ANCIENT MONUMENTS OF KASHMIR

Monuments below Srinagar

PARIHASAPURA

The base is 128' 2" square in plan, with offsets and a flight of steps on each side. Its mouldings are of the usual type, a round torus in the middle and a filleted torus as the cornice. The steps were flanked by plain rails and side walls which had pilasters in front decorated with carved figures of seated and standing atlantes. Some of these are in position, while others, which were lying about loose, have been transported to the Srinagar Museum. They are not grotesque creatures like those so commonly seen in Gandhara, but have the appearance of ordinary respectable gentlemen, whose placid features seem to indicate that the superincumbent weight sits lightly upon them. The top surface of each of the two plinths is broad and affords adequate space for circumambulation. Among the loose architectural stones lying scattered about the site are a few curious blocks in the south-eastern and south-western corners. They are round torus stones adorned with four slanting bands or fillets running round the body. As this type of torus moulding is not used in either of the bases, it is probable that it belonged to the string-course on the drum of the stupa. There are fragments of trefoiled arches also, which contained images of the Buddha and Bodhisattvas.

The large square structure to the south of the stupa is the rajavihara, or royal monastery. A flight of steps in the east wall gives access to one of its cells which served as a verandah. The monastery is a quadrangle of twenty-six cells enclosing a square courtyard which was originally paved with stone flags, some of which are extant. In front of the cells was a broad verandah, which was probably covered, the roof being supported by a colonnade which ran along the edge of the plinth. A flight of steps corresponding to the one mentioned above leads down to the courtyard. Exactly opposite to this, in the middle of the west wall, are three cells preceded by a vestibule, which is built on a plinth projected into the courtyard. It is probable that these were the apartments occupied by the abbot of the monastery. Near a corner of it is a large stone trough, which may have served as a water reservoir for bathing purposes. A couple of stone drains passing underneath cells Nos. 18 and 21 (if we begin counting from the cell to the south of the entrance chamber) carry off the rain and other surplus water from the courtyard. Externally the plinth is about 10' high. In cell No. 25 (that is the one to the north of the entrance chamber) was found a small earthen jug which contained forty-four silver coins in excellent preservation. They belonged to the time of kings Vinayaditya, Vigraha, and Durlabha. They are now exhibited in the Numismatic Section of the Srinagar Museum. The monastery was repaired at a subsequent period. The repairs are plainly distinguishable in the exterior of the wall on the eastern and western sides.

The building next to it on the south side is the chaitya built by Lalitaditya. It stands on a double base of the usual type. A flight of steps on the east side leads to the entrance, which must originally have been covered by a large trefoil-arch, fragments of which are Iying about the site. This building possesses some of the most massive blocks of stone that have ever been used in Kashmiri temples, and which compare favourably with those used in ancient Egyptian buildings. The floor of the sanctum is a single block 14' by 12' 6" by 5' 2".

The sanctum is 27' square surrounded by a circumambulatory passage. It is probable that its ceiling was supported on four columns, the bases only of which survive at the four corners. The roof, which was probably supported on the massive stone walls of the pradakshina, may have been of the pyramidal type.

The courtyard is enclosed by a rubble-stone wall which has nothing remarkable about it. In front of the temple steps is the base of a column which probably supported the dhvaja, or banner, bearing the special emblem of the deity enshrined in the sanctuary.

The flank walls of the stair were adorned with atlantes similar to those of the stupa.

Near the chaitya is the foundation of a small building of the diaper-rubble style.

While this Plateau was reserved for the erection of Buddhist buildings only, the other two were exclusively appropriated by Hindus. Perhaps the arrangement was intentional, to avoid possible friction between the two powerful religious bodies. On the karewa locally known as Gordan there are ruins of a Hindu temple which are probably all that remain of Lalitaditya's temple of Govardhanadhara. Crossing the ravine in which nestles the little village of Diwar- Yakmanpura, and ascending the Plateau opposite, are seen the immense ruins of two extraordinarily large temples - one of them has a peristyle larger than that of Martand - which may represent Lalitaditya's favourite shrines of Parihasakesava and Muktakesava.

