Other Roman Monuments

OTHER ROMAN MONUMENTS

(Within each parish the monuments are grouped as far as possible in accordance with their geographical setting.)

Affpuddle

Remains of the Roman road from Old Sarum to Dorchester, to be described in Dorset V, may be seen descending to the ford at Ashley Barn.

Arne

(49) Worgret (SY 98 NW). Occupation Debris was found in 1952 and 1955 in a field W. of Cuckoo Pound Lane, on Valley Gravel, towards the crest of the ridge (91258695) and to the S.E. overlooking the flood-plain of the R. Frome (91488669). The pottery (in D.C.M.), ranging from the 1st or 2nd century to the 4th century, seems to be domestic waste. In an army camp here in 1915 a mass of pottery, untraced, was observed lying on 'river gravel' at a depth of 3 ft. and was thought to be derived from Romano-British pottery-kilns. (Dorset Procs. XXXVI (1915), xl-xli; LXXIV (1952), 96; LXXVII (1955), 151.)

The following sites (50–54) occupy sandy Bagshot Beds around Poole Harbour, largely heathland.

(50) Stoborough (SY 98 NW). Occupation Debris, and a claylined vat or basin (Fig.), suggesting industrial activity throughout the Roman period, were found in 1952 on the N. side of Nutcrack Lane (92638638), and subsequently in fields to S. and W.

The 'vat' (in D.C.M.), roughly bell-shaped and about 3 ft. wide internally and 2 ft. deep, has a lining some 4 ins. or 5 ins. thick of puddled chalk faced with fire-hardened clay, a small eccentric sump-hole at bottom and remains of five vertical holes about 4 ins. deep in the flattened rim. The functions of a tongue of chalk about 9 ins. wide projecting obliquely from below the rim on the W. for some 3½ ft., and of two groups of stake-holes flanking the 'vat' in the underlying soil at a depth of 2 ft. 8 ins., are obscure. The 'vat' was constructed in a deposit of ash and burnt earth some 4 ft. to 5 ft. thick containing native Durotrigian and local romanized coarse pottery, much of it oxidized, not necessarily later than the early 2nd century A.D. Similar pottery in the ashy soil inside the 'vat', probably a deliberate filling, may have come from this deposit. (fn. 1) It has been suggested that it was a puddling-hole for preparing potter's clay. Neither kiln nor undoubted wasters have, however, been found.

Arne (50). Stoborough, 'vat'. A topsoil; B grey or brown ashy earth; c grey ash; d red ashy earth; e red or orange ash; f dark earth; g asy sand ;h sandy ash;1 sand;j alluvium.

In the W. half of the field (925863), S. of the lane, ploughing has revealed widespread pottery, some of the late 4th century. Test pits in paddocks further W. adjoining Stoborough village revealed what were said to be ash deposits, with pottery from Iron Age 'B' and 'C' to 4th-century Romano-British (92488633, 92468630, 92498630, 92498629, 92388627, 92508624). Further S. (92658614), beyond a brook, a borehole for clay revealed Romano-British pottery in 1962. (Dorset Procs. LXXIV (1952), 95–6; LXXVI (1954), 81–2.)

(51) Ridge (SY 98 NW). Occupation Debris was found in 1953–4 in test pits in the field E. of Redcliff Farm, at the edge of the flood-plain of the Frome (93558672). Much ashy soil, in places 1 ft. to 2 ft. thick, was found with local coarse pottery dating from at least the 2nd to the 4th century A.D., including complete vessels (some, apparently wasters, in Christchurch Museum), a quern-stone and a coin of Septimius Severus (193–211). Pottery manufacture has been suggested here. Romano-British sherds (93288673) were also observed N.E. of the farm. (Dorset Procs. LXXVI (1954), 81; LXXXI (1959), 123; Swanage Times, 4 Nov., 30 Dec. 1953.)

(52) Arne Heath (SY 98 NE). Occupation Debris from saltboiling and possibly also from pottery manufacture has been ploughed up in heathland afforestation N. of Bank Gate Cottages, a few feet above sea level some 300 yds. from the present shore of the Wareham Channel of Poole Harbour (95748723).

The remains comprise several concentrations of coarse-ware sherds (in D.C.M.), almost all oxidized and brittle, with much 'briquetage' associated, as elsewhere in Poole Harbour and the Isle of Purbeck, with crystallization of salt from brine (see Introduction, p. 526). The pottery ranges from the 1st or 2nd century to the 3rd or 4th century A.D.; the 'briquetage' includes fragments of large thin-walled containers, perhaps semicircular troughs, with knife-cut rims or edges, and perhaps one piece of a hand-moulded support. Shapeless fragments of fired clay, some with vitrified surface, may belong to associated structures. (Dorset Procs. LXXXIV (1962), 140.)

(53) Big Wood (SY 98 NE). Occupation Debris comes from rabbit-burrows in a shallow re-entrant below the 50 ft. contour W. of Big Wood(97608843). The pottery(in D.C.M.), of local Romano-British coarse ware not necessarily later than the 2nd century, is oxidized and said to be associated with red ashy soil. (Dorset Procs. LXXVI (1954), 81.)

(54) Shipstal (SY 98 NE). Occupation Debris connected with the salt industry occurs in a strip at least 170 yds. long and 25 yds. wide at the edge of the scarp bordering the mud-flats of Poole Harbour S. of Shipstal Point (central point, 98198809).

Material excavated by H. P. Smith is untraced, but the few subsequent finds of pottery, mostly oxidized, suggest activity in the 1st or 2nd century, while small fragments of'briquetage' are consistent with material from similar local saltings. Smith records, however, the characteristic hand-moulded clay supports for the brine containers, misinterpreting them as 'kiln rests', and also part of a floor of fired clay which he assigned to a pottery-kiln, but which was more probably a Salter's hearth as at Warsash (Hampshire Field Club, Procs. XIII, pt. 1 (1935), 105–9) or Ingoldmells, Lincolnshire (Ant. J. XII (1932), 239–253). Objects found recently are in D.C.M. (Dorset Procs. LVI (1934), 16–17; LXXIV (1952), 96; LXXXIV (1962), 142; J.R.S. XXIV (1934), 216; H.P. Smith, History of Poole, I (1948), 88.)

Bere Regis

The parish contains considerable remains of the Roman road from Old Sarum to Dorchester, to be dnote idinDorset V. Scattered sherds of Romano-British pottery, some of the 4th century, have been found near Bere Down Buildings (84019723), and a coin of Domitian (81–96) and one of Septimius Severus (193–211) behind the N. rampart of Woodbury hill-fort (856649).

The following sites are on Upper Chalk downland, capped locally by Plateau Gravel.

(120) Bagwood (SY 89NW, NE). Floors, Wells, Pits and Occupation Debris have been found about 250 ft. above sea level on Bere Down, in Bagwood Close and Coppice and for some 300 yds. W. of the latter, astride the Roman road between Badbury Rings (Vindocladia?) and Dorchester.

A well, supposed to lie some 40 yds. N. of the Roman road (85149714), was excavated in 1860 by William Shipp for the Dorset antiquary, Charles Warne, to a depth of 60 ft. to 70 ft. without reaching bottom. It was about 8 ft. in diameter and cut in chalk with the upper 10 ft. to 12 ft. revetted with regularly laid blocks of chalk and 'green sandstone' about 1 ft. square. The ashy filling contained much coarse pottery, some samian, nails and blocks of Kimmeridge shale. (fn. 2)

The topsoil of the close W. of the coppice, particularly S. of the line of the Roman road, (fn. 3) is noticeably dark, with Romano-British pottery, some of the 4th century; N. of the coppice there is debris from an ancient flint industry, but pottery is scarce, and signs of 'Celtic' fields peripheral to Ancient Field Groups (31) and (32) suggest that this area may have been cultivated during the life of the settlement.

In the W. part of the close, on the gentle northerly slope N. of the Roman road, excavations by Mr. G. Toms in 1962–6 revealed traces of an occupation floor (85109712) with clay roof-tiles, mortared flint and chalk, sandstone, daub, fragments of Kimmeridge shale including parts of incised tablets or panels, worked bones used for weaving, and much iron slag and clinker. This occupation, apparently of the mid or later 2nd century, overlay an earlier 2nd-century pit. Isolated postholes, two gullies and several pits were found, one of them, refilled in the late 1st century, possibly for road-metal. A chalk-cut well, 3½ ft. in diameter and at least 70 ft. deep, contained six coins, the latest of Tetricus l (270–3), large quantities of pottery and daub with plank and wattle impressions, an extensive series of animal remains including bird, rodent and fish bones, and many objects including a Purbeck stone mortar, quern-stones, an ox-goad and a bill-hook.

