286 Essay on the Avian Order of Architecture. [Sept. 



other three sides. And so far is it from being considered unfathomable 

 by the Natives, that when I directed the Maharaja's head-boatman to 

 send a small boat to the temple, he declared it would be of no use, as 

 the pond was "dry" (khushk) ! Its actual depth in November was 

 3^ feet, 2 feet only being water, and the remainder foetid mud. Vigne 

 says that it was 4 feet deep ; and as Trebeck swam to it, it is certain 

 that is is sometimes even deeper : but at no time can it exceed 5 or 6 

 feet in depth, as the banks are very low, and are besides cut through 

 for the purpose of drawing off the water for irrigation. 



6. — The Baron's estimate of the size of the temple is very nearly 

 correct. He calls it a square of not more than 25 feet, the real size 

 being 22 feet. But the actual size of the square is only 18 feet, as 

 the four porticos project two feet on each side. In the niche over the 

 northern door there still exists a squatted male figure with the Brah- 

 manical cord over the shoulder : but the figures which once adorned the 

 other niches have long since disappeared. These doorways have square 

 tops covered by pediments, which rest upon the jambs of the door, the 

 tympanum being occupied by a trefoiled niche that contains the figure. 

 This again is covered by another pediment, which also has a trefoiled 

 tympanum. The trefoiled arch rests as usual upon small pilasters on 

 each side of the door, but the pediment is supported upon bold square 

 pillars, which are attached to the building by short walls of less breadth. 

 This is an innovation, which most decidedly betokens a later date, a fact 

 already established from history : but it is also a great improvement 

 upon the earlier style ; as the boldness of the projection and the retire- 

 ment of the connecting walls afford a great and pleasing variety of light 

 and shade, which is altogether wanting in the same parts of the more 

 ancient buildings. See Plates XXI. and XXII. 



7. — The roof of the temple which is still nearly perfect, was a pyra- 

 mid resting upon a line of horizontal denticulated moulding, and 

 divided into two portions, by an ornamental band of the same moulding, 

 on a level with the summits of the four porch pediments. See Plate 

 XXII. The blankness of the upper portion is relieved by a trefoil- 

 headed niche on each side, which is remarkable for its extreme small- 

 ness and for its want of a pediment. This is also another innovation, 

 but I think not a happy one, as from the lowness of their position there 

 must have been a high unadorned bald-looking surface, left above each 



