'CELTIC' FIELD GROUPS
See introductory notes to 'Celtic' Field Groups of Central Dorset (Dorset III, 318).- The
Field Groups of North Dorset are numbered consecutively with those of Dorset III.
The 'Celtic' fields described in Dorset II and Dorset III
survive in condition varying from well-preserved earthworks to the merest traces on air photographs. Those
in the area covered by the present volume have almost
without exception been severely damaged, chiefly by
later cultivation.- From the mediaeval period onwards,
ploughing has made inroads into the 'Celtic' fields,
altering or obliterating them, a process of destruction
which has reached a climax in the last few decades under
the impact of the efficient machinery of modern arable
farming. Large blocks of fields have been totally
flattened and are now scarcely visible on the ground,
though they still survive on air photographs. But even
on photographs, particularly those taken in recent
years, 'Celtic' fields are often only faintly discernible—
the last vestiges of a lengthy process of attrition.
The fields all lie on the Chalk and can be detected over
substantial areas of former downland (see general map of
prehistoric and Roman sites in end-pocket). Their
distribution, like that of so many earthworks, coincides
to a notable extent with the areas of former downland
as recorded, for example, in O.S., 1811; areas in which
post-Roman ploughing has usually been of limited intensity, if it took place at all, until comparatively recent
times. This suggests that the 'Celtic' fields were formerly much more widespread than now appears and that
many have been obliterated in areas where arable
activity has been intensive and continuous. Where the
fields are certainly or probably related to settlements,
the latter appear to be of the Iron Age or Romano-British periods. The small square fields forming an irregular pattern on Pimperne Down (Group (73)) are
associated with an early Iron Age enclosure, Pimperne
(15). A series of elongated, but somewhat irregular,
fields and closes in the same group occupies the area
between the probable Iron Age enclosure, Pimperne
(18), and the double-enclosure settlement of Iron Age
and Romano-British date, Tarrant Hinton (18).
'Celtic' fields (Group (70)) are integrated with the Iron
Age and Romano-British settlement of Buzbury Rings
and, in part, with a series of linear boundary dykes
(Tarrant Keyneston (17–21)) in their vicinity. The
fields of Group (74) adjoin the Iron Age and Romano-British settlement on Tarrant Hinton Down (19), and
at their S. limit are cut by a linear dyke, Tarrant
Launceston (16). The Romano-British settlements on
Chettle Down (Chettle (14)) and on Blandford Down
(Tarrant Launceston (14)) lie among 'Celtic' fields
(Groups (72) and (75)), but direct relationship can now
be established only at the former settlement.
Among the 'Celtic' fields, virtually no trackways have
been found leading to and from settlements, except on
Manor Hill, Tarrant Gunville, where nearly half a
mile of track associated with a probable settlement is
visible as a soil-mark among contemporary fields (Group
(76)). On Monkton Down (Group (72)) and on Gunville Down (Group (73)) two small areas of elongated
and notably rectangular fields are conspicuous; their
form suggests that they were laid out in the Roman
period, rather than earlier.
Group (69): Tarrant Crawford. Air photographs
indicate the former existence of 'Celtic' fields, now
almost totally obliterated by ploughing, on the spur top
S.E. of Tarrant Crawford (around 926030).
Air photographs: CPE/UK 1893: 3091–2.
Group (70): Keyneston Down, Rawston Down
and Luton Down (Langton Long Blandford, Tarrant
Keyneston, Tarrant Monkton, Tarrant Rawston). 'Celtic'
fields (916047—914074—936075—935060), now largely
flattened by ploughing, and discontinuous, cover some
500 acres on the top and sides of the interfluve between
the Stour and the Tarrant, around the Iron Age and
Romano-British settlement of Buzbury Rings (Tarrant
Keyneston (16)).-They incorporate at least one other
settlement, Tarrant Rawston (4), and are associated
with a number of dykes, Tarrant Keyneston (18–21).
See map in end pocket.
Traces of a number of fields, up to 60 yds. wide and of uncertain length, are detectable in Langton Long Blandford just
W. of Buzbury Rings, around 915060. Clearly they are laid
off the linear dyke (Tarrant Keyneston (18)), between which
and dyke (19) are at least four large rectangular fields, measuring
up to 85 yds. by 130 yds.