PATAN

Sugandhes'a Temple

The Sugandhesa temple does not differ materially from other temples of Kashmir. The shrine is 12' 7" square and has, as usual, a portico in front. It is open on one side only, and has trefoiled niches externally on the other sides. These niches contained images. The temple stands on a double base, but it seems probable from the flank walls of the lower stair and the frieze of the lower base, in which the panels intended for sculpture decoration have been merely blocked out, but not carved, that the temple was never completed.

The entrance to the courtyard is in the middle of the eastern wall of the peristyle, and consists, as usual, of two chambers with a partition wall and a doorway in the middle.

Among the architectural fragments lying loose on the site, the most noteworthy are (a) two fragments of fluted columns with their capitals, (b) two bracket capitals with voluted ends and carved figures of atlantes supporting the frieze above, (c) a huge stone belonging to the cornice of the temple, bearing rows of kirtimukhas (grinning lions' heads) and rosettes, and (d) a stone probably belonging to the partition wall of the entrance, having (1) two small trefoiled niches in which stand female figures wearing long garlands and (2) below them two rectangular niches, in one of which is an atlant seated between two lions facing the spectator, and in the other are two human-headed birds.

The cornice of the base of the peristyle is similar to that of the Avantisvami temple. The cells were preceded by a row of fluted columns, bases of some of which are in situ while those of others are scattered about in the courtyard.

The attention of the visitor is called to the slots in the lower stones of the jambs of the cells. These are mortices for iron clamps which held pairs of stones together. Pieces of much-corroded iron are still extant in some of the mortices.

Sankaragaurisvara Temple

The temple itself is an imposing pile, though a great deal of its grandeur has been taken away by the concealment of its plinth. The cella is 17' square and the central stone of the floor measures 12' 6" by 10'. It has nine circular holes arranged in three rows. It is possible that these were mortices of tenons which held in position the pedestal of the idol. The left wall of the portico has a trefoiled niche which is divided into two panels. The lower and larger one contains a number of figures, of which the principal seems to be S'iva. Above it, in the upper foil, is the squatting figure of the elephant-headed god, Ganesa, whose presence here would conclusively prove, even if there were any doubt about it, that the temple was dedicated to S'iva. The jambs of the recesses on the exterior of the temple have half-engaged columns which are decorated with well-executed geometrical and other patterns. Their capitals are surmounted by human-headed birds.

A few yards to the north of the Patan dak bangalow has recently been excavated an old baoli whose waters are confined in three rectangular reservoirs which are connected with each other. The one in the middle contains a miniature temple constructed originally of three stones (Plate LVII). The top-stone is missing. It is 2' 8" square externally and is open on all four sides. The openings seem to have been closed originally with wooden doors. These little shrines belong to the time when the prosperity of the Hindus had waned, and they were not capable of devoting so much wealth to the glorification of their religion.

USHKAR AND BARAMULA

Of the monasteries and temples which Hsuan-tsang saw, and Kalhana mentions, none now remain above ground, except the ruins of a stupa and its surrounding wall, a few yards to the west of the village. On the analogy of style which is similar to that of the great stupa at Parihasapura, there can be little doubt that it is the same structure which the Kashmir chronicle states Lalitaditya built in the middle of the eighth century A.D. Only the lowest courses of its base are now in position.

An interesting fact about this stupa is that it seems to have been built over an older structure of nearly the same type, stones of which were found in situ when the silt round the base was removed some years ago. That structure may have belonged to Kushan times. This surmise is strengthened by the discovery outside the north-eastern corner of the surrounding wall, of eleven terracotta heads, besides a number of fragmentary limbs of images which display the unmistakable influence of the Gandhara school of the third and fourth century. These are now preserved in the Srinagar Museum.

Plate LVIII. (a) illustrates the head of a Bodhisattva. The unusually ornamental treatment of the hair in this fragment is noteworthy. The delicate features, rounded chin, and twisted, dandified locks secured by a beaded fillet placed sideways, make the face attractive in spite of the somewhat weary smile and the self-satisfied expression of the face.

(b): The shaggy beard, close-pressed lips, knitted eyebrows, and furrowed forehead, of this Brahman ascetic are so remarkably realistic that it would be difficult to imagine that the artist was not drawing a portrait from life. The hair is neatly brushed upwards and was probably gathered in a knot at the back of the head, where it was kept in position by an ornamental band. The ardent gaze and the prominent cheekbones are indicative of self-mortification.

