating that a temple had been dedicated to him by King Bhadravarman. Thus the practice of combining the names of a king and his patron deity in one appellation existed in Champa at this early date[350]. It is also recorded from southern India, Camboja and Java. Besides Śiva one of the inscriptions venerates, though in a rather perfunctory manner, Umâ, Brahmâ, Vishṇu and the five elements. Several inscriptions[351] give details of Śivaite theology which agree with what we know of it in Camboja. The world animate and inanimate is an emanation from Śiva, but he delivers from the world those who think of him. Meditation, the practice of Yoga, and devotion to Śiva are several times mentioned with approval[352]. He abides in eight forms corresponding to his eight names Śarva, Bhava, Paśupati, Iśâna, Bhîma, Rudra, Mahâdeva, and Ugra. He is also, as in Java, Guru or the teacher and he has the usual mythological epithets. He dances in lonely places, he rides on the bull Nandi, is the slayer of Kâma, etc. Though represented by figures embodying such legends he was most commonly adored under the form of the linga which in Champa more than elsewhere came to be regarded as not merely symbolic but as a personal god. To mark this individuality it was commonly enclosed in a metal case (kosha) bearing one or more human faces[353]. It was then called mukhalinga and the faces were probably intended as portraits of royal donors, identified with the god in form as well as in name. An inscription of 1163 A.D. records the dedication of such a kosha, adorned with five royal faces, to Śrîśânabhadreśvara. The god, it is said, will now be able to give his blessing to all regions through his five mouths which he could not do before, and being enclosed in the kosha, like an embryo in the matrix, he becomes Hiraṇyagarbha. The linga, with or without these ornaments, was set on a _snânadroṇi_ or stone table arranged for receiving libations, and sometimes (as in Java and Camboja) four or more lingas were set upon a single slab. From A.D. 400 onwards, the cult of Śiva seems to have maintained its paramount position during the whole history of Champa, for the last recorded Sanskrit inscription is dedicated to him. From first to last it was the state religion. Śiva is said to have sent Uroja to be the first king and is even styled the root of the state of Champa.

An inscription[354] of 811 A.D. celebrates the dual deity Śankara-Nârâyaṇa. It is noticeable that Nârâyaṇa is said to have held up Mt. Govardhana and is apparently identified with Kṛishṇa. Râma and Kṛishṇa are both mentioned in an inscription of 1157 which states that the whole divinity of Vishṇu was incarnate in King Jaya Harivarman I[355]. But neither allusions to Vishṇu nor figures of him[356] are numerous and he plays the part of an accessory though respected personage. Garuḍa, on whom he rides, was better known than the god himself and is frequently represented in sculpture.

The Śakti of Śiva, amalgamated as mentioned with a native goddess, received great honour (especially at Nhatrang) under the names of Umâ, Bhagavatî, the Lady of the city (Yang Po Nagar) and the goddess of Kauthâra. In another form or aspect she was called Maladâkuṭhâra.[357] There was also a temple of Ganeśa (Śri-Vinâyaka) at Nhatrang but statues of this deity and of Skanda are rare.

The Chinese pilgrim I-Ching, writing in the last year of the seventh century, includes Champa (Lin-I) in the list of countries which "greatly reverence the three jewels" and contrasts it with Fu-nan where a wicked king had recently almost exterminated Buddhism. He says "In this country Buddhists generally belong to the Arya-sammiti school, and there are also a few followers of the Aryasarvâstivâdin school." The statement is remarkable, for he also tells us that the Sarvâstivâdins were the predominant sect in the Malay Archipelago and flourished in southern China. The headquarters of the Sammitîyas were, according to the accounts of both Hsüan Chuang and I-Ching, in western India though, like the three other schools, they were also found in Magadha and eastern India. We also hear that the brother and sister of the Emperor Harsha belonged to this sect and it was probably influential. How it spread to Champa we do not know, nor do the inscriptions mention its name or indicate that the Buddhism which they knew was anything but the mixture of the Mahayana with Śivaism[358] which prevailed in Camboja.

I-Ching's statements can hardly be interpreted to mean that Buddhism was the official religion of Champa at any rate after 400 A.D., for the inscriptions abundantly prove that the Śivaite shrines of Mi-son and Po-nagar were so to speak national cathedrals where the kings worshipped on behalf of the country. But the Vo-can inscription (? 250 A.D.), though it does not mention Buddhism, appears to be Buddhist, and it would be quite natural that a dynasty founded about 150 A.D. should be Buddhist but that intercourse with Camboja and probably with India should strengthen Śivaism. The Chinese annals mention[359] that 1350 Buddhist books were carried off during a Chinese invasion in 605 A.D. and this allusion implies the existence of Buddhism and monasteries with libraries. As in Camboja it was perhaps followed by ministers rather than by kings. An inscription found[360] in southern Champa and dated as 829 A.D. records how a sthavira named Buddhanirvâṇa erected two vihâras and two temples (devakula) to Jina and Śankara (Buddha and Śiva) in honour of his deceased father. Shortly afterwards there came to the throne Indravarman II (860-890 A.D.), the only king of Champa who is known to have been a fervent Buddhist. He did not fail to honour Śiva as the patron of his kingdom but like Asoka he was an enthusiast for the Dharma[361]. He desires the knowledge of the Dharma: he builds monasteries for the sake of the Dharma: he wishes to propagate it: he even says that the king of the gods governs heaven by the principles of Dharma. He wishes to lead all his subjects to the "yoke and abode of Buddha," to "the city of deliverance."

