Matheson Collection > Macleods of Dunvegan from the time of Leod to the end of the seventeenth century

CHAPTER XIX
DUNVEGAN CASTLE
The Probable Date of the Oldest Portions .
THE Keep and the Sea-Gate
Many archaeologists maintain that none of the castles
in the West Highlands date from an earlier period than
the thirteenth century, but I do not think that this
theory can be maintained. The numerous duns and
brochs are in a sense stone castles. They were probably
built by the Celts in the seventh and following centuries
as places of defence against the Vikings. When the
Norsemen began to settle in the Hebrides, they must
certainly have required strong fortresses, and I think
it is probable that the castles, lying as they invariably
do close to the sea, were built by them in the ninth and
tenth centuries.
If this is true there is nothing improbable in the tradi-
tion which says that ' Dunvegan was standing in the
time of MacCrailt Armuinn, and had been built by one
of his ancestors in the ninth century.'
To some extent etymology confirms this tradition.
There are several derivations of the name Dunvegan.
One is Norse, Dun Vikan, the Dun of the Little Bay.
This does not seem unlikely if we consider its situation
on a little bay. Another is Gaelic, Dunbheagan, the
Little Dun. This seems unlikely if we compare the size
of the keep with that of other duns in the neighbourhood.
We could only accept that etymology by supposing that
the name was given when an earUer and much smaller
dun was standing on the rock. As stated on page 27,