(c): This is one of the most beautiful heads found at Ushkar. The oval face, small nose, sensitive nostrils, soft delicate lips, plump rounded chin, hair smoothly combed back and falling in curly tresses on the shoulders, are all essentially feminine. She is an upasika, or female lay devotee. Her soft and wistful gaze, intensified by the upturned poise of the face, shows with what a feeling of devotion these feminine worshippers approached the Master.

(d): This illustration represents the head of a contemplative young monk with shaven crown, high forehead, arched eyebrows, and large dreamy eyes. The remarkably high and narrow skull seems to be the result of lateral pressure, a practice which was once prevalent among certain tribes in Central Asia.

One of the large blocks of stones lying loose in the trench to the north of the stupa bears the Sarada letters he-sh-ka-ra incised upon its rough surface, a circumstance which proves that even in Lalitaditya's time the town was known by its colloquial name, though the dignified Sanskrit original remained in vogue in literature.

Not the least interesting object here is the compact surrounding wall built of extremely small chips of stone in mud. The entrance to the courtyard must have been on the east side, opposite to the stairs of the stupa. The lower portion of the wall is extraordinarily thick and served, no doubt, as plinth for a range of cells which ran along its entire length, the upper portion being utilised as their back wall. The foundations of the wall are pierced at intervals with openings for drains.

Other objects of interest near by are two colossal S'iva-lingas erected at the two ends of the village.

The town of Baramula, properly Varahamula, named after the Boar incarnation of Vishnu, was an important place in mediaeval times. The temple of Adi-Varaha, " Primeval Boar," destroyed by Sikandar But-shikan, is said to have been one of the most splendid in Kashmir. Its site is identified, on the strength of local tradition, with the Kotitirtha situated half a mile below the bridge. A few architectural stones may still be seen lying about at this place. The only object of interest which it now contains is the large human-faced S'iva-linga (Plate LIX).

Further down is the site of the old watch-station, known as Drang, which name it has retained from very early times. It served the double purpose of a customs post and frontier outpost where traffic could be controlled and all suspicious characters apprehended.

FATHGARH

A late memorial stele displaying very poor artistic skill has been discovered in the excavation of the cella. The lower panel contains a pair of clumsy human figures seated on stools facing each other. The upper one contains a trident.

The dadoof the portico is adorned with a rowof pedimented niches.

The open doorway faces north-west and, like the closed recesses on the exterior of the other three sides, was covered by a large trefoiled arch surmounted by a pediment. Its pilasters are, as usual, adorned with half-engaged fluted columns. This temple has a particularly massive appearance and its thorough excavation is very desirable. In the south-east corner of the area is a late brick well.

NARANTHAL

The interior is a square of 7 1/4' and is 9' 5" high. There is no ceiling.

The topmost stone of the roof has a circular mortice in the centre which was originally intended to hold the finial which crowned the apex of the pyramid. The temple is probably of the late mediaeval era, not earlier than the twelfth century A.D., and perhaps much later.

BUNIAR

The gateway is a double-chambered structure faced on each open side by a trefoil arch surmounted by a steep pediment. The lintels of the closed arches are supported on pairs of columns which were originally fluted, though the weather has now left no trace of flutes. They have a double capital, the upper one being voluted on all four sides. The walls are externally surmounted by a cornice of kirti- mukhas, alternating with miniature trefoiled niches. Upon this rests the first course of the pyramidal roof.

The flights of steps-on the eastern and western sides respectively afford entrance to and exit from the entrance chamber. The one on the roadside is buried underground, but the inner stair has been excavated. It consists of seven steps flanked by sloping rails and upright side walls. Between this stair and the temple is a small stone platform which formed the lowermost course of the stepped base of a column (most probably a Garudadhvaja).

The priest in charge of the temple has now placed in it a small stele of very crude workmanship and late date, which he has painted with vermilion. Another similar stele, still standing in the position in which it was found, is seen in front of the temple stair.

