1848.] Essay on the Arian Order of Architecture. 257 



jecting ends are carved into three upright mouldings slightly rounded 

 at top and bottom and surmounted by a straight and horizontal band. 

 The resemblance which these bear to the dentils of classical architecture 

 is remarkably striking : and I suspect that these diglyph ornaments are 

 a direct imitation of the Doric, and not an accidental likeness. In 

 either case they represent the ends of beams. In the former they are 

 the ends of the beams overlying the architrave : in the latter the lower 

 set are the ends of the beams which supported the pyramidal roof, while 

 the upper set are either the ends of the horizontal ties of the wooden 

 tresses ; or of the beams of an upper floor in the roof, a construction 

 particularly common throughout the eastern hills of the Punjab. 



8. — Each of the blank sides of the upper roof is appropriately occu- 

 pied by a niche similar in form to the doorway of the temple : but the 

 head of the niche is semi-circular and not trefoiled, while the upper part 

 of the tympanum is filled by a flowered ornament. The common tre- 

 foil was however also used in this position as may be seen in the small 

 temple which crowns the isolated Srinagar Pillar represented in Plate VI, 

 as well as in the upper part of the roof of the Pandrethan temple. Last- 

 ly the top is crowned by a melon-like ornament surmounted by a con- 

 cave-sided cone, which forms a very suitable finish to the building by 

 preserving the pyramidal form which is the characteristic feature of the 

 Kashmirian architecture. 



9. — In the interior the walls are plain, but the roof is hollowed out 

 into a hemispherical dome, of which the centre is decorated by an ex- 

 panded lotus flower. Vigne* erroneously says that the " ceiling of the 

 interior is radiated so as to represent the Sun." But, in addition to 

 my experience and knowledge of Hindu decorations in general, I have 

 the testimony of the accurate Trebeck, who states that the interior of 

 the temple of Pandrethan was " quite plain with the exception of a large 

 lotus sculptured on the roof." A reference to my drawings of the 

 two roofs, which were made from measurements, will prove the truth of 

 Trebeck' s description as well as of my own. Vigne was probably mis- 

 led by his belief that the temple was dedicated to Vishnu, as Surya or 

 the Sun-god ; but the presence of the lingam as well as the representa- 

 tions of the bullNandi, decides, beyond all possibility of doubt, that the 

 temple was appropriated to Siva. 



* Kashmir, v. 2. — p. 41. 

 2 o 



