of incorrect observation in a man of distinguished sagacity and general accuracy. I certainly know that it is a mistake to say all, for not one arch or vault properly so-called could I discover in the temples we visited. All were formed on the same corbelling principle that I have already spoken of, and I suspect that there are none otherwise formed. The nearest approach to such a construction that I saw was in the very curious two-storied building of which Raffles (Pl. 29) gives an indifferent plate, but which I will not take up your time in describing. In some of the apertures of this, there is a sort of sham key-stone found, but it is only a sham, for it really rests on corbelled projections (See Figs. 10, and 11.) Exactly the same approach to the arch is to be seen in the arcades about the Kootub, as if the builders had heard an arch described, or had seen one, but could not remember how to imitate it. This may be seen very plainly in Mr. Beato's well-known photographs of the Delhi remains. In conclusion, it may be asked, what is the object of this paper? as, with the exception of the temple at Mundót, most of the particulars must have been given by previous English writers. Well, here is an object. In a paper which the greatest living authority on Buddhist, and on all ancient Indian architecture, Mr. Fergusson, was kind enough to attach to my description of the temples at Pagán on the Irawadi, he pointed out that that account opened a new chapter in our knowledge of Buddhist architecture. In India Buddhist remains take either the form of the Tope, of the Chaitya Hall (as he calls it) or basilica, or of the Vihára or monastery. But purely image temples were not known, unless you went so far north as Cashmere and the Salt Range of the Punjab; and the Buddhist character of these was doubted from the very fact of their being such mere temples. The Pagán buildings were such, and there could be no question about their Buddhism. Now, here in Java we have exactly similar temples, and I believe those which I have described, except perhaps the ruined piles of Loro Jongran, as certainly and unmistakeably Buddhist. But not only so. The general characters also of those temples, in Java and in Burma, have a close resemblance as well as the detail of their ornaments. The ornaments of both are of Indian origin; the form and style of both are as near as could be* in the difference of * With certain remarkable exceptions. |