The temple itself stands on a double base, which is in every respect similar to other structures of its kind in Kashmir. (Plate LXXIV). A lofty trefoil arch, standing upon advanced pilasters and enclosing a rectangular entrance originally surmounted by an ornamental trefoil and steep pediment, gives access to the sanctum. The jambs of the entrance are adorned with half-engaged columns. The interior is a square of 14 feet. The pedestal of the image is placed on a broad platform. The original image, which seems to have been of Vishnu, is now replaced by small Siva-lingas originally brought from the bed of the river Narbada. The walls are covered with a coat of modern whitewash. The string course from which the ceiling springs is still visible, but the ceiling itself which Bishop Cowie saw in 1865 and described as domical, has since either fallen down or been removed. It was, no doubt, similar to the ceilings of the largcr temples at Wangath.

Externally the only decorations are the trefoils of the recesses, their pediments, and the cornice of kirtimukhas and miniature trefoils from which the roof sprang.

The quadrangle measures 145' by 119 1/2', and consists of fifty- three cells and the gateway. They are rectangular, 7' long by 4' broad. Each cell has a single trefoiled entrance enclosed in a high- pitched pediment resting on half-engaged columns. These ranges of cells are preceded by a noble colonnade which stands on a base similar to that of the temple. A transverse beam connects the capitals of the columns with the roof of the cells. Over these beams rises the entablature, only one course of which, namely the frieze of miniature trefoils, is extant (Plate LXXVI).

In the centre of each range of cells, except, perhaps, the one in which the gateway stands, is an apartment of larger dimensions preceded by a pair of taller columns which are advanced about 4' from the rest of the peristyle.

The top course of the cells is also decorated in the same way as the frieze above.

On the south side, projecting from the cornice of the upper base of the temple, is the spout of the channel which carried off the washings of the image. It seems to have been shaped originally into a makara, or crocodile's head. Immediately below it is a huge water trough carved out of a single block of stone.

The rain-water in the courtyard is carried off by a drain which runs under the south-eastern corner of the peristyle.

In cell No. 11 of the north range, beginning the reckoning from the corner nearest the gateway, is the side entrance, which was then, as now, closed with a wooden door. The monotony of the external face of the western wall is partially relieved by rows of small square projections. In its two corners are two cells opening outwards.

Immediately outside the side-door mentioned above is a square structure built of plain blocks of stone. The middle portion of each of its four walls has fallen down, and the gaps have been filled in with small chips of stone built in mud. It is diflicult to surmise what was its original purpose.

DHATHAMANDIR, OR THE BANDI TEMPLE

The only material difference in style between this temple and the one at Buniar is that it does not possess any colonnade, and that the cells, which in the former are built of granite as the temple itself, are here built of kanjur, their plinth only being of limestone. The jambs of the cells were decorated with half-engaged columns, remnants of which may still be seen in the south-western corner of the peristyle.

The central shrine faces north-east, and is built of a very beautiful green limestone, which, curiously enough, seems to have been covered with a thick coat of lime plaster, decayed fragments of which are still clinging to various parts of the walls. The pedestal of the image is extant. There are two small shrines, replicas of the main building in the north-western corner of the courtyard.

The bases of the columns which supported the trefoil arch of the gateway are still in situ. Fragments of columns as well as their beautifully carved capitals are lying about in the compound and the area outside the gateway.

On the hillside, a few yards to the south-east of the temple, are remains of two smaller shrines which, like the subsidiary temples in the courtyard, are replicas of the main temple.

FIROZPOR= DRANG

The village contains the ruins of a small temple which was probably originally surrounded by a peristyle, no part of which, except the double-chambered gateway, is now extant above ground. The chief cause of the ruin of this temple seems to have been the unchecked growth of vegetation. Several walnut trees have taken root in the masonry, and are continuing the process of destruction.

The temple faces north-east. The roof has fallen in, and has filled the whole cella with large boulders which have completely hidden the floor, and consequently made it impossible to ascertain to which particular god the shrine was dedicated. Internally it is 11' square. The ceiling seems to have been of the Narastan type. The portico was surmounted by a trefoil-arch and a pediment of the usual type.

Externally also this temple is similar to the Narastan temple, inasmuch as the corner pilasters have very slight projection and the trefoiled recesses on the sides are smaller than the arch of the portico. It stands on a base surmounted by a cyma recta moulding. Admission is gained by a flight of steps flanked by plain side-walls.

