n crannogs, as was demonstrated by my
investigations at Lochlee and elsewhere. {28a} It would be still more
necessary in a substratum of timbers that was intended (as will be
afterwards explained) to bear the weight of a superincumbent cairn.
Underneath the layers of horizontal woodwork some portions of heather,
bracken, and brushwood were detected, and below this came a succession
of thin beds of mud, loam, sand, gravel, and finally the blue clay
which forms the solum of the river valley. {28b} The piles penetrated
this latter, but not deeply, owing to its consistency; and so the blue
clay formed an excellent foundation for a structure whose main object
was resistance to superincumbent pressure.

Outside the circle of piles there was, at a distance of 12 to 14 feet,
another wooden structure in the shape of a broad ring of horizontal
beams and piles which surrounded the central area. The breadth of
this outer ring was 7 feet, and it consisted of some nine rows of
beams running circumferentially. Beyond this lay scattered about some
rough cobble stones, as if they had fallen down from a stone structure
which had been raised over the woodwork. The space intervening
between these wooden structures was filled up in its eastern third
with a refuse heap, consisting of broken and partially burnt bones of
various animals, the shells of edible molluscs, and a quantity of
ashes and charcoal, evidently the debris of human occupancy. On the
north, or landward side, the outer and inner basements of woodwork
appeared to coalesce for 5 or 6 yards, leaving an open space having
stones embedded in the mud and decayed wood, a condition of things
which suggested a rude causeway. When Mr. Donnelly drew my attention
to this, I demurred to its being so characterised owing to its
indefiniteness. At the outer limit of this so-called causeway, and
about 25 feet north-east of the circle of piles, a canoe was
discovered lying in a kind of dock, rudely constructed of side stones
and wooden piling. The canoe measures 35.5 feet long, 4 feet broad,
and 1.5 foot deep. It has a square stern with a movable board, two
grasping holes near the stem, and three round perforations (2 inches
in diameter) in its bottom. On the north-west border of the
log-pavement a massive ladder of oak was found, one end resting on the
margin of the log-pavement a