About 1869, some rock-cut apartments were discovered at the bottom of a descent on the north of the Jama' Masjid on the Uparkoṭ or fort of Junâgaṛh.1 They are of considerable interest, for though somewhat defaced, they manifest a high style of art. Few bases for example, could be found anywhere to excel in beauty of design and richness of carving those of the six pillars in the lower hall.2

The excavations (Plate IV.) opened up consist of a deep tank or bath (E) about 11 feet square, with a covered verandah on three sides of it; that on the west side is occupied with a built seat like the ásana for an idol, possibly for laying the garments upon while bathing. The pipes for the water come down the wall from the surface, pass the front of this seat, and enter a small cistern near the entrance door at the south-west corner. The water must have been raised from some well in the neighbourhood and conveyed to the supply pipe; and the small cistern may have been formed to assist in filtering the water pure into the bath.

The corridor on the south side is supported by two columns with spiral ridges on their shafts, octagonal plinths, and carved bases and capitals. The shafts of two corresponding attached pilasters on the north wall are divided into three sections each, having the grooves or ridges running in opposite spirals.

Over the bath the roof is open, and round the opening a wall still stands a foot or so above the ground-level.

In the north side over the bath is a large aperture or window into the next chamber. This apartment is entered from the northeast corner of the bath-room. It is a large chamber, 35 feet 10 inches long by 27 feet 10 inches wide, with six columns supporting the roof; the area between the first four of these, like that in the bath-room, is open to the air above, with a surrounding wall on the surface of the rock. It is also open to a hall below; and the four columns have been connected by a thin low parapet wall, about 20 inches high and 6 inches thick, now entirely destroyed. The rest of the area is occupied by the corridor on three sides, and by the space on the north where the remaining two pillars stand. In the walls on the north east, and west sides are stone bench recesses divided into long compartments, with a base moulded in architectural courses below, and a frieze above, ornamented with Chaitya-window and chequer carvings. The four pillars round the open area are square, the other two are 16-sided, and have been carved with animal figures on the abaci.

In the north-east corner a door leads into a small apartment which has a hole in the roof blackened with smoke, and which may have been used as an occasional cook-room, to prepare warm drinks, &c. for those who had been enjoying the bath. By the side of this apartment a door leads to a stair descending to the entrance of the hall below.

This lower room measures 39½ feet by 31 feet, and had evidently been filled up long before the one above it, and is consequently in a better state of preservation. It has been elaborately and very tastefully carved.3 On entering it we come on a platform on the left side, slightly raised and nearly square, with two short pillars on its west side, supporting a frame above, descending from the roof. What this was meant for is hard to say, unless the depression within was intended to be filled with cotton or other soft substance to form a dais or a seat.

Except on the west side, the remainder of the walls is surrounded by bench recesses, divided at regular intervals, as in the apartment above. Over these recesses the frieze is ornamented with Chaitya-windows having the Buddhist rail in the lower part of the opening, and two figures looking out of each; in many cases two females with something like "ears" on their head-dresses, but too indistinct to distinguish what they represent (Fig. 4, Plate III.).

The four columns in the south end of this hall are larger than the two in front of the supposed dais, but the bases of all are alike; and the bodies of the capitals are similar. The rich bases have been already alluded to, and the drawing (Fig. 3, Plate III.), where the original pattern has been truthfully restored from the different fragments still left entire, will give a better idea of them than could be done by any description.

The abaci afe carved with lions couchant at the corners, and in the middle of each is a lion, facing outwards, with a human figure on each side of it. The body of the capital consists of eight divisions round, indicated by the breaks in the ledge at the bottom, on which the human figures of the different groups stand. Most, if not all, of the figures are females, nearly nude, and some standing under foliage. They have been cut with considerable spirit, and in high, almost entire, relief: unfortunately, many of them have been much damaged,—some even since the room was excavated. In the two smaller columns, the principal member below the body of the capital is carved with the heads of animals, mostly elephants and goats or rams. On the larger columns the corresponding member is not so deep, and is a serrated torus. At the back or west side of this hall are two small rooms; that on the south with a single door, the other with three entrances between jambs slightly advanced, and with a projecting frieze.

On the north side of this is an irregular excavation, in a corner of which there seems to be a shaft of a choked-up well; but the whole excavation here is more like the work of Mahmûd Bîgarah's quarrymen in the fifteenth century4 than any portion of the original,though it is quite probable that other chambers have been quarried away.

These rooms could have been no part of a monastic establishment; and the example of the old Mehal, just to the north of this, suggests that they may have been either a sort of garden-house belonging to the palace, or possibly the bath and pleasure-house of another palace now interred under the debris that covers the whole of the Uparkoṭ. The style of carving is not unlike much that has been found about Maṭhurâ, and which I feel disposed to attribute to about the fourth century A.D.

• 1. This cave is described here because locally it forms one of the group, but from its age, probably belonging to the fourth century, it belongs to the second division of Buddhist caves according to the clasification adopted above (p. 185).
• 2. See Plates XXIII and XXIV. in Second Archæological Report. Quite close to these excavations on their south side the ground sounds hollow, and there is a line of wall cropping up, exactly similar to those round the tops of the two openings which led to the discovery of those excavated.
• 3. For drawings, &c., see Second. Arch. Report, p. 142, and plates xxi. to xxiv.
• 4. Mahmûd Bîgarah of Ahmadâbâd subdued Maṇḍalika the last or the Chuṛâsamâ kings of Junâgaṛh, and took the fort in 1469-70 where he erected the great Jama' Masjid, Arch. Sur. W. India, vol. ii. pp. 144, 165.