The gateway measures 15' 6" square externally.

Ascending the spur, at whose foot these ruins are situated, we come to a very beautiful piece of open ground in the midst of dense forest. Among the trees at one corner of it are the scanty remains of the base of a small temple.

The long lines of walling between the temple and the village are regarded by tradition as the ruins of the old watch-station.

MANASBAL

A few yards from the temple, up the hill, is a natural cavern of no particular interest. Further up is the Shahkul canal brought down from the Sindh river by Zain-ul-abidin.

On the south bank of the lake are ruins of a terraced Mughal garden, the construction of which is ascribed by tradition to Empress Nur Jahan. The garden, locally known as the Jarogha, is entirely in ruins, the only parts standing being the retaining walls and the corner bastions of the terraces.

There are a number of sculptured stones, belonging to some Buddhist shrine, scattered about on top of the low ridge on the northern bank of the lake.

ANDARKOT

There is not much left today of the Buddhist viharas and Hindu temples which are said to have been built here by Jayapida. Practically everything has been destroyed. The temple of Kes'ava is represented by a large heap of scorched and shapeless boulders. The only objects of interest are a couple of sculptured reliefs on two faces of the pilaster of the stair; and they too are woefully defaced. The relief on the flank depicts a four-armed Vishnu seated on a cushioned throne in what is technically known as the lalitasana (sportive attitude), while an attendant, probably a female, is standing on his left. The left foot of the god rests upon a small footstool. His upper right hand holds a mace, and the lower one is placed in abhayamudra (pose of grantirig immunity from fear). He is profusely ornamented, and wears the mandaramala (garland - of celestial flowers). Under his throne are an atlant and some other indistinct figures. The relief on the front panel contains a group of three figures: Vishnu seated between his two consorts.

The sculptures belong to about the seventh or eighth century A.D.

Andarkot has a particularly sorrowful interest for the Hindus of Kashmir, for it was in the fort here that the last Hindu ruler, Queen Kotadevi, surrendered to her rebellious servant, Shah Mir, on certain stipulations which he did not fulfil afterwards.

THE WULAR LAKE

The historian Jonaraja states that King Zain-ul-abidin was anxious to build such a monument to his fame as had never been built by any king before. After much deliberation he decided to construct an island-palace in the middle of the Wular lake, but he was so greatly impressed with the magnitude of the task that he sought the wise men of his time for advice regarding the ways and means of its accomplishment. They encouraged his efforts by narrating the following legend:

In ancient times there was a city at that very place where there is the lake nowadays. The city was presided over by the great snake deity, Mahapadma. But the lapse of time, and the security given by beneficent rule, brought in its train luxury and vice among the citizens, which grew with such alarming rapidity that Mahapadma resolved upon the destruction of the entire city, including its inhabitants. But, among the residents of the doomed city was a humble and virtuous potter, whom the deity resolved to spare. He appeared to him in a dream and, warning him of the impending fate of the city, told him to save himself. Next morning the potter related the warning to the citizens, but was laughed at for his pains. As soon, therefore, as the potter left the city, the waters began to rise until the entire countryside was overwhelmed by one tremendous deluge. To quote from Jonaraja: "The terrified children who stood, at first when the water was low, at the feet of their mothers, soon, as the water began to rise higher, got into their laps, then clung to their breasts, next jumped upon their shoulders, and finally sat upon their heads, as if they were their embodied vital breaths. The flood covered the quaking limbs of women as if it were an affectionate lover embracing his beloved whose limbs are trembling with emotion."

After the Naga had thus wreaked his vengeance upon the sinful city, he established his permanent residence in the newly formed lake.

" The Lord of the Snakes was in reality Kaliya, who, being trampled upon by the feet of Krishna, bore the mark of a lotus on his head, and from that time onwards bore the appellation of Mahapadma."