To this end he founded the vihâra of Dong Duong, already described, and dedicated it to Śri Lakshmîndra Lokeśvara[362]. This last word is a synonym of Avalokita, which also occurs in the dedicatory inscription but in a fragmentary passage. Lakshmîndra is explained by other passages in the inscription from which we learn that the king's name before he ascended the throne was Lakshmîndra Bhûmîśvara, so that the Bodhisattva is here adored under the name of the king who erected the vihâra according to the custom prevalent in Śivaite temples. Like those temples this vihâra received an endowment of land and slaves of both sexes, as well as gold, silver and other metals[363].

A king who reigned from 1080 to 1086 was called Paramabodhisattva, but no further epigraphic records of Buddhism are known until the reigns of Jaya Indravarmadeva (1167-1192) and his successor Sûryavarmadeva[364]. Both of these monarchs, while worshipping Śiva, are described as knowing or practising the jñâna or dharma of the Mahayana. Little emphasis seems to be laid on these expressions but still they imply that the Mahayana was respected and considered part of the royal religion. Sûryavarmadeva erected a building called Śrî Herukaharmya[365]. The title is interesting for it contains the name of the Tantric Buddha Heruka.

The grotto of Phong-nha[366] in the extreme north of Champa (province of Quang Binh) must have been a Buddhist shrine. Numerous medallions in clay bearing representations of Buddhas, Bodhisattvas and Dagobas have been found there but dates are wanting.

It does not appear that the Hinayanist influence which became predominant in Camboja extended to Champa. That influence came from Siam and before it had time to traverse Camboja, Champa was already in the grip of the Annamites, whose religion with the rest of their civilization came from China rather than India. Chinese culture and writing spread to the Cambojan frontier and after the decay of Champa, Camboja marks the permanent limit within which an Indian alphabet and a form of Buddhism not derived through China have maintained themselves.

A large number of the Chams were converted to Mohammedanism but the time and circumstances of the event are unknown. When Friar Gabriel visited the country at the end of the sixteenth century a form of Hinduism seems to have been still prevalent[367]. It would be of interest to know how the change of religion was effected, for history repeats itself and it is likely that the Moslims arrived in Champa by the route followed centuries before by the Hindu invaders.

There are still about 130,000 Chams in the south of Annam and Camboja. In the latter country they are all Mohammedans. In Annam some traces of Hinduism remain, such as mantras in broken Sanskrit and hereditary priests called Baśaih. Both religions have become unusually corrupt but are interesting as showing how beliefs which are radically distinct become distorted and combined in Eastern Asia[368].

FOOTNOTES:

[Footnote 329: Also spelt Campâ and Tchampa. It seems safer to use Ch for C in names which though of Indian origin are used outside India. The final _a_ though strictly speaking long is usually written without an accent. The following are the principal works which I have consulted about Champa.

(a) G. Maspéro, _Le Royaume de Champa_. Published in _T'oung Pao_, 1910-1912. Cited as Maspéro.

(b) A. Bergaigne, "Inscriptions Sanskrites de Champa" in _Notices et Extraits des Manuscrits de la Bibliothèque Nationale_, tome XXVII. 1^re partie. 2^e fascicule, 1893, pp. 181-292. Cited as _Corpus_, II.

(c) H. Parmentier, _Inventaire descriptif des Monuments Ćams de l'Annam_. 1899.

(d) L. Finot, "La Religion des Chams," _B.E.F.E.O_, 1901, and _Notes d'Epigraphie_. "Les Inscriptions de Mi-son," _ib_. 1904. Numerous other papers by this author, Durand, Parmentier and others in the same periodical can be consulted with advantage.

(e) _Id., Notes d'Epigraphie Indo-Chinoise_, 1916.]

[Footnote 330: _Corpus_, II. p. 11, and Finot, _Notes d'Epig._ pp. 227 ff.]

[Footnote 331: See authorities quoted by Maspéro, _T'oung Pao_, 1910, p. 329.]

[Footnote 332: Finot in _B.E.F.E.O._ 1904, pp. 918 and 922.]

[Footnote 333: _Corpus_, II. _Stêle de Po Nagar_, pp. 252 ff. and _Stêle de Yang Tikuh_, p. 208, etc.]

[Footnote 334: The statements that they came from Java and were cannibals occur in different inscriptions and may conceivably refer to two bodies of invaders. But the dates are very near. Probably Java is not the island now so called. See the chapter on Camboja, sec. 2. The undoubted references in the inscriptions of Champa to the island of Java call it Yavadvîpa.]

[Footnote 335: _Veth. Java_, I. p. 233.]

