252 Essatj on the Avian Order of Architecture. [Sept. 



excepting only that upon the Takt-i-Suliman. The wonderful temple 

 of Marttand, as the Hindu historian himself calls it, with its lofty roof 

 and highly ornamented walls, was built either in the third or the fourth 

 century ; and as its style differs fully as much from that of the plain 

 low-roofed temple of Bhaumajo, as the style of the Parthenon does 

 from that of the temples of Pcestum, a considerable interval must have 

 elapsed between the dates of their construction. The building of this 

 temple cannot therefore be placed much later than the commencement 

 of the Christian era. 



4. In plate X. I have given a plan and an elevation of this temple : 

 together with plans of the caves of Bhaumajo and of Bhima-Devi. 

 The latter is a straight narrow fissure, 1G0 feet in length, which gradually 

 widens out towards the end into two small chambers, from 1 6 to 20 feet 

 across, and from 12 to 15 feet in height. In each of these there is a 

 shapeless waterworn stone, which is considered holy by the Hindus. The 

 larger cave of Bhaumajo is 55 feet long, 25 feet broad, and from 10 to 

 20 feet in height. Baron Hugel* erroneously states that this cave is 

 about "20 feet long and 12 feet high and broad," but these dimensions 

 must certainly have been recorded from memory, for mine are given 

 from measurements made by myself. Moorcroft did not visit these 

 caves, and Vignef was deterred from entering by the stench of innu- 

 merable bats. Before I visited it I had all the bats turned out, and 

 their dung removed : but still the task of measurement was rendered 

 extremely unpleasant by a villanous smell, and still more by the myriads 

 of bugs which were swarming over the glistening walls of the temple. 



5. There are numerous dressed stones in the interior of the cave, 

 and there are also two low stone walls flanking a narrow pathway, which 

 leads to the steps of the temple. The same arrangement I have 

 observed in most of the Buddhist temples in Ladak and in Upper Kana- 

 war : and I am therefore disposed to consider this building as a Bud- 

 dhist structure. The existence of the numerous excavated cells at a 

 short distance from the cave would seem to prove the correctness of 

 this appropriation, as they appear to have been the usual accompaniments 

 of the monastical institutions of the Buddhists ; being destined either 

 for the reception of figures or for the dwellings of the priests. 



* Eng. translation, p. 36. 

 t Kashmir, v. 2 — p. 4. 



