city. To the north-east, distant about one mile from the fort, there is a large village, on another mound, called Madiya; and between the two lies a large tank called Kúnda Tál, surrounded by numerous small mounds which are said to be the remains of buildings. Originally these two places would appear to have formed one large town about 11 mile in length by half a mile in breadth, or 3 miles in circuit. The Kanungo states that Madawar formed part of the dominions of Pithora Raja, and that it possessed a large Hindu temple of stone, which was afterwards destroyed by one of the Ghori Sultans, who built the present Jama Masjid on its site, and with its materials. The stones of the mosque are squared blocks of soft grey sandstone, and as many of them exhibit cramp-holes on the outside, there can be no doubt that they must originally have belonged to some other building.

To the south-east of the town there is a large, deep, irregularly shaped piece of water called Piráli Tál. It is nearly half a mile in length, but not more than 300 feet broad in its widest part. It is filled in the rains by a small channel carrying the drainage of the country from the north-east, and its overflow falls into the Malini River, about two miles distant. This pool is only part of a natural channel of drainage which has been deepened by the excavation of earth for the bricks of the town. But in spite of this evident origin of the Madawar tank, it was gravely asserted by the Buddhists to have been produced by an earthquake which accompanied the death of a celebrated saint, named Vimala Mitra.

According to Hwen Thsang, Madipur was 20 li, or 3 miles, in circuit, which agrees very closely with what would appear to be the most probable size of the old town. The King was a Sudra, who cared nothing for Buddhism, but worshipped the Devas. There were 12 Buddhist monasteries containing about 800 monks, who were mostly attached to the school of the Sarvástivâdas, and there were also about 50 Brahmanical temples.* To the south of the town, at 4 or 5 li, or of a mile, there was a small monastery in which Gunaprabha was said to have composed 100 works; and at half a mile to the north of this there was a great monastery which was famous as the scene of Sanghabhadra's sudden

Julien's Hwen Thsang, II., 219,

death from chagrin, when he was overcome in argument by Vasubandhu. His relics were deposited in a Stupa in the midst of a mangoe grove only 200 paces to the north-west of the monastery. These two chiefs of Buddhism lived about the beginning of the Christian era, and the Stupa was still standing in A. D. 634 at the time of Hwen Thsang's visit. There is no trace now existing either of the monasteries or of the Stupa, but their sites can be fixed with tolerable certainty by the aid of Hwen Thsang's descriptions. The village of Lalpur, which is situated on a mound about three-quarters of a mile to the south-south-east of the Jâma Masjid, and which is built partly of old bricks, represents the site of the small monastery of Gunaprabha. To the north of Lâlpur, and just half a mile distant, is the shrine of Hidayat Shah, with a Masjid attached, both of which are built of old bricks. This spot I believe to be the site of the great monastery of Sanghabhadra. Lastly, to the west-north-west of Hidayat's shrine, at a distance of 200 paces, there is another shrine, or Fakir's takia, standing in the midst of a mangoe grove, like the old Stupa of Sanghabhadra, the site of which it represents almost exactly as described by Hwen Thsang.*

Besides the mangoe grove there was a second Stupa which contained the relics of Vimala Mitra, who, as a disciple of Sanghabhadra, must have lived in the first century of the Christian era. The legend relates that, on passing the Stupa of his master Sanghabhadra, he placed his hand on his heart, and with a sigh expressed a wish that he might live to compose a work which should lead all the students of India to renounce the "Great Vehicle" (Maha Yâna), and which should blot out the name of Vasubandhu for ever. No sooner had he spoken, than he was seized with frenzy, and five spouts of burning hot blood gushed from his mouth. Then feeling himself dying, he wrote a letter "expressing his repentance for having maligned the Mahd Yana, and hoping that his fate might serve as an example to all students." At these words the earth quaked, and he expired instantly. Then the spot where he died suddenly sank and formed a deep ditch, and a holy man who witnessed his end exclaimed-To-day this master of the scriptures, by giving way to his passions, and by persisting in erroneous opinions,

See Plate No. XLII. for map of Madawar.

MADAWAR, OR MADIPUR KASHIPUR, OR GOVISANA. 251

has calumniated, the Mahá Yána, for which he has now fallen into everlasting hell." But this opinion of the holy man would appear to have been confined to the followers of the Maha Yána, for the brethren of Vimala Mitra, who were Sarvástivádas or students of the lesser vehicle, burned his body and raised a Stupa over his relics. It must be remembered also that Hwen Thsang, who relates the legend, was a zealous follower of the Mahá Yâna, and this no doubt led him to overlook the manifest contradiction between the statement of the uncharitable arhat, and the fact that his brethren had burned his body in the usual manner. This legend, as well as several others, would seem to show that there was a hostile and even bitter feeling between these two great sects of the Buddhist community.

The site of Vimala Mitra's Stupa is described as being at the edge of the mango grove, and from the details of the legend it is clear that it could have been at no great distance from the Stupa of Sanghabhadra. It would appear also that it must have stood close by the great ditch, or hollow, which his opponents looked upon as the rent in the earth by which he had sunk down to "everlasting hell." Now the mangoe grove which I have before mentioned extends only 120 paces to the westward to the bank of the deep tank called the Pirwáli Tál. I conclude, therefore, that the Stupa of Vimala Mitra must have stood close to the edge of this tank and on the border of the mangoe grove which still exists in the same position as described by Hwen Thsang.

