173 
the length of the solar year, the amplitudes, or rather 
their differences, are directed to be determined by marking 
the sun's position at rising, on a horizontal circle of consi- 
derable magnitude. Besides which, the astronomical build- 
ings at Benares, Delhi, and Jyepoor, though all probably 
only a few centuries old, indicate the purposes for which 
they were required ; and these purposes we have seen are 
not of very recent date among the Hindoos. The present 
buildings therefore may be of modern origin, but we cannot 
expect them to be ancient in a country, where, from the 
perishable nature of the materials employed in architecture, 
cities have succeeded cities, leaving scarcely any traces 
of their sites. These structures may, however, and most 
probably have, only succeeded others of a similar kind, but 
of much older date. That of Benares has been described 
by Sir R. Baker in the Philosophical Transactions for 1775 ; 
those of Jyepore and Delhi I have not seen described, but 
at the latter place, the large and lofty circular roofless 
building, or rather wall, pierced with horizontal rows of 
openings like windows, would seem as if intended to be 
employed for the purposes of a horizontal circle. 
Mr. J. Bentley, in bis papers in the Asiatic Researches, 
vols. vi. and viii., has inferred from the minimum errors in 
the Surya Siddhanta, that it was composed about A.D. 
1067, and this is confirmed by other circumstances (v. Prin- 
sep. Tables, p. 29) : he also concludes that Varaha-mihira 
was the author of this work. But by reasoning entirely 
analogous to that with respect to the Surya Siddhanta, he 
was led to conclude that the system of astronomy, composed 
by Brahma-gupta, may be referred to A.D. 536; and this 
date nearly agrees with that assigned to the Brahma- 
Siddhanta by the Hindoos themselves. Now Mr. Colebrooke 
has observed, that Brahma-gupta, in a work of acknowledged 
authenticity, quotes by name Varaha, who flourished 