Excavations at the S. side of the Roman road in the N.W. corner of the coppice (85169712) disclosed a gravelled area of the late 3rd or 4th century overlying natural gravel, covered with a layer of cobbles in which were a few isolated post-holes; above this a 2 ft. deposit of burnt occupation debris, including daub and building rubble without roof-tiles, was dated to the 4th century by pottery and a small coin series ending with an issue of 341–6. W. of this point surface finds of the earlier 2nd century to the mid 4th century suggested occupation for at least 300 yds. along the Roman road, and a trial excavation at 84909700 revealed much 2nd and 3rd-century pottery S. of the road.

Warne's identification of the site with the Ibernio of the Ravenna Cosmography is without foundation (cf. Archaeologia XCIII (1949), 35), nor do any known features connect the site with an official posting-station (mutatio). (Archaeologia XXXIX (1863), 85–92, reprinted in Warne's Ancient Dorset (1872), 201–8; Hutchins l, 147–8; Dorset Procs. LXXXIV (1962), 103–6, 115; LXXXV (1963), 99–100; LXXXVI (1964), 110–2; LXXXVII (1965), 98–9; LXXXVIII (1966), 116–7.)

(121) Muddox Barrow Coppice (SY 89 NE). Occupation Debris comes from Bere Down N.E. of the coppice some 250 ft. above sea level (85389688). It comprises a few Durotrigian or Romano-British sherds, found in a dark layer about 2 ft. below surface in laying water-pipes in the arable field some 400 yds. S.E. of the remains at Bagwood. (Dorset Procs. LXXIII (1951), 102–3.)

Bincombe

Remains of the Roman road from Dorchester to Radipole can be seen crossing Ridgeway Hill (see Roman Roads, p. 528). Samian ware and Romano-British coarse pottery have been found in Chalbury hill-fort at 69548379 and 69448396 (Ant. J. XXIII (1943), 103, 108). Scattered sherds, some of the 3rd or 4th century, occur on Bincombe Hill (688847) among 'Celtic' fields (see Ancient Field Group (8)).

(77) West Hill (SY 68 SE). An Inhumation Burial was found in 1943 during ploughing below the crest of the Chalk ridge (69968494). The adult skeleton, extended E.-W. with head to W., lay in a cist lined at the sides and covered with flat lime-stone slabs, the top about 1 ft. below surface. There were no associated objects, but analogous burials in Dorset, where datable, are of the Roman period. (Unpublished note and photographs by V. F. M. Oliver, in D.C.M.)

Broadmayne

A coin of Constantine I (not Constantius II as stated) of 324–30 comes from Broadmayne village (730865; Dorset Procs. LXXVII (1955), 152). Pits and debris indicating settlement into the 4th century have since been found in the village (726866; Dorset Procs.LXXXVIII (1966), 103).

Chaldon Herring

A few Romano-British shale objects and sherds, one of New Forest ware, were found in 1958 E. of Holworth Farm (771832; Dorset Procs. LXXXI (1959), 132). Romano-British sherds also come from among 'Celtic' fields in the Warren (790805; see Ancient Field Group (15)).

Chickerell

(34) Putton Brickyards (SY 68 SW). Occupation Debris and Inhumation Burials have been found in the brickfields on Oxford Clay (649800). Of two burials found in 1924, presumably extended, one had a coarse ware bowl (D.C.M. 1907.3.9) perhaps of the 3rd century; (fn. 4) 20 other skeletons were said to have been found in ploughing previously. Occupation debris was 1st and 2nd-century samian ware, coarse pottery including a 3rd-century sherd, and a small grotesque human face in Kimmeridge shale (1933.18.1). A burial (64838000) in a stone cist aligned N.-S. was destroyed in 1963, and three more in 1965, apparently in plain graves similarly orientated, with heads to N., one of them at 64828001. (Dorset Procs. XLVI (1925), xlvi, lxix; LVI (1934), 15; LXXXV (1963), 100–1; LXXXVII (1965), 115–8; J.R.S. XIV (1924), 235; XXI (1931), 241; Bath Herald, 13 Sept. 1924.)

(35) Buckland Ripers (SY 68 SW). Inhumation Burials have been found on the crest of the Langton ridge (Cornbrash) in the S.W. corner of the former parish of Buckland Ripers (631819).

According to Ellis several unaccompanied adult inhumations in covered slab-lined cists, with some charcoal and traces of burning on the stones, were exposed in quarrying 'Tatton Hill' in 1826 and subsequently reburied together. They were presumably extended. Hutchins adds that eight or ten were in cists, about 1 ft. below surface. In 1928 an unaccompanied adult skeleton extended E.-W. with head to W. was found at the same depth nearby (63178191) in a covered cist without end slabs, 6½ ft. long, 20 ins. wide at shoulders and 15 ins. at knees. The posture and type of grave are characteristic of the Roman period in the region. The cemetery apparently extends into Portesham parish (Portesham (63)). (G. A. Ellis, History and Antiquities of Weymouth and Melcombe Regis (1829), 254–6; Hutchins II, 492; Dorset Procs. L (1928), 122–4.)

Church Knowle

Romano-British sherds and a piece of shale armlet were found in the ditch of a Bronze Age barrow (see Round Barrows, Church Knowle (41)) on the E. part of Knowle Hill (Dorset Procs. LXXVI (1954), 53, 55).

Sites (57–58) lie on the heathland of the Bagshot Beds.

(57) East Creech (SY 98 SW). Villa; a tessellated pavement and other remains of buildings about 200 ft. above sea level were partly explored in the 19th century some 300 yds. N. of the Chalk downs.

The site, long under plough, is marked on O.S. by an antiquity symbol at 93538275. (fn. 5) A slight ridge, from which the ground falls away in all directions save W., runs E. on the long axis of the field to some 15 yds. S. of the map symbol, and for the last 60 yds. The crest is slightly flattened, suggesting an artificial platform some 7 yds. wide, with some loose heathstone rubble in the E. end. Around the point marked by the symbol and to the N., much Romano-British material occurs over an area some 50 or 60 yds. square, especially about 93518279, consisting of heathstone, Purbeck limestone, clay roof and flue-tile fragments, wall-plaster, a few large red tile and limestone tesserae, pottery, and Kimmeridge shale waste. The adjoining marshy field to S. has banks and probably drainage ditches of doubtful age, but recent test-borings for clay are not thought to have revealed occupation.

After deep ploughing in 1869, internally-plastered walls were traced by the tenant, Mr. Pike, who found a Tuscan column (see Fig.) about 4 ft. high, apparently of Purbeck stone. (fn. 6) A pavement with simple lattice design executed in coarse tesserae of red tile on a 'white' ground, and measuring about 16 ft. by at least 11 ft., was found in 1888, while Hutchins records 'many small thin triangular or kite-shaped stones of different dimensions having fine points and thin edges', suggesting opus sectile. A note in D.C.M. states that about 100 or 150 yds. N. of the pavement 'was found a basement, of stone, containing the lower courses of (apparently) a stoke hole, from whence issues a side channel, having a side branch, and showing evident marks of fire. This may be the basement of a hypocaust.' This feature, apparently 12 ft. square according to The Builder, may alternatively have been a corn-drying oven; its position is doubtful since a disused clay-pit already existed in the area specified.

Some tesserae are in Poole Museum. Recent surface finds (D.C.M. 1936.18; 1963.I) show occupation from at least 2nd to 4th century. Turned Kimmeridge shale armlet cores occur, and Hutchins records a vessel full of them from a neighbouring field. A shale industry is possible, and although some arable was probably available on the mixed downwashed soils at the foot of the Chalk ridge, the latter was never here occupied by 'Celtic' fields. (Hutchins I, Addenda (1874), 716; Warne, Ancient Dorset (1872), Appendix, 282; D.C.C., 13 Dec. 1888; The Builder LV (1888), 478; Anon. (Mrs. J. E. Panton), Fresh Leaves and Green Pastures (1909), 234; photographs of pavement, R.C.H.M. records.)