Zain-ul-abidin's wise men then wound up their story by the naive assurance that he, being an incarnation of Hari, could do as he pleased in the domain of his liege subject Mahapadma, and that his efforts would assuredly be crowned with success. He therefore set about vigorously for accomplishment of his purpose. Large cargo boats, which are still plentiful in Kashmir, were filled with boulders and sunk at the spot selected. The island was named Zainalanka after its builder, and still retains that name. The king built a small but very beautiful mosque at one corner of it, the ruins of which are still in existence. It is built of large blocks of limestone, similar to those used in the mediaeval temples of Kashmir. Even the mouldings are those which are so common in the ancient Hindu temples. The principal decorative feature externally is a double row of trefoiled niches, which are flanked by pretty fluted columns surmounted by ribbed capitals. The lower row of niches stands on a quirked ovolo course, and is surmounted by a filleted torus. The corner pilasters have square filleted capitals. The two flank walls of the porch survive.

The interior was paved with large stone flags. The floor was on a level with the filleted torus outside.

A large assortment of architectural fragments of Hindu temples, such as fluted columns, stones from pilasters, jambs, etc., are scattered round about the site, which tends to prove that there was once a Hindu temple also on the island. This hypothesis is further strengthened by the presence of a big Siva-linga, which is partly submerged, on the east side of the island. It is probable that the temple also was built by Zain-ul-abidin, for none of his successors on the Kashmir throne was capable of such a colossal task. There are indications to show that there were series of steps on all sides leading down to the water, though the principal landing seems to have been on the east side.

The only other structure on the island is a small domed chamber of brick masonry measuring 15' 7" square internally. This was also partly built of Hindu materials. The string-courses on top of the walls consist of projecting wooden beams. The entire surface of the chamber seems to have been coated with painted plaster. Externally the walls were adorned with shallow arched recesses which were originally covered with plaster. They were separated by glazed tiles. The square impressions of these tiles are visible on the plaster backing wherever it still adheres to the walls. Fragments of painted and glazed tiles are lying in the debris.

The structure seems to be a tomb, and is probably contemporary with the mosque.

In 1874, Dr. Vincent-Smith discovered on this island a stone slab bearing a Persian inscription which mentioned the name of Zain-ul- abidin and contained the date A.H. 847 - A.D. 1442-43.

GARUR

The ceiling is built of overlapping stones.

WANGATH

Narannag is the modern name of ancient Sodaratirtha, which has been, since very early times, an important place of pilgrimage in Kashmir. The site probably owes its sanctity to the existence of a large spring, near which have been built two groups of temples belonging to the mediaeval era. All the temples are more or less in ruins, chiefly owing to the inroads of vegetation. The stone used is the greyish granite which is found in abundance at the place. The first group, that is the one nearer Wangath, comprises six temples situated within an enclosure wall. Judging by their positions, as they are by no means symmetrically disposed in relation to one another, and by the difference in their architectural details, it is probable that the various structures were built at different dates (Plate LXII).

(1) The principal temple is a square of 25', and, except in a few particulars, does not differ from other temples of Kashmir. The first point of departure from the usual style are the entrances. We have seen the temples of the vimana type, which have all the four sides open. But this one has two entrances opposite each other in the north-east and south-west sides. The second distinction is its domed ceiling, though there is no doubt that externally the roof was pyramidal. A large quantity of lime has been used in the masonry of the temple. The ceiling is built of circular courses of kanjur stone, and is crowned at the apex by a full-blown lotus. The dome springs from four large corner stones, which cut off the angles formed by the walls. On two sides of the string-course upon which the dome rests are eight rectangular slots (four on each side), which seem to have been intended to hold the rafters of a canopy over the image. The interior measures 17' square. In the centre of the floor is a square space which is unpaved. It marks the site of the pedestal of the image. The mortices of the tenons of the doors can still be seen in both entrances.

The two sides which are closed are decorated externally with square-topped recesses, each of which contains the pedestal of an image which was probably a replica of the one in the sanctum itself. The core of the roof consists of rubble-stone masonry in lime.

This temple has been identified by Sir Aurel Stein with the Jyeshthega temple of Lalitaditya.

(2) The small temple to the left possesses niches on three sides intended for images. Their pedestals with tenons are in situ. Part of the kanjur backing of the dome is intact on top of the walls.

(3) The temple immediately touching the porch of the preceding shrine is a single square structure, plain both internally and externally. One stone of the roof, which is the only one in situ, shows that the ceiling was not domical, but was built of overlapping stones.

(4) Immediately behind is the basement of another small temple. Its superstructure has fallen down.

