THE SEVEN PAGODAS. 



141 



thing of this kind might account for the destruction of its 

 interior. Its central shrine and two adjacent cells have been 

 cut out and thrown into one chamber ; the three large sculp- 

 tured tableaux have been carefully obliterated or " dished," 

 together with the dvarapal warders ; the cankha and cakra 

 of Vishnu have been cut on the end walls of the terrace, and 

 a colonnade, the commencement of a pillared hall like that 

 built in front of the pastoral bas-relief No. 15, begun ; but 

 the scrap of Saiva inscription has been allowed to remain 

 intact on the floor. 



Against the theory of the total overthrow and expulsion 

 of Siva, however, we find that many of the tableaux contain- 

 ing Siva as the principal figure, are still in their original sites 

 apparently, although the liiiga that seems to have been their 

 invariable accompaniment has been removed ; and in one 

 instance, the Atiranaeandesvara at Saluvankuppam No. 58, 

 we see three copies of the Siva tableau and three lingas still 

 in situ with them, all in one shrine. These caves may be 

 due to a Saiva revival, and in the latter instance (No. 58), 

 the shrine with its emblems and altar-pieces may have escaped 

 spoliation from having been covered by the sand, the level of 

 which is now about as high as the roof of the mandapa, which 

 can only be entered by first excavating the flood of sand 

 which pours in and buries the whole with every sea-breeze 

 in the dry season. 



The Tamil inscriptions (Nos. 35, 6£), and 61) of the 

 eleventh, twelfth, and thirteenth centuries, show that the 

 Vaishnava sect was dominant during that period (1073 to 

 1235 A.D.) 



The older Sanskrit inscriptions show conclusively that Si\a 

 was supreme here in the time of the Pallavas, who are known 

 from other sources to have ruled this part of the country 

 from Kanci (Conjeveram), of which Mamallapuram is said 

 to have been the seaport during the early centuries of the 