(58) Greenspecks Clay-pit (SY 98 SW). A Working-floor and Occupation Debris found c. 1912 in clay-workings (92658270) some 350 yds. W. of East Creech Farm indicate existence of a workshop of the Kimmeridge shale industry, now perhaps wholly destroyed, which may have originated in Iron Age 'C'.

The material (in D.C.M. and the Furzebrook office of Pike Bros., Fayle and Co.) consists of shale waste, mainly trimmed discs and broken or unfinished armlets, some probably cut by iron tools, almost entirely of the hand-cutting industry elsewhere being replaced by the lathe in Iron Age 'C' and Roman times. An unusual conical object (D.C.M. 1912.37.2) and a circular plaque (1919.2.11) found with 'Roman pottery' are lathe products. Some indurated pieces of the working-floor, containing shale chips, flint flakes and sherds, are preserved at Furzebrook, with 1st-century native ware and sherds probably of the 3rd century presumably from this site. (Dorset Procs. LXX (1948), 43–4; LXXIII (1951), 89–91; LXXV (1953), 65, 69; D.C.M. Accessions Book, 17 Dec. 1913.)

(59) Bare Cross (SY 98 SW). Inhumation Burials were found in 1859 on the crest of a saddle in the Purbeck Hills (Upper Chalk), 375 ft. above sea level. They were discovered in the road cutting and an adjoining chalk-pit, probably that at 93088210 behind a W.-facing cross-ridge dyke (see Dykes, Church Knowle (56a), Fig. opp. p. 509). Parallel chalk-cut graves 1 ft. or so apart, some 2 ft. deep and 20 ins. wide, lay E.-W. with head to W. Of six examined all were female or young people according to the anthropologist J. Thurnam; one, a woman of at least 40, lay E. of three individuals of about 12, 15, and 18 years. The regular disposition and presumably extended posture argue Roman date. (Purbeck Papers I (1859–60), 232–3.)

Sites (60–62) lie on the limestone uplands of the Purbeck and Portland Beds.

(60) Bradle Farm (SY 98 SW). Inhumation Burials were found ¼ mile S. of the farmhouse on the N. slope some 300 ft. above sea level (c. 92958020). A covered cist of stone slabs, 11 ft. long and 1½ ft. wide, found in 1888 1 ft. below surface, contained a central adult skeleton with head to E. and hobnails (in D.C.M.) at feet, with other bones beyond the head. A similar, smaller coffin-shaped cist was found c. 1876 about 100 yds. away in the same field. (D.C.C., 17 May 1888; Notes and Queries for Somerset and Dorset I (1890), 47.)

(61) Smedmore Hill (SY 97 NW). Occupation Debris and waste from manufacture of shale armlets were found after pipelaying in 1956 above the 500 ft. contour near earthworks of a settlement (Church Knowle (55)). The material, mainly from 93167955 about 70 yds. S.E. of the earthworks, consists of Iron Age 'A' ware, samian, 3rd or 4th-century coarse ware, several shale armlet cores of Calkin's classes A, C and D, and a flint lathe-tool. (Dorset Procs. LXXVIII (1956), 76); LXXXV (1963), 102–3.)

(62) West Orchard Farm (SY 97 NW). Occupation Debris comes from an undefined area (93657926) on both sides of the field wall some 200 yds. S.E. of Newfoundland copse, about 550 ft. above sea level. A trial excavation in 1956, N. of the wall, partly exposed a rough floor of limestone blocks of unknown date overlying an Iron Age 'A' deposit (93687925). Excavated material and surface finds include waste from manufacture of hand-cut shale armlets, much Iron Age 'A' pottery (some haematite–coated), sherds of Iron Age 'C' or early Roman date, and a 2nd–century enamelled brooch, but no demonstrably later material. Flint flakes and scrapers occur in the field to W. (Dorset Procs. LXXV (1953), 69; LXXVII (1955), 126–7; LXXXV (1963), 102.) Excavation S. of the field wall in 1967 revealed pits and further traces of flooring, and suggested two phases of Iron Age 'A' occupation associated with a hand-cut shale armlet industry. The later occupation, marked by Durotrigian pottery and other finds, some certainly of post-conquest date, was associated with waste from a similar shale industry perhaps using metal tools, and a few shale lathe–cores. There was evidence for a flint industry of Mesolithic character in the vicinity. (Dorset Procs., LXXXIX (1967), 141–3.)

Coombe Keynes

(21) Manor Farm (SY 88 NW). Occupation Debris comes from N.E. of Coombe Wood about 175 ft. above sea level on Upper Chalk near its junction with Reading Beds of the Frome valley (83688552). A silage pit in 1950 showed dark patches 3 ft. or 4 ft. below surface containing sherds of the 1st or 2nd to the 4th century. There are other remains nearby (see Wool (50)). (Dorset Procs. LXXII (1950), 82; LXXIII (1951), 85.)

Corfe Castle

Roman pottery and in some cases shale objects have come from the following sites not listed as monuments: (i) clay-pits S. of Norden Farm (Dorset Procs. LXXIII (1951), 88–9; Ant. J. XXXI (1951), 193–4); (ii) Challow Hill (Dorset Procs.LXXXVII (1965), III; (iii) Town's End, SY 96308117 (Dorset Procs. LXXXVI (1964), 117); (iv) Lynch (Purbeck Papers III (1859–60), 220; Dorset Procs. LXXVII (1955), 150); (v) Chapman's Pool (Dorset Procs. XIII (1892), 183). For the Roman objects and intrusive inhumations probably of that age, in two barrows at Afflington, see Round Barrows, Corfe Castle (183–4); for two settlements, both probably of Roman date, associated with 'Celtic' fields, see Settlements, Corfe Castle (221–2).

Sites (224–230) occupy sandy Bagshot Beds, largely heathland, in and around Poole Harbour.

(224) Green Island (SZ 08 NW). Occupation Debris, perhaps Iron Age 'A' to late Roman, with waste from Kimmeridge shale armlet manufacture, comes mainly from the N. half of the island below the 50 ft. contour.

Pottery including New Forest ware, and shale, were found before 1925 in the N.W. cliff scarp at 00588674, and above the W. cliff at 00458656. Test pits by H. P. Smith in 1951 near The Hermitage (00608664) and in five places N. and S.W. of the lagoon and two or three W. of the N. landing-stage, nearly all showed occupation remains, including Iron Age 'C' pottery, samian, and shale from both hand-cut and latheturned armlet industries; at The Hermitage, hand-cut waste showing use of iron tools occurred with 'C' sherds; iron slag was found on the E. shore N. of the lagoon and in one of the test pits S. of it. Subsequent excavation by Mr. J. B. Calkin near the N. shore (00608673) revealed an occupation layer with some Iron Age 'A' and 'B' pottery but mainly 'C'; a Hengistbury class H sherd (fn. 7) and fragments of amphorae suggesting a regional pre-Caesarian trade with the Continent; also shale workers' waste, mostly class A cores from turned armlets with a few hand-cut roughs, and worked flints, some approaching the standardized Romano-British type of lathetool. Oxidized sherds (in D.C.M.), some of the 3rd or 4th century, come from near the old landing-stage (00708679). Some of Smith's finds are in Poole Museum. (Dorset Procs. LXXV (1953), 53–4, 60, 64–6, 69; LXXXV (1963), 104–5; LXXXVIII (1966), 121; H. P. Smith, History of Poole I (1948), 89; O.S. Archaeology Division records.)

Buried remains, mostly below low water mark, of a stoneramped causeway connecting the W. shore with the mainland some 200 yds. W. of Cleavel Point, save for a gap in the centre of the channel, were confirmed by excavation and diving in 1959. The causeway, undated, was probably surfaced with stone flags on a log corduroy laid axially (Dorset Procs. LXXXVIII (1966), 158–60.)

(225) Furzey Island (SZ 08 NW). Occupation Debris comes from two places in the low cliff face on the W. shore. A shallow earth-filled depression, unexcavated, some 3 ft. below surface and about 12 ft. wide in exposed section (00918694), yielded a small amount of native coarse pottery, not closely datable. Unworked shale fragments are said to occur. Some 200 yds. N.W., a narrow feature, perhaps a gully (00838710), had a similar sherd and a fragment possibly of Romano-British salting 'briquetage'. (Dorset Procs. LXXXV (1963), 104.)

