If this was not the site of Uxellodunum, we may pretty safely believe it to
have been that of some important _oppidum_ of the Gauls. A circumvallation
there could never have been in a strict sense, for where the plateau rests
upon high calcareous walls there was no need of a fortification. But
elsewhere, where the position was accessible from the valley, it was
protected by a strong wall. On the northern side this rampart can be
followed for a considerable distance without a break. In one spot the soil
which has collected about it has been dug away, leaving the masonry bare.
It is not composed of loose stones of various sizes, like that of the
Celtic city at Murcens, but of small flat stones neatly laid together, with
layers of mortar between; a circumstance that sets one conjecturing and
doubting. The wall appears to have been six or eight feet thick. The line
of it now only rises very slightly above the edge of the plateau.

I met a peasant who owned the highest part of the tableland, and who
managed to grow crops upon it. Near his cottage he pointed out the remains
of an ancient structure, which he called the fort. The masonry was of the
same character as that of the wall. Near to it were fragments of ancient
pottery and tiles, which he had dug out of the ground. The tiles were
very heavy and flat, with turned-up edges, so that they could hang one to
another. There were holes, too, for the nails which held them to the roof.
Thrown on one side were human bones, which had from time to time been
turned up by the plough. The peasant told me that his father, while digging
rather deeply, had found a skeleton wearing a bracelet and part of a
helmet. A visitor to the Puy d'Issolu, many years ago, was allowed to take
these remains away, together with a quantity of iron arrow-heads, on his
promise to come back and pay 600 francs for them. He never came back.

The view from the Puy takes in an immense expanse of the solemn Limousin
country. To the south is the stone-strewn Quercy, while to the north and