The principal group of temples at Brambanan is or has been that of Loro Jongran. They are so utterly ruined that, even when very near them, you scarcely make out anything but great cairns of stones heaped together. It must have been a tremendous earthquake that produced such ruin. Closer examination shows among the chaos many fragments of rich mouldings and sculpture, and some of the basements, highly adorned with vases and festoons, are tolerably perfect. The largest pyramid of ruin you ascend to a height of some five and thirty feet, and find entrances to cells opening to the four cardinal points. The most remarkable circumstance about this ruin is that three of those cells contain very fine and purely Hindu figures. That to the north is an eight-armed goddess standing triumphantly on a dead buffalo and grasping in one of her four left arms the curly wig of a little monster. It is evidently the same subject that is represented in Moor's Hindu Pantheon, plate 35, and therein entitled Durga or active Virtue slaying Maheshásura or Vice personified.” This is the figure called by the Javanese Loro Jongran, and giving its name to the temple. It appears to be common among Javan remains, as you will find half a dozen in the plates to Raffles. To the west is Ganesha with his elephant head; and to the south a fine Jupiter-like bearded Siva with the trident.* The fourth entrance was obstructed by fallen stone, and I was too tired to attempt to crawl in. It is to the east, and probably was the entrance to a central chamber. From the height at which these cells stand they must evidently have formed an upper story of the temple. They are carved on great slabs standing against the wall without being attached to it, and I have some doubts if they are the original occupants. The cells otherwise seem exactly parallel to those of the cruciform Buddhist temples already described, and to which class nearly all the other Brambánan temples appear to belong. These are, however, the most ancient, as we may guess from their utter ruin. The other and more perfect temples cannot have been standing when the tremendous earthquake occurred which rattled these down into such a chaos. They may therefore have been the remains of a more ancient Brahminical sanctuary, as we know from the travels of Fahian that in his time (the beginning of the 5th century) Brahmins existed in

There is an engraving of this in Crawfurd. Indian Arch. II. pl. 27.

Java, but Buddhism did not.* I do not take up more time with these, as there is a full description of them inserted in Raffles.

The only other group of temples that I will notice is that called Chandi Sewu, or the Thousand Temples, also described in Raffles. The group consists of one large central cruciform temple, as usual, with three blind porches and a fourth on the east giving access to the interior. But this is surrounded by four successive squares of small cells or temples, the outer square of which is upwards of 500 feet in the side. Many of these small cells are obliterated, and without more time than I had it would be difficult to say accurately their original number. A plan however is given in Raffles, which shows that the inner square has 8 temples to the side, the next has 12, and the two outer squares 20 and 22 respectively. I note this, because I suppose its accuracy may be assumed, and because its discrepancy from my own notes shows how apt a hurried notice in such matters is to err, even when there is a desire to be accurate.† My notes mention only 3 squares, containing respectively 8, 12 and 24 temples to the side, and I took some pains to allow correctly, by pacing, for the intervals where numerous temples were obliterated. However, I am amused to find that a man who probably had no such plea of haste as I, and is an observer by profession, Dr. F. Junghuhn, the author of the chief physical account of Java, in a paper on the same subject as my own, declares that there are 176 in the 4 squares, respectively 28, 36, 52, and 60. The whole number will be, according to Raffles's plan, in the four squares 240, besides four pairs placed intermediately between the 2d and 3d squares and flanking the avenues of approach.

The central temple is greatly shattered, and the image (a great Buddha I doubt not‡) which it contained, is gone. It stands with its porches on a terrace slightly elevated. There are no figures

"On fut ainsi pendant quatre-vingt-dix jours; alors on arriva à un royaume nommè Yepho thi. Les hérétiques et les Brahmanes y sont en grand nombre, il n'y est pas question de la loi de Foe."-Relation des Royaumes Bouddhiques, 360.

I may apologize for such inaccuracy by the fact that I was only recovering from a long illness, and was incapable of exertion in a hot sun.

What Crawfurd says in speaking of this is misleading: "Each of the smaller temples had contained a figure of Buddha, and the great central one, consisting of several apartments, figures of the principal objects of worship, which, in every case that I have had an opportunity of examining, have consisted of the destroying power of the Hindu triad or some of his family." The central temple of Chandi Sewu was empty then as it is now, and this merely states* a foregone (and I believe quite mistaken) conclusion.

* Indian Archipelago, II. 196.

sculptured upon it, the decorations apparently having been panels of diaper work chiefly. I give a sketch of the beautifully executed. doorway, chiefly on account of the singular ornament at the lower angle of the door-frame, representing what I must call for shortness an arabesque sea monster, and exactly similar to a constant ornament over the openings of the great Pagán temples on the Irawadi. (See Fig. 8). It is found also in Southern India. The small cells or chapels are each about 10 or 11 feet square. Their walls are carved with mythological figures in bas-relief, and each has been crowned by a small dagoba of the genuine Buddhist pattern. They all open outward, except the 3d row which stands back to back with the outer row,* and each has contained a cross-legged Buddha, of which some remain. There are groups of modern temples about Calcutta and Burdwan, somewhat similar in general arrangement. Mr. Fergusson appears to doubt whether he should not class this as a Jain temple. I know little about Jains, but will answer for it that any Burmese would find himself at home in it as a monument of unmistakeable Buddhism.

Guarding the outer end of the avenue, by which we approached, are two gigantic warders, standing or rather kneeling, about 9 feet in height, with club grasped in the right hand, and a snake which twists round the body grasped in the left, with crisped hair and great staring eyeballs; also closely resembling the similar figures in marble and in stucco which are so common in Burma. (Fig. 9.)

The central temple is apparently that which is represented in the plates to Raffles† as the "Great Temple at Brambanan," whilst one of the cells is represented,‡ as one of the smaller temples at Brambanan." It strikes me, however, that they are both very inaccurate, and the elaborate restoration of the great temple which is given in Pl. 40 is, I have no hesitation in saying, preposterously improbable.

In conclusion, as it is a point of some interest, I may note that Mr. Crawfurd says,§ that, though the interior vault of the temples is a false one, "the builders of Brambanan had possessed the art of turning an elliptical arch and vault, for the entrances or doorways are all arched, and the roofs all vaulted." I think this is another instance

* This from Raffles's plate.

† 2d edition, Pl. 39.

Pl. 41.

§ History of Indian Archipelago, II. 196.