on of some holed
tree-roots in Mr. Bruce's paper to the Scottish Antiquaries: {30}

'Midway between the centre and the outside piles of the structure what
looked at first to be tree-roots or snags were noticed partly imbedded
in the sand. On being washed of the adhering soil, holes of 12 inches
wide by 25 inches deep were found cut in them at an angle, to all
appearance for the insertion of struts for the support of an upper
structure. On the outside, 14 inches down on either side, holes of 2
inches diameter were found intersecting the central hole, apparently
for the insertion of a wooden key or trenail to retain the struts.
These were found at intervals, and were held in position by stones and
smaller jammers.'

The outer woodwork formed the foundation of another stone structure,
of a horseshoe shape, having the open side to the north or landside of
the tower, which doubtless was intended as a breakwater. By means of
the ladder placed slantingly against the wall of the central stone
building access could be got to the top in all states of the tides.

The people who occupied this watch-tower ground their own corn, and
fared abundantly on beef, mutton, pork, venison, and shell-fish. The
food refuse and other debris were thrown into the space between the
central structure and the breakwater, forming in the course of time a
veritable kitchen-midden.

Besides the causeway on the north side, Mr. Bruce describes 'a belt of
stones, forming a pavement about six feet wide and just awash with the
mud,' extending westwards about twenty yards from the central cavity,
till it intersected the breakwater. {31} These so-called pavements
and causeways were probably formed during the construction of the
tower with its central pole, or perhaps at the time of its demolition,
as it would be manifestly inconvenient to transport stones to or from
such a place, in the midst of so much slush, without first making some
kind of firm pathway. Their present superficial position alone
demonstrates the absurdity of assigning the Dumbuck structures to
Neolithic times, as if the only change effected in the bed of the
Clyde since then would be the deposition of a few inches of mud. At a
little distance to the west of these wooden structures there is the
terminal end of a modern ditch ('the burn' of Mr. Alston), extending
towards the sho