Page:Archaeological Journal, Volume 5.djvu/231

 UPON THE ARCHITECTURE OF WESTERN INDIA. 177 -I2i — 2i Hi 5r J^ /3 u n 13 Eig. 7 is a pilaster in a cell adjoining the _ temple. In this the tall pedestal has been banded, which is an iniprovement upon fig. 6. Fig. 8 is in a cell which forms part of a sqnare cave. The shaft, neck, and capital are fluted. The neck is double. The foliage, pendant in fig. 6, is here raised, and becomes the true ornamental abacus called phul- band. Fig. 9 represents one of two balustrade columns which attract the eye of the visitor as he ascends to the first platform. The excavation of which these columns form the entrance has been left incomplete. The in- tention of the monks had been to make another temple, but they have only just made a beghming. These columns them- selves are unfinished. The abacus. A, has not been caned, B has not been moulded, the band round the cushion capital is plain. The neck is not cylindrical, but narrow be- low, and is separated from the shaft not by a single or double roll but by a canal. In the varieties of this order above described, the outline of the shaft is a curve of inflec- tion, but this has the simple parabolic cur- vature of the moulding. Fig. 10 is the outline of an imperfectly finished pillar, in a square cave, which has been conunenced at the most remote part of the monastery, beyond the Asoka grove, which occupies the ancient artificial lake. The shaft is exactly similar to that of fig. 9, except that the upper half of the neck is converted into two mouldings, the canal E is omitted, and thus the lower part of the neck is confused with the shaft, whose outline then becomes a curve of inflection. This comparison seems to shew that there is some intimate connection between the balustrade shaft and the parabolic moulding. The columns (fig. 11) of the Elephanta cave belong also to the balustrade order. The abacus is a thin phid-band. The 1i J3 n&