To the S. of Buzbury Rings heavy ploughing has totally
flattened the fields, but traces are visible as far S. as 916047. Air
photographs suggest that fields have been relaid near the linear
earthwork, Tarrant Keyneston (20); some appear to be related to it, others to be cut by it.- A ploughed out hollow-way
or dyke, apparently integrated with 'Celtic' fields, runs N.E.
from 920050 to a roughly circular enclosure about 130 ft. across
'Celtic' field lynchets, but few complete fields, extend E.
from Buzbury Rings towards Tarrant Rushton, along the steep
slope N. of Keyneston Down, as far as 933060. Air photographs
show traces of former 'Celtic' fields on Rawston Down, especially around 924067. Farther N., in Tarrant Monkton, traces
are visible on Luton Down, around 915071, and also over an
extensive area around, but mostly N. of Luton Drove, between
921074 and 936076. Within this area, soil-marks indicate the
existence of a probable settlement at 913076, comprising at least
one small enclosure approached by a long, narrow funnel and
associated with further ditches.
Air photographs: CPE/UK 1934: 1129, 3155–8, 5156–7;
CPE/UK 1893: 3068; 58/3250: 0064–6, 0077–8; HSL/UK/
62/263: 2593, 2596; N.M.R. ST 9307/1–6.
Group (71): The Cliff (Tarrant Rawston). Vestiges
of 'Celtic' fields are traceable on the shoulder of The
Cliff, a steep river cliff rising to 300 ft. above O.D.
and facing N.W. over the Tarrant valley.- Because of
destruction by strip cultivation and later ploughing no
complete field survives.
Air photographs: CPE/UK 1934: 3151–3.
Group (72): South Tarrant Hinton Down—
Race Down—Monkton Down (Tarrant Hinton,
Tarrant Launceston, Tarrant Monkton).- 'Celtic' fields,
now almost entirely destroyed by cultivation, are visible
on air photographs over much of the western parts of
these parishes. They are not demonstrably continuous,
but they may once have been so, and some at least were
almost certainly associated with the extensive settlement
on Blandford Down (Tarrant Launceston (14)).
On S. Tarrant Hinton Down very faint traces of 'Celtic'
fields are visible between 919100 and 929097 on the top of a
ridge and extending down its E. slope into Tarrant Launceston.
They lie close to the settlement on Blandford Down and to the
enclosure in Tarrant Hinton (21), but the remains are so disturbed
that no certain physical connections are determinable.
S.W. of Tarrant Hinton village (around 931106) air photographs reveal traces of 'Celtic' fields on a low spur overlooking
the Tarrant valley.
Remains of 'Celtic' fields, severely damaged by recent development, survive within the area of the military camp on
Race Down, around 918086. Further E. (around 934090) air
photographs indicate the former existence of 'Celtic' fields, now
totally flattened by ploughing, on the N.E. slope of a dry valley.
On Monkton Down (926080), 'Celtic' fields of markedly
rectangular and rather elongated form, measuring up to 60 yds.
across and 120 yds. long, are traceable on the W.-facing slope
of a spur overlooking Pond Bottom. Faint traces are visible
extending E. to the soil-mark enclosures Tarrant Monkton
(20), and S.E. towards Group (70).
Air photographs: CPE/UK 1845: 4064–5; 58/3250: 0077–9,
0112–3; HSL/UK/62/263: 2593; N.M.R. ST 9307/3, 9308/5.
Group (73): Pimperne Down—Hinton Bushes
(Pimperne, Tarrant Gunville, Tarrant Hinton). 'Celtic'
fields are traceable over much of the northern part of
Pimperne parish and in the adjacent parts of Tarrant
Gunville and Tarrant Hinton. They cover some 350
acres and are not continuous, though they may formerly
have been so. Enclosures in the area, some of them
certainly settlements of Iron Age and/or Romano-British date, are likely to be associated with the 'Celtic'
fields. (Map in end pocket.)