(226) Fitzworth (SY 98NE). Occupation Debris, Iron Age and Roman, with remains of a hut and salters' waste, comes from a wide area in the N. of the low-lying Fitzworth peninsula.

Two oval areas of Romano-British domestic occupation with much shell refuse, each with adjacent working area, were examined in 1947 some 400 yds. N.E. of the farmhouse. The larger domestic area of some 50 yds. by 25 yds. (99228676) had pottery of the 2nd to 4th centuries A.D. and, on the W., remains perhaps of a round hut (99208676) about 20 ft. wide, with sandstone footings and central hearth; the smaller (99388670) had similar pottery and two coins of the later 3rd and early 4th centuries. The working areas (99268674, 99338668) showed no structural remains but much oxidized coarse pottery of the 1st century and later, unfired clay, and 'briquetage', mostly broken brine-containers and props, but including fire-bars possibly from pottery-kilns. Subsequent ploughing revealed similar material in the same field: burnt clay (99318677), sherds and 'briquetage' (99208682, 99178672, and especially 99298667). Romano-British sherds also come from the field to N.W. (99148687), near Iron Age 'A', 'B' and 'C' sherds exposed in a bomb-crater further W. (99038687). (Dorset Procs. LXX (1948), 42–3, 51, 57–8; LXXI (1949), 62–3; LXXXIV (1962), 137, 142.)

(227) Ower (sy 98 NE, SZ 08 NW); A Pottery-kiln, and Occupation Debris including salters' waste, have been found in low-lying arable fields in the N.E. part of the Ower peninsula where the South Deep approaches Cleavel Point.

The pottery-kiln of 'horizontal-draught' type, excavated in 1951 underneath the low bank bordering the mud-flats of Poole Harbour (00048618), consisted of a 10 in. thick dome of fired clay, some 1½ ft. high and 4 ft. in diameter internally, on a level clay floor 2 ins. thick, with two flues sub-rectangular in cross-section, about 1 ft. wide and 6 ins. high, set against the S. and N.W. sides. The dome was broken into on W., probably to extract the load after firing, but does not seem to have possessed vents unless these had been plugged with clay.

Oxidized coarse ware sherds found close to the kiln include a type (cf. Arch. Ael. xxxv (1957), no. 314) hardly datable before c. 200 as well as others of Durotrigian origin. Excavation about 30 yds. S.W. (00018616) showed similar oxidized ware including a 3rd or 4th–century type (cf. ibid., no. 228).

Salters' waste, consisting mainly of fragments of large thin-walled containers and hand-moulded props, abounds at the kiln site, and is associated with oxidized sherds in a thin layer below topsoil, visible in the scarp bordering the mudflats for some 230 yds. (00198609 to 00008619) and reappearing further W. for about 50 yds. (99928618 to 99878618). Excavation showed 1st and 2nd-century occupation debris, including a shell midden, at two places, 116 yds. S. (00038606) and 250 yds. S.W. of the kiln (99898599), the latter with oxidized sherds, perhaps wasters. Romano-British sherds also come from ploughing near Newton Bay (99998595) and, with 'briquetage', from the field W. of the kiln (99888603). (Dorset Procs. LXXIII (1951), 91–2; LXXXIV (1962), 115–6, 141.)

(228) Rempstone (SY 98 SE). Occupation Debris comes from Rempstone Hall and neighbourhood.

The lower half of a rotary quern and part probably of an upper stone, both of heathstone, were found after ploughing in the field W. of the Hall, the former at about 99068249. A few Romano-British sherds have been found here; a waste core from a turned shale armlet comes from the stream in the grounds of the Hall (99168246). (Dorset Procs. LXXXV (1963), I03–4.)

A deposit of similar cores with a shale object 'like the bowl of a large glass or rummer...' was found in 1845 in draining a withy-bed, presumably that adjoining the main road at 995820, in which lies the Rempstone stone circle. This is probably to be identified with the conical object (D.C.M. 1908.28.1), apparently a waste product of the lathe, formerly said to come from Norden, Corfe Castle. (Devonshire Assoc. Trans. II (1867–8), 630; W.A.M . XLIV (1928), 113; Purbeck Papers I (1856–7), 86; Hutchins I, 563; Warne, Ancient Dorset (1872), Appendix, 229–30.)

(229) Brenscombe Farm (SY 98 SE). Villa; remains of two mosaic pavements were found in 1961 and 1967 about 500 yds. N. of the foot of the Chalk downs some 150 ft. above sea level (97898272). The site is scheduled as an ancient monument.

The first fragment, 6 ins. below surface and orientated approximately N.E. to S.W., measures some 7 ft. by 3 ft. exclusive of remains of mortar bedding, and is of approximately ½ in. tesserae of chalk and brown ferruginous sandstone. The pattern, much disturbed, shows parallel brown and white bands of varying width with traces of a rectangular panel to S.W. within a chevron border. Excavation revealed spreads of stone debris, or rough cobbling, on all sides. A superior mosaic with a foliate scroll was partly exposed near by in 1967.

Surface debris includes pieces of Roman brick, roof and flue-tile, small red brick tesserae, a few sherds of samian ware and some coarse ware of the 3rd or 4th century. (Dorset Procs. LXXXIV (1962), 113–4; LXXXV (1963), 103; LXXXVIII (1966), 120; LXXXIX (1967), I44.)

(230) Norden (SY 98 SE). Inhumation Burials, Floors and Occupation Debris, including objects from a Kimmeridge shale industry, have been found above the 50 ft. contour in the Norden Clay Works, some 300 yds. N. of the Corfe gap in the Chalk range.

In and perhaps before 1882 several burials in 'stone coffins', probably slab cists, were found in the Matcham pits (957826), with remains thought to belong to a road at a depth of 18 ins., and, nearby, a 'fine collection of Roman pottery', mostly 'urn-shaped vessels, decorated with a lozenge pattern around the neck'. These vessels, preserved by Lord Eldon but not traced, were thought to indicate the proximity of a kiln.

Roman remains have been found recently between the clay-pits beside the Wareham road. N.E. of the road were surface finds (956827), remains of a stone-paved floor with pottery, some of the 4th century (95788261), a small cist containing Romano-British sherds (95648265), and, beyond the railway, a chalk and limestone floor at a depth of 2½ ft., associated with a coin and sherds of the 4th century (95948264). Finds S.W. of the road occurred between the mineral railway (95568266) and 'North Castle' (95788250). A deposit 1 ft. thick of chalk and limestone capped with small stone blocks, 1½ ft. below ground surface but rising N.W. in conformity with it, was exposed in 1964 in an excavation 3 ft. square (95558266). It was observed in a pipe-line trench for 6 yds. to 10 yds. S.E. until destroyed by a clay-pit, and is possibly a road, perhaps that claimed to have been found c. 1882. It overlay Roman material including a coarse ware sherd of the early 2nd century and part of an oblong shale plaque with incised geometric decoration (in Christchurch Museum). A piece of a carved shale table leg (in D.C.M.) and some 20 small siltstone cubes, perhaps unused tesserae, were unstratified except for a single cube below the deposit.

Cores from turned shale armlets are said to have been common at Norden; a few are in D.C.M. Also in D.C.M. are two more shale plaques, one circular, found before 1859 with a pair of bronze dividers in lowering the road N.W. of St. Edward's Bridge, and part of a carved table leg from a clay-pit near Hill Coppices (95068257; Ant. J. XXXI (1951), 193–4). These objects suggest a manufactory of furniture and other objects of shale besides armlets. (Dorset Procs. VIII (1887), XXXIX–XL;XIII (1892), 188; LXXIII (1951), 86–9; LXXV (1953), 69; LXXVII (1955), 126, 150; LXXVIII (1956), 91; LXXXVI (1964), 116–7; LXXXVII (1965), 111–3; LXXXVIII (1966), 119; LXXXIX (1967), 145; Purbeck Papers I (1859–60), 225; II (1867), 53–4; Hutchins I (1861), 471; H. J. Moule, Dorchester Antiquities (2nd edn. 1906), 44.)

Sites (231) and (232) lie on Upper Chalk.