It seems probable that the people of Madáwar, as pointed out by M. St. Martin, may be the Mathe of Megasthenes who dwelt on the banks of the Erineses. If so, that river must be the Mâlini. It is true that this is but a small stream, but it was in a sacred grove on the bank of the Málini that Sakuntala was brought up, and along its course lay her route to the court of Dushmanta at Hastinapur. While the lotus floats on its waters, and while the Chakwa calls its mate on its bank, so long will the little Málini live in the verse of Kálidás.

V. KASHIPUR, OR GOVISANA.

On leaving Madipur the Chinese pilgrim travelled 400 li, or 66 miles to the south-east, and arrived in the kingdom of Kiu-pi-shwang-na, which M. Julien renders by

Govisana.* The capital was 14 or 15 li, or 23 miles in circuit. Its position was strong, being elevated, and of difficult access, and it was surrounded by groves, tanks, and fish ponds. There were two monasteries containing 100 monks, and 30 Brahmanical temples. In the middle of the larger monastery, which was outside the city, there was a Stupa of Asoka, 200 feet in height, built over the spot where Buddha was said to have explained the law. There were also two small Stupas, only 12 feet high, containing his hair and nails.

According to the bearing and distance from Madipur, as given by Hwen Thsang, we must look for Govisana somewhere to the north of Muradabad. In this direction the only place of any antiquity is the old fort of Ujain, which is just one mile to the east of Kâshipur. According to the route which I marched, the distance is 44 kos, or 66 miles. I estimate the value of the kos by the measured distance of 59 miles between the Post Offices of Bareli and Muradabad, which is always called 40 kos by the Natives. The true bearing of Kashipur is east-south-east, instead of south-east, but the difference is not great; and as the position of Kashipur is equally clearly indicated by the subsequent route to Ahichhatra, I feel quite satisfied that the old fort of Ujain represents the ancient city of Govisana which was visited by Hwen Thsang.

Bishop Heber describes Kâshipur as a "famous place of Hindu pilgrimage which was built by a divinity, named Kashi, 5,000 years ago."+ But the good Bishop was grossly deceived by his informant, as it is well known that the town is a modern one, it having been built about A. D. 1718 by Kashi Nâth, a follower of Raja Devi Chandra, or Deb Chand, of Champawat, in Kumaon. The old fort is now called Ujain; but as that is the name of the nearest village, it seems probable that the true name has been lost. The place itself had been deserted for several hundred years before the occupation of Kashipur; but as the holy tank of Dron Ságar had never ceased to be visited by pilgrims, I presume that the name of the tank must have gradually superseded that of the fort. Even at the present day, the name Dron Sagar is just as well known as that of Kashipur.

*Julien's Hwen Thsang, II., 233.

Travels, Vol. II., p. 216.

The old fort of Ujain is very peculiar in its form, which may be best compared to the body of a guitar. It is 3,000 feet in length from west to east, and 1,500 feet in breadth, the whole circuit being upwards of 9,000 feet, or rather less than 2 miles. Hwen Thsang describes the circuit of Govisana as about 12,000 feet, or nearly 2 miles; but in this measurement he must have included the long mound of ruins on the south side, which is evidently the remains of an ancient suburb. By including this mound as an undoubted part of the old city, the circuit of the ruins is upwards of 11,000 feet, or very nearly the same as that given by Hwen Thsang. Numerous groves, tanks, and fish ponds still surround the place. Indeed, the trees are particularly luxuriant, owing to the high level of the water which is within 5 or 6 feet of the surface. For the same reason the tanks are numerous and always full of water. The largest of these is the Dron Sagar, which, as well as the fort, is said to have been constructed by the five Pandu brothers for the use of their teacher Drona. The tank is only 600 feet square, but it is esteemed very holy, and is much frequented by pilgrims on their way to the source of the Ganges. Its high banks are covered with sati monuments of recent date. The walls of the fort are built of large massive bricks, 15 inches by 10 inches by 2 inches, which are always a certain sign of antiquity. The general height of the walls is 30 feet above the fields; but the whole is now in complete ruin, and covered with dense jungle. Shallow ditches still exist on all sides except the east. The interior is very uneven, but the mass has a mean height of about 20 feet above the country. There are two low openings in the ramparts, one to the north-west and the other to the southwest, which now serve as entrances to the jungle, and which the people say were the old gates of the fort.*

There are some small temples on the western bank of the Dron Sagar; but the great place of worship is the modern temple of Jwâlâ Devi, 600 feet to the eastward of the fort. This goddess is also called Ujaini Devi, and a great fair is held in her honour on the 8th day of the waning moon of Chaitra. Other smaller temples contain symbols of Mahadeva under the titles of Bhutesar, Muktesar, Nágnáth, and

*See Plate No. XLII. for a map of Ujain or Govisana.