'Celtic' fields, now much damaged by ploughing, survive on
Pimperne Down between 275 ft. and 380 ft. above O.D., on
the S.E. slope of a dry valley. They extend in a narrow band
N.E. from the Iron Age enclosure, Pimperne (15), towards
Pimperne Fox Warren (900109). Most of those fields which are
complete vary in area between ¼ and ½ acre.
To the S.W. of the foregoing, faint traces of 'Celtic' fields
are discernible on Camp Down (around 881089), but they have
been much altered by strip ploughing and later cultivation.
To the N., on Gunville Down and extending into Pimperne
(904117), are further remains of 'Celtic' fields, now heavily
ploughed. They are notably rectangular and elongated (though
some internal divisions may have disappeared) with the long
axes aligned N.W.-S.E. Faint traces of fields are also visible
E. of Pimperne Wood.
E. of Pimperne Down, towards and around Ferns Plantation
(910106), 'Celtic' fields now almost totally flattened by cultivation occupy the slopes and summit of a spur. Even the
smallest of these fields is as much as 1 acre in area.
Further E., S. of Hinton Bushes and mainly in Tarrant
Hinton parish, a series of elongated 'Celtic' fields and angular
closes, now largely flattened, occupy the area between enclosure Pimperne (18) and settlement Tarrant Hinton (18), and
almost certainly are associated with them. The boundary dyke
extending S.W. from Tarrant Hinton (18) appears to delimit
these fields on the S. One of the few complete fields measures
130 yds. by 33 yds.
Air photographs: CPE/UK 1845: 4069–71, 6065–70; CPE/UK
1934: 2159–60, 4154–8; CPE/UK 1944: 2136–8, 2318–20,
3320; 58/3250: 0111, 0136–8.
Group (74): Tarrant Hinton Down and Launceston Down (Tarrant Hinton, Tarrant Launceston).
'Celtic' fields formerly covered much of Tarrant Hinton
Down in the N.E. of that parish (946126—950111), but
they have been levelled by ploughing and for the most
part are visible only as faint traces on air photographs.
On the N. they adjoin and appear to be associated with
an Iron Age and Romano-British settlement, Tarrant
Hinton (19). Further S. they surround and incorporate
a small group of round barrows, Tarrant Hinton (46–
50); their relationship to the dyke (952119) extending
W. from Long Crichel is uncertain. At their S. limit
the fields are clearly cut by the W. end of another dyke
Tarrant Launceston (16,a).
Air photographs: CPE/UK 1845: 4149–50; 6059–61;
F22/58/1090: 0094–5; CPE/UK 1934: 2150–1; HSL/UK/62/263:
2589; C.U.A.P. AMO 9, 10.
Remains of 'Celtic' fields on the S. of Hyde Hill Plantation in Tarrant
Launceston form part of a Group which lies mainly in Long Crichel;
they are, therefore, reserved for treatment in Dorset V.
Group (75): Chettle Down and Hookswood
Common (Chettle and Farnham). 'Celtic' fields formerly covered much of the N. and W. of Chettle
parish, between 940152 and 940135, but the majority
have been severely damaged by ploughing and now are
visible only on air photographs. They survive in old
pasture on Chettle Common, around 940145, and may
be traced into Chettle Chase Coppice, but these examples
have been damaged by later digging and are partly
obscured by scrub. They lie on a gentle S. slope and are
defined by lynchets or low spread banks, rarely more
than 2 ft. in height; though ploughed, they are visible
to the E., up to and S. of the settlement in Chettle (14).
Several of the fields or closes near the settlement are
irregular in shape. Traces of fields extend as far E. as
Hookswood Common (945152). Further traces, detached from the main block, occur N. of Farnham near
Half Hide Coppice (955162) and also to the N.W. in
Tarrant Gunville, just S. of enclosure (33).
Air photographs: CPE/UK 2038: 3065–9; C.U.A.P. ANC 4, 6.
Group (76): Stubhampton Down and Manor
Hill (Tarrant Gunville). 'Celtic' fields cover much of
the N.W. part of Tarrant Gunville, but have largely
been flattened by intensive cultivation, especially in
recent years. They lie between 300 ft. and 500 ft. above
O.D. on a series of spurs separated by dry combes at
the head of the Tarrant valley (903143—923157).
