stand. In the north-west corner of the courtyard is a well, 7 feet in diameter, for the use of the monks, which is finished off with a circular parapet at the top. This parapet would appear to be a later addition; for, about 3' 6" below, the well is square externally where it pierces an original brick paving. Connected with this original brick floor in the south-west corner is a sort of platform, which in some ways recalls similar features existing at Taxila. Along the cloistered walls were originally placed stone figures on pedestals, which Dr. Spooner thinks were removed when the monastery was deserted. In the centre of the east side, and projecting from the back wall, was the chapel enshrining a colossal stucco figure of the seated Buddha, of which merely indications of the legs remain. Immediately in front of the sanctum, and at a slightly lower level, is a raised platform which, from the stone bases of the columns still in situ, appears originally to have been an open-pillared portico. It is worthy of note that the large gateway of the sanctum seems at some later period to have been blocked in the middle by a masonry pier, on either side of which was left a small door. This is apparent through the straight line of the jambs of the wide original gateway being still visible in the brick wall face.

In the entrance to this monastery Pandit Hirananda Sastri discovered a very important copper-plate inscription of Sri Devapaladeva, the third sovereign of the Pala Dynasty.

The plate is surmounted by a seal bearing the Dharmachakra emblem flanked by two gazelles, and the inscription it contains, which is in Sanskrit written in Devanagari character, records the grant of five villages in the Gaya and Rajagriha districts of the Srinagarbhukti (i.e. Patna Division) for the upkeep of the Nalanda monastery and the provision of comforts for the monks and bhiksus arriving there from all quarters. The inscription further records that Devapaladeva made this endowment at the request of the king of Sumatra, Sri Balaputra Deva-it would seem in return for an equivalent grant in that country. The inscription is dated in the 38th regnal year of Devapaladeva, corresponding to 891 a.d.

Nalanda: Metal image of standing Buddha in Abhaya (protection) Mudra.

Photo-engraved & printed at the Offices of the Survey of India, Calcutta, 1923

Several other finds of interest were made in this colonnaded verandah.

To pass on to the fourth level, this would appear to result from the reoccupation of the site after its earlier abandonment by the monks ; the ground level in the interval having risen somewhat as a result of silt deposit and natural earth accumulation. The verandahs, in which the stone figures were erected to inspire the monks with nobler aims, were apparently no longer desired, and therefore filled in. Evidence of this is to be found in the fineness of the outer face of the brick walls exposed to the courtyard, the corresponding inner face having been left rough and uneven. It was at this time that the necessity arose for the two main flights of steps, the one leading from the entrance up to the top storey of the monastery, and the other down therefrom into its courtyard; for the monks, it seems, still utilized the well in the courtyard which was now finished off with the higher circular parapet at its top. It may be observed that the level of the well, as added to, was slightly higher than the foot of the steps leading down to the courtyard. Another addition of considerable interest was made at this time in about the centre of the north wall of the court, where two structural of brick with corbelled entrances, 3′ 10′′ wide, curiously resembling the rock-cut caves at Barabar, were erected. These have vaulted roofs; the side walls rising vertically 3' 9", beyond which the vault is carried up another 5' 9". The chambers are identical and measure 15' 6" by 11'8" internally, the thickness of the entrance wall being 4 feet. "As these caves (at Barabar) are cut with curving ceilings, their imitation here has necessitated the covering of the chambers with a vaulted roof, which constitutes not actually the first but is among the first examples prior to Muhammadan influence. No exact date can yet be assigned to this building, but that it is pre-Muhammadan seems certain, and this invests the arch with real importance. The bricks seem to have been specially made for the purpose. They are not very large ancient bricks and not squared in the usual way, but laid as stretchers

caves

The verandah in front yielded

with the sides a little slanted." several sculptured fragments of some interest, which include one remarkable plaque representing the principal events in the life of Buddha, though the Mahaparinirvana scene at the top is broken and missing. The plaque appears to have been lodged originally in the niche between the doors of the two caves. It was found in fragments, and these have been carefully reset together.

Lastly, there remains to be noticed the sixth level of occupation which, however, may or may not be the earliest on this site. Beneath the floor at the south-east corner of the verandah of the monastery at a depth of about 6 feet below the level of the cloistered walk, and again in one of the rooms on the south side, still another pavement of lime-concrete was discovered, as well as a couple of brick walls in the same corner.

Further excavation carried out in 1921-22 disclosed the existence of a brick paving in the southern half of the monastery, which feature probably extends over the whole courtyard. This paving was found at a level of some 22 feet below the extreme top parapet of the monastery court and 3' 6" below the courtyard level of the fifth successive stratum referred to previously. A feature of interest uncovered at the sixth level in the centre of the south wall was the remains of what appeared to be a low chabutra, inset in which were duplicate panels in low relief of bird-bodied men worshipping a lotus plant. From the style of this decorative relief Sir John Marshall considers it to date probably from the seventh century or "possibly the sixth".

The level of the last stratum in relation to the fifth (assigned) on the evidence afforded by the copper-plate of Devapaladeva found here to approximately the ninth century A.D.) would seem to indicate independently a date of about the sixth or seventh century for the sixth stratum, thus confirming that suggested by the style of the decorative panel in question.

Two further subsidiary levels in the succession of monastic structures erected on this same site are discernible in the snall

1 Annual Progress Report: Archæological Survey, Central Circle, 1916-7.