(231) Castle Hill (sy 98 se). Occupation Debris has been found on the N. and W. parts of the hill occupied by the castle. Sherds of 2nd-century samian and coarse ware, and small fragments seemingly of Roman flue-tile, come from excavations in the West Bailey (95838228); a few sherds of coarse ware are recorded from the surface of the scarp S. of the latter and at the outer base of the N.E. wall of the Inner Ward (95948233). (Med. Arch. IV (1960), 38; Dorset Procs. LXXXV (1963), 104.)

(232) West Hill (SY 98 SE). Occupation Debris comes from ploughing above the 300 ft. contour on the crest of the narrow ridge overlooking the Corfe gap (954823; Fig. p. 97). Romano-British sherds, some of the 3rd or 4th century, flint flakes, part of an unfinished hand-cut shale armlet, and 26 bronze coins (all of the 4th century down to Gratian (367–83) save two of the 2nd and 3rd centuries) have been found since 1952 on the crest of the hill. (Dorset Procs. LXXIV (1952), 93; LXXXI (1959), 108; LXXXVI (1964), 117; LXXXVIII (1966), 120.)

Sites (233) and (234) occupy loamy Wealden Beds.

(233) Sandyhills Copse (SY 98 SE). Occupation Debris of Iron Age 'C' or early Roman date comes from ploughing W. of Sandyhills Copse and the adjacent part of Corfe Common, on and near the crest of a ridge some 150 ft. above sea level. A concentration of coarse ware sherds and some burnt and unburnt clay occurs close to the copse (96958157); a scatter of similar sherds, flint flakes and shale fragments has been found on the common about 100 yds. S.W. (96888149). A waste core from a hand-cut shale armlet comes from E. of the copse. The objects are in D.C.M. (Dorset Procs. LXXXIV (1962), 114–5.)

(234) Blashenwell (SY 98 SE). Inhumation Burials, Floors and Occupation Debris including waste from both hand-cut and lathe-turned shale armlet industries and ranging from Iron Age 'A' to Roman times, have been found in and near the old marl-pit (951805) in the calcareous tufa deposit N. of Blashenwell Farm. The site is about 150 ft. above sea level, close to the Purbeck marble outcrop.

The geology and ecology of the tufa, apparently deposited by the neighbouring stream draining from a synclinal trough in the Upper Purbeck Beds and containing a Mesolithic industry (see p. 511) have been much studied. Deposition of tufa ceased before the Roman period and it was overlaid by 1 ft. or more of topsoil containing Roman debris and a mixed Roman and post-Roman fauna indicating locally open conditions.

A contracted skeleton of a youth in a Purbeck stone cist was found in 1895 some 4 ft. deep in the tufa at the S. end of the pit (95188047). In 1908 another cist, with floor and cover slabs, containing an adult skeleton probably extended with head to N., was exposed but not cleared nearby in the S. face above the tufa a few feet from the road, while indications of at least two more graves were seen in the side of the road cutting. Some 6 ft. to 8 ft. W. of the cist a compact mass of unmortared herringbone masonry, some 2 ft. high and 4 ft. long or wide, appeared in the face, based in the tufa. In 1965 a tightly contracted inhumation facing E. with head N. was found between two end slabs, resting on undisturbed tufa at a depth of 2½ ft. A later pit, cutting through the grave but avoiding the burial, which was perhaps of the Early Bronze Age, contained a Romano-British sherd.

In 1936 Dr. W. H. C. Frend recorded 'floors of laid paving-stone, traces of hearths and a large amount of coarse pottery and "coal money" (fn. 8) . . . brought to light through the construction of a farm road to the farmhouse'. In 1938 two vessels found near human vertebrae in topsoil at the roadside between the pit and the farmhouse were exhibited at the Institute of Archaeology, London; (fn. 9) one of these, in the Institute's collection, is of Iron Age 'C' or early Roman date. The site, now under plough, yields Iron Age 'A', 'C', and Romano-British pottery, flint and shale waste from a hand-cut armlet industry, and equivalent waste from manufacture of turned armlets, mainly in and near the S. end of the pit. Stone slabs may indicate a floor some 20 yds. S.W. (95168045). Photographs of the 1908 discoveries are in D.C.M. and R.C.H.M. records. Some remains were noted in 1965 a quarter of a mile N.N.W. (Dorset Procs. XVII (1896), 67–75; XXIX (1908), xl; LXX (1948), 43; LXXV (1953), 52, 69; LXXXVII (1965), 97–8; LXXXVIII (1966), 120; LXXXIX (1967), 145; W.H.C. Frend, 'Iron Age and Roman sites in Purbeck' (MS. in D.C.M., 1936), 13; Procs. Geologists' Association LXVI (1955), 87–8.)

Sites (235–239) are on the limestone plateau of the Purbeck and Portland Beds. For position of (237–9) see Fig. p. 631.

(235) Encombe Obelisk (SY 97 NW). A Building and Occupation Debris including waste from the manufacture of turned shale armlets, were excavated in 1954 about 550 yds. W. of the obelisk, on the upper N. slope of a limestone ridge above the 500 ft. contour (94277898).

The building, aligned approximately N.-S. towards the S.W. corner of a 'Celtic' field (Ancient Field Group (21)), was rectangular, 51 ft. by some 20 ft. to 25 ft., and probably thatched; its E. side was indicated approximately by surface debris and no internal partitions were observed. Surface rubble suggested that the building may have stood in a compound. The remaining walls, some 6 ins. high at most, below 6 ins. of topsoil, were 2 ft. wide, of rubble faced with selected or roughly dressed limestone blocks, unmortared and resting on weathered Portland Stone brash. Some flagstones remained in the N.W. corner. A rectangular cist, about 4 ft. by 2 ft. and 1½ ft. deep, sunk from floor level beside the N. wall-footings, and floored and lined on the three outer sides with flagstones including reused roof-tiles, may have been an addition. It contained five prepared shale discs and 27 lathe-cores of Calkin's class C. Surface finds included similar cores and examples of classes A and B, flint lathe-tools and pottery of the 1st or 2nd century to the 4th century, as well as Iron Age 'A' sherds and a few rejects from a hand-cut armlet industry. (Dorset Procs. LXXV (1953), 54, 69; LXXVI (1954), 80–1; LXXXVIII (1966), 114–5, 120.)

(236) Quarry Wood, Encombe (SY 97 NW). An Inhumation Burial was found c. 1788 in quarrying near the former Encombe Lodge. The skeleton, under 'two very large flat stones set up edgeways', was accompanied by an earthen vessel of about one gallon capacity filled to the brim with 'coal-money'; the quarry is evidently that at 949791 above the 400 ft. contour on the N.E. side of the North Gwyle. (Purbeck Papers I (1859–60), 226.)

(237) Kingston Plantation (SY 97 NE). Occupation Debris including Kimmeridge shale waste comes from ploughing on the plateau S. of the plantation, about 475 ft. above sea level (953788). The material includes samian ware, coarse ware of Iron Age 'C' or early Roman date and some of the 3rd or 4th century, shale armlet lathe-cores of Calkin's class C, and a few Iron Age 'A' sherds and rejects from a hand-cut shale armlet industry. (Dorset Procs. LXXV (1953), 69; LXXVI (1954), 80.)

(238) Kingston Barn (SY 97 NE). Occupation Debris, including an altar, has been found in arable about 450 ft. above sea level S. of Kingston Barn (963788). Samian sherds, coarse ware of the 1st or 2nd to the 4th century, fragments of shale, and a few large stones possibly from structures, come from W. of the footpath, with, at 96267894, a roughly made Roman altar of Purbeck stone some 14 ins. high (now in D.C.M.; Plate 228). Sherds of Iron Age 'A', and shale waste, occur in the field (962788) to W. (Dorset Procs. LXXIV (1952), 93; LXXVI (1954), 79; LXXXIII (1961), 85–6.)

(239) Westhill Wood, Encombe (SY 97 NE). Occupation Debris comes from ploughing above Westhill Wood near Hounstout Cliff, about 475 ft. above sea level (950778). The debris consists of samian ware, coarse pottery of the 1st or 2nd century to the 4th century, a coin of Carausius of c. 293, some clay roof-tile fragments, chert and flint flakes, and shale waste including a few cores from turned armlets. A mound (95067783), some 75 ft. in diameter and 2 ft. high, shows much limestone rubble. (Dorset Procs. LXXXI (1959), 108.)

