Journal, r.a.s. (oeylon). [Vol. XXIl. 
lower — southern — end widens out into a little tank formed by 
an embankment thrown from rock to rock. At the time of 
my visit (December, 1909) this intervening grass land was a 
rather swampy meadow, the favourite haunt of wild buffaloes. 
But in the days when the monastery flourished it too must 
have been submerged by the waters of the tank. With the 
large area of non-absorbent rock to collect and transmit the 
rain, it is probable that a very considerable population might 
have subsisted by means of this reservoir ; always supposing, 
of course, that no water was drawn off for irrigation purposes, 
and of this there is no indication in the way of either sluice or 
channel. In this dry land, and remote from the river, water 
conservance was of the first importance, and I am inclined to 
attribute an early date to this little reservoir. 
But the history of Tantri-malai goes back far beyond the 
tank — at least I think so — and it seems to me that in its 
various remains we can read of four different ages of occupa- 
tion, ending in tragic abandonment. It may be that in this 
I have tried to see too much in the stones and their record ; 
but the evidence, such as it is, forms the subject of this 
Paper, and those who cannot accept the conclusions may 
perhaps form others from the observations. 
1. — Earliest Period. 
In the south-eastern quarter of the area of rocks there 
stands a boulder, which by its overhanging sides forms a shelter 
convenient alike to man and beast. At the present day it is 
obviously the den of beasts, and in it are bones which show 
it has been used by leopards. 
But in prehistoric times it must have been the refuge of 
human beings, for in it I found chipped fragments of chert 
and a piece of crystal. These were not in any sense imple- 
ments, but they may have been the waste chips thrown 
away while implements were being made. For they are of 
material quite foreign to their surroundings, such as could 
hardly have come to be where they were by accident. Of 
their age, I am quite incompetent to form even an estimate ; 
but to their newness I can put a limit — the limit is 2,000 
years ago. The cave where they were found has a drip -ledge 
