had had time to dry. The spot where they were found is evidently that where they were baked; for, besides the failures, there was also a large accumulation of broken pieces, all mixed in a deep deposit of ashes and the other refuse of a potter's kiln.
Most natives who have seen them think they were meant to hold either gunpowder or oil, which is what the shape suggests; but the material, on account of its weight, seems unsuitable for such a purpose, if the flask was to be carried about on the person; while the pointed bottom makes it awkward for storing. The idea has also been hazarded that they were meant to be filled with gunpowder and then exploded as a kind of fire-work; but, if this were their object, there would scarcely have been so much trouble spent on their ornamentation. A third theory, which has found much favour on the spot, but which at the time I was inclined to reject as altogether untenable, is that they were intended to form a balustrade for a balcony or the roof of a house. At first my own impression was that they were not at all likely to be of the same age as the bricks. The site might have been originally occupied by a fort or a monastery and then deserted for centuries before the potters came and set up their kilns on it, making use for their houses and workshops—of any old building—materials they happened to light upon. But finally I came to the conclusion that the balustrade theory was not so very far wrong, and that these curious objects were manufactured in such numbers in order to serve as finials for miniature Buddhist stupas. The dedication of such votive memorials was a recognized duty on a pilgrimage, and it would obviously be a convenience for worshippers to have an establishment for their manufacture and sale in immediate connection with the shrine. This view is strongly confirmed by the discovery on the same spot of what is unmistakeably a finial. It is of similar configuration and has a similar orifice at one end, which in this case is clearly intended for the admission of a supporting rod. But later again I found a circular flask which is of the same material and of equal weight and is ornamented in exactly the same style. It is, however, easy to grasp in the hand, and apparently was intended to hold oil or some similar fluid, for pouring out drop by drop. Thus the only definite conclusion at which it is safe to arrive is that various articles for different uses were turned out at the same factory, all being characterized by ornamentation of a peculiar local pattern.