(5-6) Of the two temples behind, one has its entrance facing north-east, and the other faces south-east. The ceiling of the former was of the overlapping type while that of the latter was probably domical.

(7) Immediately to the left of the latter is the ruined plinth of another temple, smaller than any of those described above.

The gateway of the enclosure is situated, curiously enough, at the north-west corner, and not in the middle of the wall, as is the case in other temples. It is, as usual, a two-chambered structure. In each chamber are the bases of two columns which supported the roof of the porch. Fragments of circular columns which stood on the bases- are lying strewn round about the site. The retaining wall of the plinth on which this group of temples stands is built of closely packed pieces of slate, and in places where vegetation has not played havoc with it, it is very compact and beautiful.

A broad causeway paved with huge blocks of granite leads to the second group of temples, situated about two hundred yards farther off.

Between the two groups of temples are a number of structures which in themselves deserve to be classed into a third group. All of them have fallen down, but one, the base only of which is in existence, is unique in Kashmir. It appears to have been a spacious pillared hall or barahdari about 100' long by 67' broad. The bases of the columns are in situ. They are eight in number on the longer side and four on the shorter. The staircase is built between the central pair of columns facing the first group of temples. The slots in the landing on the top of the stairs seem to have been intended for holding the posts of screens.

The lower group of temples is likewise enclosed in a massive rectangular stone wall pierced by a two-chambered gateway. Inside the wall are half a dozen structures, all in ruined condition and partly buried under the ground. The largest temple has been identified by Sir Aurel Stein with that of Siva-Bhutesvara. Internally it is a square of 17', and was similar to the largest temple in the first group. The small temple to its right has a round-headed entrance. There are two small shrines on its south, the first of which differs from all the structures at this site in that it is built of small pieces of slate in lime mortar, placed upon a granite base. The peculiar feature of the second is an immense rectangular stone trough carved out of a single gigantic block of granite. Among the objects lying scattered about the compound, the most noteworthy are two fragments, in the south-east corner of the quadrangle, of an octagonal pedestal of a S'iva-linga, and two stones in the same corner of the quadrangle, each decorated with a pair of remarkably well- executed geese.

The long raised line of brick masonry on the enclosure wall is the debris of the cellular quadrangle which usually surrounds the Kashmiri temple.

Outside the north-western corner of the peristyle is the large spring, the real lodestone of the pilgrims, which was the occasion of all this magnificence. Its cool, delicious water, perhaps, contributed to some extent to its sanctity. The water issues out of the mountain side through the masonry walls of the tank. Slightly higher up, and more than half-buried under the ground, is a small temple whose double pyramidal roof is the only existing roof in the whole group of temples here. This has been identified by Sir Aurel Stein with the temple of Bhairava, near which the turbulent baron Dhanva was decapitated for encroaching upon the lands and revenues of these temples.

Before leaving the site, the visitor should not fail to observe the unusually massive stone retaining wall of this group of temples. The granite blocks are of so extraordinary a size and are so beautifully dressed and finely joined as to give the impression that their carvers regarded them more in the light of wooden beams than as close-grained intractable boulders to be chipped into shape with vigilant care and inexhaustible patience. It is probable that this wall served the double purpose of protecting the temple enclosure from being overwhelmed by the debris of the hill above, and also as the back wall of the range of cells on this side. The site has from time immemorial enjoyed a very high degree of popular esteem, which remains unimpaired to this day, thanks to the pilgrims who, after consigning the ashes of their dead relatives to the Gangabal lake, make it a point to offer worship here. The temples were endowed with extensive estates, and the priests in charge seem to have been a particularly influential body. In the later mediaeval period, after the death of Avantivarman, these temples shared the misfortunes which came upon Kashmir with ever-increasing violence. The temple treasury was plundered by Bhadresvara, the minister of Samgramaraja (A.D. 1003-1028); a conflagration in the reign of Uchchala (A. D. 1101 - 1111) inflicted much damage upon the buildings; during the reign of Jayasimha (A.D. 1128-1155), Hayavadana, a rebel baron, had the temples " sacked by marauding hillmen." Sumanas, a brother of Rilhana, the minister of Jayasimha, built a matha or congregation hall here. It is possible that the pillared hall is this same matha. Further excavations may throw light upon the question.