Vol. VIII, No. 9.] Bhatta Bhavadeva of Bengal. 339 
[WV he 
o add a rough ground plan of the whole structure for 
facility ve understanding the general plan. 
The s temples of Orissa have yet to be studied 
scientifically ‘ 80 far as I have been able to examine them 
d 
under three groups. These groups, if named after a aera oen 
specimen of each class, would be somewhat as follow 
(i) The Parasuramesvara group, iccaeeats the 
oldest, and architecturally very interesting. It 
includes, in uvanesvara town, the Parasu- 
ramesvara, the Sisiresvara, the Kapalini (mis- 
called Vaitala Deula), and possibly the Utta- 
resvara; at Mukhalingam, District Ganjam, the 
Somesvara, and so on. 
(ii) The Lingaraja group, numerous and architecturally 
nesvara town, the Lingaraja and the "Yameévara : 
in taal ee J agannatha and the jal anda 
Konarak, the Sun-temple, and s 
(iii) = Brahmesvara group, found alricak exclusively i in 
anesv. 
includes the Brahmesvara, the Siddhesvara, the 
Kedaresvara, and probably the Rajarani Deula. 
Architecturally, the Ananta-Vasudeva falls in the second 
carvings, it i el itation of the Lingaraja, one might 
sa smaller edition thereof. Unlike most temples in Bhu- 
vanesvara, it fa ever, the west (and not the east) ; 
and it is the ve temple i in that town dedicated to Vasudeva 
and not to Siv 
C.—His ANCESTORS, 
The inscription has luckily preserved for us a list of 
Bhavadeva’s ancestors up to the seventh generation. This 
list is shown on the heehee page in the form of a genealogical 
hart. 
! The District Gazetteer of Puri, pp. 54, 241. 