In the area of Stubhampton Down 'Celtic' fields covered much
of Earl's Hill and the slopes of the dry combe immediately S. of
it (907142). Few complete fields remain visible, even on air
photographs. On the lower slopes mediaeval and later strip
cultivation had largely removed the 'Celtic' lynchets before the
destructive effects of more recent ploughing.
Immediately to the N.E. 'Celtic' fields covered most of the
spur between Stubhampton Bottom and Ashmore Bottom
(around 914147). Though much damaged by later ploughing, a
number of large fields, up to 150 yds. by 80 yds., appear to have
existed, as well as smaller ones.
To the E., on the spur of Manor Hill, air photographs indicate an irregular pattern of earthworks associated with a track
and probably representing a settlement, all now flattened by
ploughing. The track, which appears to lie among 'Celtic'
fields, runs N. from 92201504 for some 400 yds. to the probable
settlement area around 921154 and then turns sharply E.N.E.;
after 200 yds. it turns N. again and after a further 200 yds. is
lost. The presumed settlement incorporates what appears to be
an almond-shaped enclosure, some 400 ft. by 300 ft. overall.
Air photographs: CPE/UK 1944: 4319–22; CPE/UK 2038:
Group (77): Sutton Hill and Bareden Down
(Fontmell Magna, Iwerne Minster, Sutton Waldron).
'Celtic' fields appear formerly to have covered much of
the summit and the slopes of the Chalk escarpment in
these parishes. Subsequent destruction has left discontinuous areas of remains, all levelled by ploughing
except on the steeper slopes, and visible only on air
Air photographs show 'Celtic' fields at the top of the escarpment in Fontmell Magna, between 877167 and 887165 on
either side of the old Blandford-Shaftesbury road, and on the
spur of Fontmell Down around 878180. A few lynchets survive on the upper slopes of the spur immediately S. of Littlecombe Bottom, with later strip lynchets below them. To the
S. further remains are detectable on Sutton Hill, between
877157 and 892157, and on the steep slope overlooking Coombe
Bottom (876160). In Iwerne Minster remains of 'Celtic' fields
lie around the head of a dry combe on Bareden Down (885153).
Air photographs: CPE/UK 1934: 4324–6; CPE/UK 2038:
Group (78): Ashmore. Evidence of 'Celtic' fields
survives both N. and S. of the village. On the slopes of
the combe head above Boyne Bottom (905185), near
the N. edge of the parish, are at least four large 'Celtic'
fields; they measure up to 130 yds. by 80 yds. and are
defined by low lynchets. Air photographs suggest
traces of other fields, now ploughed flat, adjacent to
and S. of the village at 911172 and 910166. The lyncheted angles of 'Celtic' fields, much reduced by ploughing and in the past mistaken for barrows, survive on the
spur immediately N. of Well Bottom (916167).
Air photographs: CPE/UK 1811: 1078–9; CPE/UK 2038:
4073–7; N.M.R. ST 9018/1–3.
Group (79): Melbury Down and Compton
Down (Compton Abbas and Melbury Abbas). 'Celtic'
fields have survived in the vicinity of Melbury Down,
on the slopes of the deep and narrow dry valley which
extends E. from Melbury Abbas village. Remains seen
on air photographs S. and S.E. of Breeze Hill have now
been levelled by ploughing or are obscured by the
conifers of Gardiner Forest. They cover at least 100
acres on the often steep slopes of the spurs and reentrants forming the N. side of the valley, between
500 ft. and 800 ft. above O.D. (894196—905196).
Other remains occur to the S., around 891189, at a
similar height on the northward-facing side of the valley;
presumably they formerly extended on to the higher,
more level ground to the S. Across the spur to the S.W.,
in Compton Abbas, remains of 'Celtic' fields occur on
the steep slopes around the head of the combe E. of
the village (884187); scarps are seen, but no complete
fields. Ploughing above and below has removed all
indication of their former extent.
Air photographs: CPE/UK 1811: 1079–81, 3078–80.