The wall of this fortification is of brick, and the foundation is traced without difficulty. We rode round it within side, and as far as we could calculate it could not measure less than two miles round. It is of an irregular figure, but in a position rendered remarkably strong by two rivers, which run their course round three-fourths of the wall, and then unite. The depth from the wall to the rivers is from fifty to a hundred feet, and in some places still more; in many parts the descent is nearly perpendicular. Where the rivers do not encircle the walls, a deep moat is cut from one river to the other, which is easily flooded by stopping the course of either river. It is about seventy-five feet wide and not less than fifty deep, even in its present state, filled up no doubt considerably, and in many parts cultivated. There appear to be several dwellings, if not villages, within the walls, and a good deal of cultivation, principally of tobacco.

The next morning we proceeded to visit the ruins at Kédal and Jágu, the former about seven miles, the latter nearly four from Málang, in a south-easterly direction.

At Kédal are the remains of a very beautiful temple of stone: its present height about thirty-five feet. The building is supported by a lion at each of the four cornices, and one on each side of the steps of the entrance. In the centre of each of the lower departments, between the lions, are figures in relief upon the wall. The mouldings and sculpture on this temple are in the same style as those of Brambánan and Bóro Bódo, but of still greater beauty. The building is surrounded by a square wall, and in the front is a raised terrace. The chamber appears to be of the same form as most of the temples in Java. Over the entrance is an immense gorgon head, and in the chamber itself a deep hole.

There are no Hindu images or other traces of Hindu mythology, except what may be afforded by the lions, and the figures in relief above mentioned. These represent the same principal figure, but with different attributes. On one side three immense serpents entwine over the head, the tail of one of them evidently held in the right hand; on another a water-pot, with a serpent's head attached to it, is on the head of the figure; and on the other there is a female figure with a serpent, the female reclining over the head.

This temple is just within the skirts of a forest.

At Jágu, also, we found the ruins a few yards within a forest, but these appeared to have been more extensive than the preceding.

The base of the principal building is much larger than any of the temples we visited in the eastern part of the island, and there appeared to have been originally two or three terraces rising one above the other to the height of thirty feet. The form of the entrance still appears, but the roof, sides, and back part of the building, have entirely given way. Behind the ruin, and apparently in the same spot on which it originally fell, lies a dilapidated image of a Hindu deity. The pedestal of this image is perfect, and lies near it. The head had been carried to Mélang some years ago by a Dutchman. On the back stone we observed an inscription, evidently in the Devanágari character, and which the Sepoy who accompanied us declared to be Sanscrit. The characters on each side were extremely distinct, but those at the back of the head of the figure were defaced.

This building is most richly ornamented with carved work, and various devices in relief are cut in the first, second, and third stories. One of these relievos represents a battle between an army of apparently polished people and an army of Rasáksa. The figures are very rudely carved and disproportioned, but in general richness of effect may be compared to the style of the ornaments at Bóro Bódo. There are a variety of processions and achievements represented in different parts, but no where could we observe any image or particular object of devotion. Along the cornices, which are most splendidly rich, we noticed birds and beasts of various descriptions interwoven. In one part a palm tree between two lambs approaching each other, in another a perfect boar, apparently led to the sacrifice.

At a short distance from this principal building, say fifty yards, stands the remains of what would appear to have been an elevated terrace of about twelve feet high. The ascent is on one side, by regular stone steps, still perfect.

Previously to leaving Málang we took sketches of two images brought in from the fort, and also of the image of a man, peculiar from the manner in which the hair was tied. At Málang, also, I received from the Tumúng'gung a small square stone-box, containing a golden lingam; this had been discovered three months before, about a cubit under ground, by a peasant, while digging for stones to build his cooking place. The lingam had originally two very small red stones within it, something like rubies; one of them was lost before it was delivered to me, the other by the party examining it.