(240) Eldon Seat, Encombe (SY 97 NW). Huts of Iron Age 'A', and Occupation Debris from Iron Age 'A' to the Roman period including waste from shale armlet industries, have been excavated near the Kimmeridge Clay cliffs W. of Freshwater Steps, about 250 ft. above sea level (939776).

Excavations by Prof. B. Cunliffe, in 1963–4, where an area with Iron Age and Roman surface debris is crossed by an E.-W. lynchet of Ancient Field Group (22), revealed remains of an early phase of Iron Age 'A' showing features derived from the Late Bronze Age, succeeded by a phase associated with haematite-coated pottery. A circular hut of the earlier phase, 22 ft. in diameter with post-holes and a S.E. porch, was associated with a paved area to S. overlying some earlier post-holes. A broadly contemporary hut to N. was of 24 ft. diameter with remains of a clay and earth floor, a clay hearth and a small adjacent clay oven partly lined with sherds. Between them slighter traces of a hut perhaps 17 ft. by 24 ft., floored with broken limestone and with a hearth of stone slabs capped with clay, belonged to the later phase. Debris from a hand-cut shale armlet industry was associated with both phases, but other objects suggest normal domestic activities such as spinning and weaving, especially in the second phase. The lynchet, apparently developing after this occupation had ceased, had Durotrigian sherds towards its base and early Roman above, doubtless derived from nearby occupation of Iron Age 'C' or early Roman date, with evidence for a latheturned armlet industry, which Austen also recorded in the vicinity in 1856.

Previous excavations here, including those of c. 1936 which are probably wrongly recorded as at 94097748 some 250 yds. nearer Freshwater Steps, were thought to indicate the presence of roughly cobbled floors and drystone walls. They produced abundant material like that found in 1963–4, with Romano-British sherds of the 1st or 2nd century and perhaps the early 3rd century. (J. H. Austen, Purbeck Papers I (1856–7), 90; Dorset Procs. LXXI (1949), 52; LXXV (1953), 69; LXXVII (1955), 150; LXXXIII (1961), 83–4; LXXXV (1963), 98–9; LXXXVI (1964), 109.)

Corfe Mullen

Remains of the Roman road from Badbury Rings to Hamworthy can be seen on Barrow Hill and on much of the parish boundary with Poole further to the S. (see Roman Roads, p. 530). Romano-British sherds are recorded from a garden at Lamb's Green (99599871; Dorset Procs. LXV (1943), 58), an oven (99619885) to N. beside the Roman road and a pit (99729877) with early pottery to N.E. (ibid. LXXXVIII (1966), 115); see Fig.p. 530.

Sites (24) and (25) occupy the Plateau Gravel ridge overlooking the Stour valley to N.

(24) East End (SY 99 NE). A Pottery-kiln, Pits, Ditches, and Occupation Debris have been found about 200 ft. above sea level in the railway ballast pit beside the destroyed Roman road (992983).

A settlement, possibly established in Iron Age 'B' or 'C' and surviving at least to the mid 4th century, is indicated by storage and rubbish pits and alleged hut sites observed since 1927, a pottery-kiln, excavated in 1932, ascribed to c. A.D. 50–60 with contemporary pits and ditches, and possibly by a late Roman kiln. Early wares, including Continental imports, were sherds of Hengistbury class B, butt-beakers, Belgic platters, amphorae and S. Gaulish samian, with metallic-lustred ware of New Forest type indicating the latest known occupation. Iron slag suggested smelting; milling on a commercial or military scale is implied by a piece of upper stone (catillus) of a Niedermendig lava donkey-mill (see Fig.) in Poole Museum. (fn. 10) One or possibly two pits lined with clay contained carbonized grain, mainly Spelt wheat. (fn. 11)

The stoke-hole of the kiln alone survived (99219830). Its filling and that of four pits and two ditches some 15 ft. long found within 20 yds. of it, yielded analogous sherds and an as of Caligula of A.D. 40. Kiln wares included ring-neck flagons, mortaria, and thin-walled romanized versions of Durotrigian cooking-pots probably unintentionally oxidized. Six oxidized vessels (four in B.M.), found together in two rows in 1929, were probably rejects from a late 3rd or 4th-century kiln producing cooking-pots and bowls of Gillam's types 147 and 228 (Arch. Ael. xxxv (1957), 180–251). Objects are in B.M., D.C.M., Poole Museum and Christchurch Museum. (Ant. J. xv (1935), 42–55; Dorset Procs. LXV (1943), 57–8; Procs. Bournemouth Natural Science Society XLI (1951), 52–3; M. H. Callender, Roman Amphorae (1965), 156.)

(25) Cogdean Elms, East End (SY 99 NE). A Cremation Burial was found in 1865 in the old Corfe Mullen gravel pit, E. of the Roman road (995982). A Durotrigian bead-rim jar of Brailsford's class 4, containing burnt bones of an adult, a worn imitation as of Claudius I (41–54) and fragments of two glass phials distorted by fire, was found with amphora fragments 'in a gravel pit near the Cogdean Elms Inn, 1865' (B.M., Register of Antiquities, 1892.9–1.365–8). Several urns, undated, with charred bones, were found before 1847 apparently in the same pit. (Arch. J. LXXXVII (1930), 286; P.P.S. XXIV (1958), 116; Dorset Procs. XI (1890), 22.)

Dorchester

For Roman remains, see pp. 531–92.

Holme, East

Three coins, two of Constantine I (306–37), are recorded from gardens of Holme Priory, SY 897860 (Swanage Times, 17 Feb. 1954, and information from Mr. J. B. Calkin).

(13) Swanage Railway Bridge (SY 98 NW). Occupation Debris has been found in cleaning the bed of the Frome S. of Worgret (90808616). Romano-British pottery, possibly derived from a site on the Valley Gravel terrace to N., was first noted in 1899. Recent finds of coarse pottery (in D.C.M.) include sherds of the 3rd or 4th century. (Dorset Procs. XX (1899), 149.)

Kimmeridge (fn. 12)

Sites (16–18) are on Kimmeridge Clay.

(16) Metherhills (SY 97 NW). An Inhumation Burial and Occupation Debris indicating manufacture of shale armlets come from a low hill S. of the village overlooking Kimmeridge Bay (913792).

An oval stone cist lined with water-worn boulders, apparently plastered and containing human remains with a rough hand-cut shale ring on either side of the skull, and in the same field a large hoard of turned shale armlet cores, were found before 1856. Subsequent examination by Warne showed similar cores to a depth of 18 ins. Excavation on the S. slope (91357917) in 1946 indicated shale-working in an oval area some 40 yds. long below the 200 ft. contour; the remains were flint lathe-tools and worked flakes, slabs of unworked shale, and Romano-British sherds including 4th-century ware. (Purbeck Papers I (1856–7), 89, 90; Arch. J. XVI (1859), 300; Warne, Ancient Dorset (1872), Appendix, 329; Dorset Procs. XIII (1892), 183; LXXV (1953), 52, 69.)

(17) Kimmeridge Bay (SY 97 NW). Debris from salt-boiling exists above high water mark near the site of the former pier below the Clavel Tower.

The remains consist of a deposit at least 4 ft. thick of burnt and unburnt shale and fine grey soil in the talus at the foot of the hill, about 8 ft. above high water mark, extending for some 25 yds. (90807872 to 90817870) and containing saltboiling 'briquetage' and some vesicular slag; the deposit is cut through by remains of walls of relatively recent date. Burnt shale appears at a higher level some yards further S. The 'briquetage' includes hand-moulded supports and pieces of coarsely-gritted flat-bottomed containers about 1 in. thick. (Dorset Procs. LXXXIV (1962), 140, 142.)

(18) Swalland Farm (SY 97 NW). Occupation Debris associated with shale armlet industries, probably lron Age and Roman, comes from 400 yds. S.W. of the farmhouse (92547798).

The site, about 650 yds. N.E. of Clavell's Hard where the 'blackstone' band of the Kimmeridge shale outcrops at the base of the cliff, yields shale waste from a hand-cut armlet industry probably of the Iron Age, and from a turned armlet industry of the Roman period. A few sherds, perhaps Iron Age 'A' and Romano-British, come from 150 yds. W.N.W. (92427804). (Dorset Procs. XIII (1892), 183; LXXV (1953), 69; LXXXV (1963), 99.)

Knighton, West

Sites (26) and (27) are on the edge of Reading Beds overlying Upper Chalk.

