No. 63. — 1910.] TANTRI-MALAI. 
81 
It is as plain as any written history could be. It is an 
instance of the true value of archaeology as a stimulus to the 
imagination. In what is now a solemn jungle there must have 
been a tragedy. Tantri-malai lies right in the track from 
Mannar to Anuradhapura ; and it was at Mannar that the 
Tamil invaders used to land, and to Anuradhapura that they 
used to march. 
It remains to find out the date of these great images. Mr. 
Parker places them, and indeed the whole of Tantri-malai, 
in the second or first century B.C., or at the latest in the first 
century a.d. Mr. Bell , Archseo logical Commissioner , attributes 
them to the time of Parakrama Bahu the Great, in the 
twelfth century a.d. Fortune placed in my way the means 
of deciding this wide difference of opinion ; for I found an 
inscription which seems to settle the point. 
On a rock adjacent to that on which the dagaba and great 
images are, there are three buildings which by their more 
modern style and infinitely superior preservation palpably 
belong to a period quite distinct from that of the other ruins. 
One is a square building on a rock, having a circumambient 
verandah of broad dressed slabs of stone ; another is a chamber 
hewn from the rock and walled up in front with cut and dressed 
stone ; and the third is a little box -like structure on the 
summit of the rock in which the cave is hewn. Its stonework 
resembles that in the cave below, and is neat work adorned 
by slight pilasters in relief. Mr. Bell in his Report appro- 
priately calls ijb a " campanile," adding its traditional name, 
Pot-gula, or " library." 
These three buildings are quite obviously of the same date 
and style. From the evidences of sudden cessation of work, 
never afterwards resumed, and especially from the half -cut 
steps, it may be confidently assumed that no important stone 
work was undertaken at Tantri-malai subsequent to the 
dispersal of the masons. The images cannot therefore be of 
earlier date than the three buildings just described, for these 
exhibit much skill in stone work, and the hewing out of a 
rock chamber with a cubical content of 1,170 ft. (22 ft. 7 in. 
by 8 ft. by 6 ft. 5 J in.) indicates that both time and labour 
were available. We can therefore roughly date the images 
a 36-10 
