Archaeologia Islandica - 01.01.2005, Blaðsíða 112

Birna Lárusdóttir, Gavin Lucas, Lilja Björk Pálsdóttir and Stefán Ólafsson
became clear that the midden was deeper
than the exposed layers showed. The vis-
ible deposits consisted mainly of peatash
but were slightly mixed with lenses of
charcoal and wood ash in between.
It was decided to dig a trench
straight through the center of the midden,
towards the brook, to get as clear a pic-
ture as possible of its composition and
depth as well as to retrieve a representa-
tive assemblage of finds. The trench was
5.8 x 1 m and just over 2 m deep. The
midden stratigraphy can be divided into
four different phases, based on the com-
position of layers (fíg. 6). In the follow-
ing text the lowest phase will be
described first.
I. At the bottom there were fíve distinc-
tive layers which are probably not mid-
den layers but rather surface layers of
some sort, homogenous and very com-
pact, nearly free of any ash deposits,
bones and other waste. They were
sealed by a layer of iron panning.
Because of the limited width of the
trench it was not certain whether they
belonged to the inside of a structure.
There was a pile of rocks covering the
bottom of the southem side of the trench,
making it diffícult to distinguish between
layers. This pile could be part of a wall
collapse. All the layers that belong to
this phase are more or less horizontal,
ending on an edge where land starts slop-
ing down to the brook. If they are the
remains of floor layers they would prob-
ably have belonged to an animal house
rather than a dwelling. A small amount
of fmds was retrieved from this phase,
among them a fragment of red earthen-
ware which cannot be dated accurately
Figure 6. Section drawing from trench 1.
110