28 
I now can recollect, and about six feet wide. There seems to have 
been some attempt at elegance in the construction of this one. The 
centre apartment had three more branching from it, at right angles, 
which, with the passage forming the entrance, gave it the shape 
of across. This was evidently a subterraneous house. Some of 
these caves have sustaining walls of dry stone works, to confine the 
sides, and support the flags which form the ceiling. Some of them 
are excavated into the hard gravel, with the flags resting on no other 
support; and so low, that you can only sé erect inthem. That is, 
from three to four feet from the floor to the ceiling. I have not seen 
any higher than four feet. If my conjecture be true, they seem to 
have continued the use of caves, even after they had adopted round 
forts; for in many of these forts are found similar caves. In one 
which I opened on my own demesne, I found two, exactly like those 
I had explored, where no fort at all stood. So similar were they 
to each other in construction, that no doubt remains upon my mind, 
that they were executed by the same people, and for the same pur- 
pose. I believe them to have been used as dormitories, and depots 
for their most valuable effects; as places of shelter from wild beasts ; 
and as permanent residences, particularly for their women and chil- 
dren. The tradition of the country makes them granaries ; but for 
granaries they could never have been intended, as it would have 
been very difficult to convey grain into them, through long and nar- 
row passages, not more than two feet square ; and, for granaries, 
where was the necessity of separate apartments inside of each other? 
It is also scarcely credible that corn, at so early a period, could 
have been at all in use; and still more incredible, that this country 
could have produced it. 
It may not be uninteresting, and not entirely foreign to my subject 
to mention, that on the adjoining farm to Staigue, within less than 