(26) Little Mayne Farm (SY 78 NW). Occupation Debris has been found near the farmhouse (72308708). A late 1st-century samian cup and other Roman sherds (in D.C.M.) were found in 1930 and 1936 in tree-planting S.W. of the house above the 200 ft. contour (Dorset Procs. LXXIV (1952), 99).

(27) West Knighton (SY 78 NW). Occupation Debris has been found S. of the village (73278735). A quern-stone of Wessex type and a sherd of Iron Age 'C' or Roman date come from trenching in 1941 in an army camp W. of the road to Broadmayne about 200 ft. above sea level (Dorset Procs. LXXIV (1952), 100).

Langton Matravers

A few Romano-British sherds are recorded on the limestone ridge seawards of Spyway Barn, SY 995773, and on Wealden Beds near Windmill Barn, SZ 00928005 (Dorset Procs. LXXXV (1963), 102).

(42) Wilkswood (SY 97 NE). Occupation Debris, probably connected with the Purbeck marble industry and possibly with shale-working, has been found on Upper Purbeck Beds below 200 ft. on the N. slope of the limestone ridge (99377941).

The remains, excavated at the E. edge of a modern quarry in the Purbeck marble outcrop, consisted of debris apparently of a midden covering some 20 sq. yds., with scattered material for perhaps 100 yds. E. The bulk of the datable sherds, including samian, amphorae, colour-coated bowls and white flagons, as well as native wares, ranged from mid to late 1st century A.D., with a few pieces of Iron Age 'A' and late Roman ware. A sawn and polished piece of Purbeck marble, several worked and unworked fragments of shale, four turned armlets, and many flints including notched tools, suggest marble and shale-working, although characteristic armlet cores and lathe-tools are wanting. The site lies below remains of 'Celtic' fields of Ancient Field Group (28). (Dorset Procs. LXXV (1953), 52–3; LXXXI (1959), 121–2; Archaeological News Letter I, no. 11 (1949), 15.)

Sites (43–44) lie on Middle Purbeck limestone.

(43) Acton (SY 97 NE). An Inhumation Burial and Occupation Debris, probably connected with shale armlet industries of the Iron Age and Roman period, have been found in quarrying about 400 ft. above sea level near Blacklands.

The occupation site (98957774), largely destroyed by earlier quarrying, was tested in 1945, yielding a few Iron Age 'A' sherds, burnt daub, sling-stones, several unfinished hand-cut shale armlets and turned armlet cores of Calkin's class C, shale waste, and large concentrations of flint flakes including rough tools, at a depth of about 3 ft. (Dorset Procs. LXX (1948), 43; LXXV (1953), 69.)

The burial, some 500 yds. N. (989782), was extended with feet to S.E. in a covered stone cist without grave-goods, and is probably Romano-British (Dorset Procs. LXXI (1949), 66).

(44) Putlake Farm (SZ 07 NW). Inhumation Burials and Occupation Debris have been found S. of the farmhouse above the 300 ft. contour.

A collapsed stone cist, with two pieces of a coarse ware mug, perhaps a local copy of a 2nd-century type, lying on the cover-stones, was found in making a silo in 1957 (00197837). Remains of two inhumations without cists were found nearby about 1 ft. below surface, one with bowl fragments of Brailsford's Durotrigian class 1 (cf. P.P.S. XXIV (1958), 103); more burials were suspected in the W. side of the silo. Surface debris in the field, including beach-pebbles (sling-stones?), shale, and Romano-British sherds, suggest nearby occupation. (Dorset Procs. LXXXI (1959), 122–3.)

Two Roman coin hoards contained in vessels are untraced. One came from an extension c. 1842 to Leeson House (004786); the other, including a coin of Constantine, was ploughed up at 'Langton' in 1817 (Hutchins I (1861), 560; D.C.M. Parishes Index).

Lulworth, East

There are no known sites in the parish, but several discoveries suggest occupation near Arish Mell (SY 854803): a 4–lugged stone mortar (D.C.M. 0.231.1) 'from a barrow at Arish Mell'; a bronze trumpetbrooch (1934.7.2), found in 1932 at a depth of 18 ins. in laying electric cable 20 yds. from Maiden Plantation; and a rough quern-stone found near quantities of iron slag 'in breaking up a portion of the Down near Arishmill' (Warne, C.T.D. (1866), cpf, 26). A bronze mounting in the shape of a cockerel (1886.9.127) also comes from the parish.

Lytchett Minster

A few scraps of coarse ware, apparently Romano-British, occur in plough-soil within Bulbury Camp, where a hoard of Celtic metalwork probably dates from about the time of the Roman conquest (see Hill-forts, Lytchett Minster (30)).

Morden

(57) Charborough Park (SY 99 NW). Occupation Debris and Inhumation Burials have been found in the park. The former consisted of Romano-British sherds (not traced) found in 1956 in replacing a post in the deer-fence in High Wood about 150 yds. W. of Charborough Tower, on an outlier of Reading Beds over Upper Chalk about 300 ft. above sea level (92789752). The burials were found in the Chalk in making a plantation fence on the estate in 1864 about 100 yds. N. of a large barrow 'towards the west end of Charborough Down'; 14 to 16 skeletons, extended with feet to E., were placed parallel 2 ft. apart in a line N.-S. (Dorset Procs. LXXXV (1963), 102; Hutchins III, 505–6.)

Moreton

(31) Moreton Park (SY 88 NW). Occupation Debris comes from the lower N. slope of Fir Hill below the 100 ft. contour, on Valley Gravel. The remains consist of a few sherds of Iron Age 'C' or Romano-British coarse ware from plough-soil E. of the sunken fence (80808867). four turned shale armlet cores of Calkin's class C are preserved at Moreton House, residue of a quantity 'found in an earthen pot in digging the ditch to the fence on the addition being made to the Fir hill on the west side in 1807'. (Dorset Procs. LXXXV (1963), 99; LXXXVII (1965), III.)

Osmington

Part of a hand-moulded clay prop (in D.C.M.) probably from salt-boiling, comes from an unrecorded site on the cliffs at Osmington (Dorset Procs. LXXXIV (1962), 115).

(43) Landslip (SY 78 SW). Occupation Debris comes from W. of Sandy Barrow close to the N. edge of the landslip, on the crest of the Upper Greensand ridge above the 300 ft. contour. The remains consist of a scatter of Romano-British sherds, with some perhaps of Iron Age 'A', found in 1962 in ploughsoil for at least 100 yds. E. of the lane from Osmington (72778258 to 72908264). A few Romano-British sherds come from 250 yds. S., in the landslip (73908236). Other finds, not precisely located, include sherds probably of Iron Age 'A', and Romano-British ware, some of the 3rd or 4th century. Objects are in D.C.M. (Dorset Procs. LXXXIV (1962), 113; LXXXV (1963), 98–9.)

(44) Ringstead (SY 78 SW). An Inhumation Burial was found in 1926 in a chalk-pit at Spring Bottom (74708210). A shallow grave in Lower Chalk contained an extended skeleton with a one-handled bead-rim beaker of Romano-British coarse ware; iron nails in two lines by the skeleton, some with traces of wood, suggest a coffin. Objects are in D.C.M. (MS. note in D.C.M.)

Owermoigne

Native vessels, Iron Age 'C' or early Romano-British, were found outside and in the ditch of an enclosure at Bowley's plantation (see Enclosures, Owermoigne (36)).

Poole

Substantial remains of the Roman road from Badbury Rings to Hamworthy survive on the parish boundary with Corfe Mullen on the W., and further S. (see Roman Roads, p. 530). In addition to remains described below, intrusive inhumation burials, possibly Romano-British, have been found in a Bronze Age barrow near Cogdean Elms (see Round Barrows, Poole, Barrow Hill Group (P)). A beaker (in Poole Museum) and sherds, of New Forest ware, come from Branksea Avenue by Poole Harbour (989902; H. P. Smith, History of Poole I (1948), 88–9); Romano-British pottery is reputed to have come from the Lake Clay-pits (981906), and coins and tesserae from a field near St. Michael's Church, Hamworthy (993911; H. P. Smith, op. cit., 64, 75–6; Dorset Procs. XI (1890), 21). For a site of military character, lying mainly in the parish of Pamphill, see p. 529, and Fig. p. 530.