[Footnote 336: See "La Chronique Royale," _B.E.F.E.O._ 1905, p. 377.]

[Footnote 337: _Corpus_, II. p. 259. Jinendra may be a name either of the Buddha or of a grammarian. The mention of the Kâśikâ vṛitti is important as showing that this work must be anterior to the ninth century. The Uttara Kalpa is quoted in the Tantras (see Bergaigne's note), but nothing is known of it.]

[Footnote 338: _B.E.F.E.O._ 1904, p. 973.]

[Footnote 339: From Mi-son, date 1157 A.D. See _B.E.F.E.O._ 1904, pp. 961 and 963.]

[Footnote 340: = Chinese Mei shan, beautiful mountain. For an account of the temples and their history see the articles by Parmentier and Finot, _B.E.F.E.O._ 1904, pp. 805-977.]

[Footnote 341: But contemporary inscriptions have been discovered. _B.E.F.E.O._ 1902, pp. 185 ff.]

[Footnote 342: Doubtless because the capital was transferred to the south where the shrine of Po-nagar had rival claims.]

[Footnote 343: See especially the article by Parmentier, _B.E.F.E.O._ 1902, pp. 17-54.]

[Footnote 344: XXVI _Corpus_, II. pp. 244, 256; date 918 A.D.]

[Footnote 345: Śivamukham: probably a mukhalinga.]

[Footnote 346: Also Yäpunagara even in Sanskrit inscriptions.]

[Footnote 347: Parmentier, _l.c._ p. 49.]

[Footnote 348: This is only a very rough description of a rather complicated structure. For details see Parmentier, _Monuments C̆ams_, planche XCVIII.]

[Footnote 349: Inscrip. at Mi-son of 658 A.D. See _B.E.F.E.O._ 1904, p. 921.]

[Footnote 350: Other examples are Indrabhadreśvara, _Corpus_, II. p. 208. Harivarmeśvara, _B.E.F.E.O._ 1904, p. 961.]

[Footnote 351: _E.g. B.E.F.E.O._ pp. 918 ff. Dates 658 A.D. onwards.]

[Footnote 352: Yogaddhyâna, Śivârâdha, Śivabhakti. See _B.E.F.E.O._ 1904, pp. 933-950. Harivarman III abdicated in 1080 and gave himself up to contemplation and devotion to Śiva.]

[Footnote 353: See _B.E.F.E.O._ 1904, pp. 912 ff. and esp. p. 970. I have seen a kosha which is still in use in the neighbourhood of Badami. It is kept in a village called Nandikeśvara, but on certain festivals it is put on a linga at the temple of Mahakut. It is about 2 feet high and 10 inches broad; a silver case with a rounded and ornamented top. On one side is a single face in bold embossed work and bearing fine moustaches exactly as in the mukhalingas of Champa. In the tank of the temple of Mahakut is a half submerged shrine, from which rises a stone linga on which are carved four faces bearing moustaches. There is said to be a gold kosha set with jewels at Śringeri. See _J. Mythic. Society_ (Bangalore), vol. VIII. p. 27. According to Gopinatha Rao, _Indian Iconography_, vol. II. p. 63, the oldest known lingas have figures carved on them.]

[Footnote 354: _Corpus_, II. pp. 229, 230.]

[Footnote 355: _B.E.F.E.O._ 1904, pp. 959, 960.]

[Footnote 356: See for an account of same _B.E.F.E.O._ 1901, p. 18.]

[Footnote 357: _Corpus_, II. p. 282.]

[Footnote 358: In several passages Hsüan Chuang notes that there were Pâśupatas or other Śivaites in the same towns of India where Sammitiyas were found. See Watters, _Yüan Chwang_, I. 331, 333; II. 47, 242, 256, 258, 259.]

[Footnote 359: Maspéro, _T'oung Pao_, 1910, p. 514.]

[Footnote 360: At Yang Kur. See _Corpus_, II. pp. 237-241.]

[Footnote 361: For his views see his inscriptions in _B.E.F.E.O._ 1904, pp. 85 ff. But kings who are not known to have been Buddhists also speak of Dharma. _B.E.F.E.O._ 1904, pp. 922, 945.]

[Footnote 362: Apparently special forms of deities such as Śrîśânabhadreśvara or Lakshmînda Lokeśvara were regarded as to some extent separate existences. Thus the former is called a portion of Śiva, _B.E.F.E.O._ 1904, p. 973.]

[Footnote 363: Presumably in the form of vessels.]

[Footnote 364: _B.E.F.E.O._ 1904, pp. 973-975.]

[Footnote 365: _B.E.F.E.O._ 1904, p. 975.]

[Footnote 366: _Ib._ 1901, p. 23, and Parmentier, _Inventaire des Monuments Chams_, p. 542.]

[Footnote 367: Gabriel de San Antonio, _Breve y verdadera relation de los successes de Reyno de Camboxa_, 1604.]

[Footnote 368: See for the modern Chams the article "Chams" in _E.R.E. and Ethics_, and Durand, "Les Chams Bani," _B.E.F.E.O._ 1903, and "Notes sur les Chams," _ib._ 1905-7.]