covered. The distance from the Guptár Ghât on the west, to the Râm Ghât on the east, is just 6 miles in a direct line, and if we suppose that the city with its suburbs and gardens formerly occupied the whole intervening space to a depth of two miles, its circuit would have agreed exactly with the smaller measurement of 12 kos. At the present day the people point to Râm Ghât and Guptar Ghât as the eastern and western boundaries of the old city, and the southern boundary they extend to Bharat-Kund, near Bhadarsá, a distance of 6 kos. But as these limits include all the places of pilgrimage, it would seem that the people consider them to have been formerly inside the city, which was certainly not the case. In the 'Ayin Akbari,' the old city is said to have measured 148 kos in length by 36 kos in breadth,* or, in other words, it covered the whole of the province of Oudh to the south of the Ghâghra river. The origin of the larger number is obvious. The 12 yojanas of the 'Râmâyana,' which are equal to 48 kos, being considered too small for the great city of Rama, the Brahmans simply added 100 kos to make the size tally with their own extravagant notions. The present city of Ajudhya, which is confined to the north-east corner of the old site, is just two miles in length by about three quarters of a mile in breadth; but not one half of this extent is occupied by buildings, and the whole place wears a look of decay. There are no high mounds of ruins, covered with broken statues and sculptured pillars, such as mark the sites of other ancient cities, but only a low irregular mass of rubbish heaps, from which all the bricks have been excavated for the

* Gladwyn's translation, ii. 32.

houses of the neighbouring city of Faizâbâd. This Muhammadan city, which is two miles and a half in length by one mile in breadth, is built chiefly of materials extracted from the ruins of Ajudhya. The two cities together occupy an area of nearly six square miles, just about one-half of the probable size of the ancient capital of Râma. In Faizâbâd the only building of any consequence is the stuccoed brick tomb of the old Bhao Begam, whose story was dragged before the public during the famous trial of Warren Hastings. Faizâbâd was the capital of the first Nawâbs of Oudh, but it was deserted by Asaf-ud-daolah in A.D. 1775.

In the seventh century the city of Visakha was only 16 li, or 2 miles, in circuit, or not more than one-half of its present size, although it probably contained a greater population, as not above one-third or perhaps less of the modern town is inhabited. Hwen Thsang assigns to the district a circuit of 4000 li, or 667 miles, which must be very much exaggerated. But, as I have already observed, the estimated dimensions of some of the districts in this part of the pilgrim's route are so great that it is quite impossible that all of them can be correct. I would therefore, in the present instance, read 400 li, or 67 miles, and restrict the territory of Visákha to the small tract lying around Ajudhya, between the Ghagra and Gomati rivers.

18. SRAVASTI.

The ancient territory of Ayodhya, or Oudh, was divided by the Sarjû or Ghâgra river into two great provinces; that to the north being called Uttara Kosala, and that to the south Banaodha. Each was again

tances from Bahraich and Gonda.* In shape it is an almost semicircular crescent, with its diameter of one mile and a third in length curved inwards and facing the north-east, along the old bank of the Rapti river. The western front, which runs due north and south, for three-quarters of a mile, is the only straight portion of the enclosure. The ramparts vary considerably in height; those to the west being from 35 to 40 feet in height, while those on the south and east are not more than 25 or 30 feet. The highest point is the great north-west bastion, which is 50 feet above the fields. The north-east face, or shorter curve of the crescent, was defended by the Rapti, which still flows down its old bed during the annual floods. The land ramparts on the longer curve of the crescent must once have been defended by a ditch, the remains of which yet exist as a swamp, nearly half a mile in length, at the south-west corner. Everywhere the ramparts are covered with fragments of brick, of the large size peculiar to very ancient cities; and though I was unable to trace any remains of walls except in one place, yet the very presence of the bricks is quite sufficient to show that the earthen ramparts must once have been crowned by brick parapets and battlements. The portion of the parapet wall, which I discovered still standing in the middle of the river face, was 10 feet thick. The whole circuit of the old earthen ramparts, according to my survey, is 17,300 feet, or upwards of 3 miles. Now this is the exact size of 20 li, or 33 miles, which Hwen Thsang gives to the palace alone; but, as the city was then deserted and in ruins, he must have mistaken the city itself for the palace.

It is certain at least that the suburbs outside the walls must have been very limited indeed, as the place is almost entirely surrounded with the remains of large religious buildings, which would have left but little room for any private dwellings. I am therefore quite satisfied that the city has been mistaken for the palace; and this mistake is sufficient to show how utterly ruined this once famous city must have been at so distant a period as the seventh century, when the place was visited by Hwen Thsang. As Fa-Hian describes the population as already very inconsiderable in A.D. 400, while the Ceylonese annals speak of Khiradhára, king of Sawatthipura between A.D. 275 and 302, the great decline of Srâvasti must have taken place during the fourth century, and we may perhaps not be far wrong in connecting it with the fall of the Gupta dynasty in A.D. 319.

Srávasti is said to have been built by Raja Sravasta, the son of Yuvanáswa of the Solar race, and the tenth in descent from Surya himself. Its foundation therefore reaches to the fabulous ages of Indian history, long anterior to Râma. During this early period it most probably formed part of the kingdom of Ayodhya, as the Vâyu Purana assigns it to Lava, the son of Râma. When Srâvasti next appears in history, in the time of Buddha, it was the capital of King Prasenâjit, the son of Maha Kosala. The king became a convert to the new faith, and during the rest of his life he was the firm friend and protector of Buddha. But his son Virudhaka hated the race of the Sakyas, and his invasion of their country and subsequent massacre of 500 Sâkya maidens, who had been selected for

Wilson,Vishnu Purana,' book iv. p. 2; Hall's edit., vol. iv. p. 263.

tances from Bahraich and Gonda.* In shape it is an almost semicircular crescent, with its diameter of one mile and a third in length curved inwards and facing the north-east, along the old bank of the Rapti river. The western front, which runs due north and south, for three-quarters of a mile, is the only straight portion of the enclosure. The ramparts vary considerably in height; those to the west being from 35 to 40 feet in height, while those on the south and east are not more than 25 or 30 feet. The highest point is the great north-west bastion, which is 50 feet above the fields. The north-east face, or shorter curve of the crescent, was defended by the Rapti, which still flows down its old bed during the annual floods. The land ramparts on the longer curve of the crescent must once have been defended by a ditch, the remains of which yet exist as a swamp, nearly half a mile in length, at the south-west corner. Everywhere the ramparts are covered with fragments of brick, of the large size peculiar to very ancient cities; and though I was unable to trace any remains of walls except in one place, yet the very presence of the bricks is quite sufficient to show that the earthen ramparts must once have been crowned by brick parapets and battlements. The portion of the parapet wall, which I discovered still standing in the middle of the river face, was 10 feet thick. The whole circuit of the old earthen ramparts, according to my survey, is 17,300 feet, or upwards of 3 miles. Now this is the exact size of 20 li, or 33 miles, which Hwen Thsang gives to the palace alone; but, as the city was then deserted and in ruins, he must have mistaken the city itself for the palace.

*See Map No. XI. for its position>