DOI Artikel:
Lamb, R. C.; Rees, S. E.: Ard cultivation at Sumburgh, Shetland
DOI Seite / Zitierlink: https://doi.org/10.11588/diglit.49001#0121
ARD CULTIVATION AT SUMBURGH,
SHETLAND
By
R. C. Lamb and S. E. Rees
Rescue excavations were carried out in 1974,
in advance of airport improvements, on a
complex prehistoric settlement site on the
low-lying isthmus N of Sumburgh Head, at
latitude 590 53’ N, longitude i°x8’ W. This
site lies 1.4 km NW from Jarlshof, on land
which supported a flourishing agriculture
until the seventeenth century, when it was
overwhelmed and buried to at depth of 2-3
metres in blown sand. The excavations pro-
duced ard marks at three distinct strati-
graphic horizons, and one hundred and fifty
stone ard points illustrating all stages of
manufacture and various conditions of wear
(Rees 1979, 249, 254).
Ard marks were the earliest feature repre-
sented in the excavated area, and reappeared
in two subsequent periods associated with
stages in the development of the farmstead
buildings. The original cultivated field be-
came the site of a succession of timber struc-
tures, themselves followed by a complicated
series of stone houses. There were two
stone-built houses within the excavated area,
one of which throughout its history retained
a circular plan, the other beginning as an
oval house and later being converted to the
heel-shaped form known at many prehis-
toric sites in Shetland. Radiocarbon analysis
of material from the post-holes of the timber
structures, this material comprising heather
(probably used for bedding) and general
domestic refuse which had been swept into
the holes when the posts were pulled out,
gave dates of 1679 B.C. and 1550 B.C. Ap-
plying a correction and allowing that the
timber structures themselves were some un-
known length of time in use, these dates
suggest that the ard cultivation was under
way by 2000 B.C. Use of the ard seems to
have been continuous throughout the subse-
quent life of the settlement, the latest phases
of which correspond to the Late Bronze Age
at Jarlshof (Hamilton 1956, 18-39). Because
of the great thickness of deposit and the
complexity of the stone structures, the area
excavated was only 300 m2; the earliest ard
marks occurred in the northern one-third of
this area, having been obliterated under the
southern of the two houses by the disturb-
ance which accompanied its construction.
The earliest ard marks were cut through
thin brown soil into clean sand; the original
thickness of soil is not known as most of it
had been scraped off when the house was
built. In plan they show no immediately
apparent regularity, being a confusion of
irregularly aligned marks, many of them by
no means straight (fig. 1 and 2). In this and in
their asymmetrical V-section they resemble
those at Store Vildmose, Denmark (Nielsen
1970); they are closely but irregularly
spaced, 10-60 mm wide and up to 50 mm
deep within the sand, their asymmetrical
sections indicating angles of tilt of the imple-
ment of between 6° and 180, with no sugges-
tion of the two particularly favoured ranges
of tilt angles postulated by Nielsen (1970,
159). Many of the intersections of the marks
exhibit dragging zones identical with the
SHETLAND
By
R. C. Lamb and S. E. Rees
Rescue excavations were carried out in 1974,
in advance of airport improvements, on a
complex prehistoric settlement site on the
low-lying isthmus N of Sumburgh Head, at
latitude 590 53’ N, longitude i°x8’ W. This
site lies 1.4 km NW from Jarlshof, on land
which supported a flourishing agriculture
until the seventeenth century, when it was
overwhelmed and buried to at depth of 2-3
metres in blown sand. The excavations pro-
duced ard marks at three distinct strati-
graphic horizons, and one hundred and fifty
stone ard points illustrating all stages of
manufacture and various conditions of wear
(Rees 1979, 249, 254).
Ard marks were the earliest feature repre-
sented in the excavated area, and reappeared
in two subsequent periods associated with
stages in the development of the farmstead
buildings. The original cultivated field be-
came the site of a succession of timber struc-
tures, themselves followed by a complicated
series of stone houses. There were two
stone-built houses within the excavated area,
one of which throughout its history retained
a circular plan, the other beginning as an
oval house and later being converted to the
heel-shaped form known at many prehis-
toric sites in Shetland. Radiocarbon analysis
of material from the post-holes of the timber
structures, this material comprising heather
(probably used for bedding) and general
domestic refuse which had been swept into
the holes when the posts were pulled out,
gave dates of 1679 B.C. and 1550 B.C. Ap-
plying a correction and allowing that the
timber structures themselves were some un-
known length of time in use, these dates
suggest that the ard cultivation was under
way by 2000 B.C. Use of the ard seems to
have been continuous throughout the subse-
quent life of the settlement, the latest phases
of which correspond to the Late Bronze Age
at Jarlshof (Hamilton 1956, 18-39). Because
of the great thickness of deposit and the
complexity of the stone structures, the area
excavated was only 300 m2; the earliest ard
marks occurred in the northern one-third of
this area, having been obliterated under the
southern of the two houses by the disturb-
ance which accompanied its construction.
The earliest ard marks were cut through
thin brown soil into clean sand; the original
thickness of soil is not known as most of it
had been scraped off when the house was
built. In plan they show no immediately
apparent regularity, being a confusion of
irregularly aligned marks, many of them by
no means straight (fig. 1 and 2). In this and in
their asymmetrical V-section they resemble
those at Store Vildmose, Denmark (Nielsen
1970); they are closely but irregularly
spaced, 10-60 mm wide and up to 50 mm
deep within the sand, their asymmetrical
sections indicating angles of tilt of the imple-
ment of between 6° and 180, with no sugges-
tion of the two particularly favoured ranges
of tilt angles postulated by Nielsen (1970,
159). Many of the intersections of the marks
exhibit dragging zones identical with the