1S66.] Buddhist Monasteries and Temples. 69 



in that direction and the bed of the stream itself. The ridge is 

 manifestly an artificial work, and was originally intended either as a 

 wall to the ancient city, or as a rampart thrown np against it and the 

 neighbouring fort of Raj Ghaut. The latter supposition was that 

 held by Mr. James Prinsep, who imagined that it w r as cast up by 

 the Mohammedans in their attack upon Benares, and was specially 

 directed against the fort. This supposition may be true, although 

 it is difficult to perceive how it could have been of much service 

 either in an attack on the fort or on the city, especially in a period 

 when artillery was not in use. Had it reached as far as the river 

 Ganges, we could understand how, by severing the fort from the city, 

 it might have been a source of damage to both, but the south-western 

 extremity is not near the Ganges by a third of a mile or perhaps 

 more. "We are inclined to think, however, that this extremity was 

 once connected with that river, but at a time far more ancient than 

 the Mohammedan conquest of India. On the whole, it appears not 

 unlikely that this long embankment was the old boundary of the city 

 in the early periods of its history, which was possibly employed for 

 offensive purposes by the Mohammedans on the extension of the city 

 to the south and south-west, and the consequent abandonment of 

 this means of defence by the inhabitants. The embankment may 

 have been originally carried on to the Ganges in a straight line with 

 its present direction ; or, making a short circuit, may have entered 

 it by Tilia Nala, on the banks of which are the remains of a Buddhist 

 temple, which will be hereafter described. In this case, a portion of 

 it must have been thrown down and swept away to make room for 

 the growth of the city, and there is good ground for supposing that 

 the city extended in a narrow band on the banks of the Ganges 

 about as far as the Man-mandil observatory, even before the Christian 

 era. Should this idea be correct, it would follow that the most 

 ancient site of the city of Benares was situated within the limits of 

 this wall, stretching across from the Burna to the Ganges, cutting off 

 a tongue of land as far as the confluence of the two rivers, and 

 including the high land of the Raj Ghaut Fort, which was, in all 

 probability, once well populated. The city must have been then of 

 small extent, as compared with its existing dimensions, unless, as we 

 10 



