Archaeologia Islandica - 01.01.2004, Qupperneq 87

ICELANDIC FARMHOUSE EXCAVATIONS: FlELD METHODS AND SITE CHOICES
removed and a horizontal surface created
where originally the floor had been on a
slight angle. The floor deposits closest to
the walls were probably thin and difficult
to detect whereas the floor deposits along
the central axis of the building are quite
distinct. Bruun had clearly instructed his
workers not to cut through floor layers or
other deposits, nor to remove stones or
other intemal features. As far as can be
seen this was heeded but where the
deposits were ephemeral the workers
could only take their cue from more dis-
tinct features and dig "blindly" the areas
in between. Once the removal of the soil
infilling the structure was completed
Bmun will have started his actual inves-
tigation. This consisted of a more careful
investigation of certain features, hearths,
pits and doorways, which were more
carefully excavated and recorded. Bruun
also established the thickness of the walls
by recording the depth of the entrances as
well as observing this in a couple of
trenches placed through the walls. On
the whole however he did not examine
the outer sides of the walls or the area
surrounding the building. At Gásir sev-
eral test trenches were dug from the base
of the main excavation trench, no doubt
mainly to establish the thickness and
nature of underlying deposits. These are
as a mle quite small, usually rectangular
and typically 50-70 cm wide. At both
Hofstaðir and Gásir a number of trenches
were also dug outside the main excava-
tion areas. At Hofstaðir most of these
test trenches were placed adjacent to the
main excavation area, where surface
irregularities suggested there might be
structures undemeath. Bmun's pits and
trenches failed to establish if these actu-
ally were buildings. These trenches -
some of them quite large - are very dif-
ferent from the main excavation area
which seems to have been dug in a very
orderly manner and with some confi-
dence. The adjacent trenches were clear-
ly not laid out on the expectation that
buildings would be found and are in
some cases extremely irregular, reaching
different depths, with small and large off-
shoots and even shallow tunnels showing
lateral burrowing. These trenches reflect
considerable confusion, more or less
indiscriminate digging in search for some
foothold to base a strategy on. When
none was found the trenches were aban-
doned and the area got summarily treated
in the reports.
Inside the main excavation area of the
Hofstaðir long house, Bmun created a
detailed plan based on accurate measur-
ments, showing walls, entrances, hearths
and pits, surface deposits and the location
of artefacts. (Fig. 4) The retrieval of arte-
facts was apparently not a major concem
- as is evidenced by the substantial num-
ber of objects retrieved in later cam-
paigns from Bruun's backfíll, but he
meticulously described those he did
recover and marked their find spots on
the plan. In addition to the main plan
Bmun drew larger scale plans of individ-
ual features, created a schematic eleva-
tion showing a cross-section of the long-
house, made sketches of the site and took
a series of photographs. His excavation
archive - preserved at the National
Museum in Copenhagen - contains some
fieldnotes and preliminary sketches but
on the whole most of the information
85