The Antiquities of Orissa v1.c9.19

v1.c9.19

[42] They all begin with a thick rectangular projecting base-moulding called *upāna*, corresponding exactly with the plinth of the European orders. Over it occurs either a thin receding tile of the same description, or a quirked cyma-recta in a reversed position, covered by a thin tile. The plinth and the tiles are invariably left plain, without any carving on their face, but the upper edges of the tiles are frequently broken by small projections, shaped like lancet-heads, and placed at long intervals. (See woodcut No. 20) At Benares these are called *rāmarekhās*, and are most extensively employed in the ornamentation of all flat, horizontal mouldings. In Orissa, they are used in conjunction with both flat and rounded members, and project either upwards, or downwards, or both upwards and downwards at the same time, serving the purpose of crests in the former, and of cusps in the latter, position. In upper mouldings they are richly carved. In plain work they spring from the moulding and form an unbroken part of it, as shown on the upper side of the woodcut; but when decorated with carving, they are entirely independent. Some plain ones are also so represented as to appear separate from the band on which they are placed. (See the lower part of the woodcut.) On the ledges of sloping roofs and cornices they serve the purpose of finials. Everywhere they constitute the most marked and characteristic feature of the ordinance, and there is not a single ancient temple in Orissa where they are wanting. When ornamented, they have generally a scroll or beaded border, with men, animals, or flowers, in the middle. Sometimes they are so shaped as to look like what in Gothic architecture are called crockets. Ordinarily in general appearance they are very like miniature Saracenic battlements, and if enlarged and brought into close proximity, they would no doubt serve the purpose of those ornaments. This arrangement is actually seen on the outer wall of the Puri enclosure, and there the lancet-heads are sufficiently enlarged to form regular battlements. The walls, however, cannot be taken to be of the same date as the temple, as the priests have a tradition that the Marhattas rebuilt that part of the wall which has large battlements, and the other portion may, likewise, be the result of a previous attempt at improvement. Among the ruins at Koṇārak I noticed several large battflements, each formed of a single stone, measuring 2′ 2″ by 1′ 8″ and having a flat band round the edges in front; but they were not *in situ*, and I could not satisfy myself as to the position they had formerly occupied, It is well known that the Marhattas transferred larger quantities of building materials from Koṇārak, in the middle of the last century, for the erection of the refectory or Bhogamaṇḍapa at Puri, and it is possible that the battlements were carried at that time for use at Puri. They are exactly of the same pattern as those on the eastern wall of the Puri enclosure. The doubt regarding the origin of the Puri battlements is, however, immaterial to the question at issue, viz. as to whether the Hindus ever built battlements on the top of their walls and cornices, for the triangular and pyramidal battlements to be met with at Udayagiri and Sāñcī (*vide ante* p. [17], woodcuts, Nos. 1 and 2) leave no room for doubt on the subject; and the transition of the straight lines of triangular battlements of Udayagiri into the curvilinear ones of Puri is so easy, that a Saracenic theory is not at all needed to account for it.