(402) Hamworthy (SZ 09 SW). Pits, Ditches, Ovens and Occupation Debris including salt-boilers' waste, of Iron Age and Roman date, have been found on Valley Gravel beside Holes Bay at Carter's Tile-Works, at or near the termination of the Roman road from Badbury Rings to the harbour (002904).

Excavation by H. P. Smith in 1926–33 (fn. 13) and 1949, supplemented by information of discoveries made in extensions to the premises and gravel-digging, indicated occupation over at least 6 acres from the W. end of the Tile-Works to the gardens of Ivor Road. The site is now a store-yard behind recently reclaimed land. The main features were three oval pits, some 5 ft. in diameter and 8 ft. deep, twelve ditches, apparently some 2½ ft. wide by 6 ft. deep, and eight oval or circular dished areas (five on Smith's 1928 plan) about 9 ft. or 10 ft. in diameter, some with traces of hearths, regarded by him as hut-sites. Hearths were also found outside the latter. The ditches, mostly, it seems, U-shaped with vertical sides, appear to have been straight, discontinuous features between 40 ft. and at least 125 ft. long; most ran N.-S. with a few N.W.-S.E., but three, disposed radially, met in a 'silt-hole'. All were presumably for drainage, although one (Smith's T4), 40 ft. long and presumably of normal depth, ended in an oven (Smith's 'pottery-kiln') described below, while another (T 9), of flask-shaped section, was filled with the local mottled red and white clay.

The pits and some of the ditches, reaching present watertable, and all the 'hut' areas, were filled with black earth containing most of the 'late Celtic' (Iron Age 'C') pottery. All were assigned to the pre-Roman phase, but at least one pit and one ditch, cut through the 'Early Iron Age stratum', appear to be of Roman date (Dorset Procs. LVI (1934), 20, plate 11).

The dished areas, excavated some 10 ins. in the gravel, and perhaps working areas rather than hut-sites, were between 40 ft. and 50 ft. apart and adjacent to pits and ditches. One (H 3) had a circular hearth of clay and flints, with iron and iron slag nearby; another (H 5) had remains of a clay hearth overlying its black occupation soil. Both were sealed by compact shingle, containing Roman sherds, believed to belong to a branch of the Roman road, running E.S.E. parallel with the S. side of Messrs. Carter's premises (see Road II, p. 531).

Further structural evidence for the Roman phase is scanty, and only a few pieces of roof and flue-tile occurred. Two floors of 'coarse red tile' (broken brine-containers) were thought to belong to rectangular huts; one (RH) overlay a hearth of the pre-Roman phase. The circular clay oven at N. end of ditch 4, with base some 5 ft. below surface, is probably of Roman date. As preserved in the Old Town House, Poole, its internal diameter is 18 ins.; the walls, some 8 ins. thick, survive to 14 ins. high, and the slightly concave floor is 12 ins. thick. The damaged front suggests an aperture some 9 ins. wide. No trace of an upper floor survived, and its size, and the absence of wasters, preclude its use as a pottery-kiln. It was associated with hand-moulded clay supports and pieces of thick straight-sided containers with rounded corners, of gritty clay, which were generally widespread in the Roman stratum and in the upper filling of the ditches nearest Holes Bay. This 'briquetage' indicates a salt-boiling industry in this phase. Two or three oven bases were found together in 1949 to N. (00249045) with similar 'briquetage' and Romano-British sherds.

Many fragments of shale, a few unfinished hand-cut armlets and three lathe-turned armlet cores of Calkin's classes A and C, suggest shale industries in both phases, but specialized tools for shale-working are unrecorded among the many flints.

The pottery (in Poole Museum) comprises a little native Iron Age 'B' (Southern Third 'B') ware and much Iron Age 'C' and Romano-British pottery including New Forest and other 4th-century types. Five vessels of the 3rd or 4th century came from harbour mud at the new quay (01049002) in 1931. Extraneous pre-Roman wares include a few pieces of 'Glastonbury' ware and of 'graphite'-coated jars comparable with class H at Hengistbury Head, Hampshire, possibly from N. France. (fn. 14) Fragments of amphorae need not be earlier than the Roman conquest. The initial occupation need not be dated before the early 1st century B.C., although a coin of Tauromenium, Sicily, of the 3rd century B.C., found near the site, and a Siculo-Punic coin of the 4th century B.C. from the harbour shore, are local examples of a number of pre-Imperial coins from the Mediterranean occurring in the hinterland of Poole Harbour and thought to indicate pre-Roman trade. (fn. 15)

Imported early Roman wares include Gallo-Belgic terra nigra and terra rubra, ring-neck flagons, colour-coated beakers, and pre-Flavian samian, The only coins from the site are two of Claudius of A.D. 41–2. This material points to Claudian occupation, although the best evidence for a military stores base (cf. Arch. J. CXV (1959), 57) is the complete Niedermendig lava slave or donkey-mill from the site, in the British Museum, (fn. 16) with part of another at the Claudian site at Corfe Mullen (Corfe Mullen (24)) 5 miles N. on the same Roman road. (Dorset Procs. LII (1930), 96–130; LIV (1932), 5, 13–14; LVI (1934), 11–21; LXXI (1949), 66; LXXXIV (1962), 137, 142; H. P. Smith, History of Poole I (1948), 51–76.)

Sites (403–4) occupy sandy Bagshot Beds.

(403) Turlin Moor (SY 99 SE). Occupation Debris was found in 1963 in a housing estate in the N. part of the Hamworthy peninsula (978916). The material from drainage trenches comprises samian and Romano-British sherds, and some fragments resembling salt-boiling 'briquetage', in black soil some 2½ ft. below surface (Dorset Procs. LXXXVI (1964), 115–6).

(404) Sterte (SZ 09 SW). Occupation Debris and a Coin Hoard have been found near the E. shore of Holes Bay, Poole Harbour (00939165).

The hoard (in Poole Museum), found in remains of a coarse ware jar 2 ft. below surface in laying water-mains in 1930, consists of 965 mid 3rd-century coins (964 antoniniani,1 as) ending with Aurelian (270–5), and was doubtless part of the same Aurelianic hoard as were over 300 coins (not traced) discovered in a broken urn in the same meadow in 1833. Iron Age 'C' and Romano-British sherds have also been found. (Dorset Procs. LII (1930), 127–8; H. P. Smith, History of Poole I (1948), 87–8, 90–1; Bournemouth Echo, 12 Nov. 1930; Num. Chron. new ser. IX (1869), 283–5; 5th ser. XIII (1933), 229–32.)

The provenance of 34 antoniniani (in Poole Museum) stated to have been dug up as a hoard in 1936 in allotments at Baiter (018902) is doubtful (Num. Chron. 5th ser. XVIII (1938), 300; H. P. Smith, op. cit., 91). The series resembles that of the Dorchester hoard found at the same time (Dorchester (194a)).

Portesham

An intrusive burial of a child in a stone cist of Roman date came from a barrow on Ridge Hill in 1885 (see Round Barrows, Portesham (43)). 'Human remains found A.D. 1858' are recorded on O.S. maps near The Buildings (61168606); a group of undated inhumations in stone cists aligned N.W.-S.E. was found on Corton Farm in 1765 (Hutchins II, 761–2) and one on Corton Down (63668674) in 1936 (MS. note by C. D. Drew, in D.C.M.). Three cist burials extended with heads W., found in a landslip on the downs near Portesham, were described in Dorset Procs. XXIV (1903), xliv-v; another on an adjacent hill was thought Saxon.

(63) Tatton Farm (SY 68 SW). Inhumation Burials have been found on the Langton ridge (Cornbrash) S.E. of Langton Cross. Many burials in stone cists were ploughed up in 1883 on 'Tatton Hill' (approx. 629819) adjoining similar remains in Chickerell parish (Chickerell (35)), and skeletons found on Tatton Farm before 1845 probably come from the same cemetery. (Dorset Procs.L (1928), 123; Gentleman's Magazine (1845), pt. i, 79.)

(64) Portesham Withy Bed (SY 58 NE). Occupation Debris was found c. 1901 at about 350 ft. above sea level 'exactly six furlongs due west of St. Peter's Church' (approx. 590859). The site, near the junction of Upper Greensand and Lower Chalk, occupied a shelf in the W. slope of a re-entrant valley, among 'Celtic' fields (see Ancient Field Group (4)). The remains, from foundations 1½ ft. deep for a gamekeeper's house, included flint and chert flakes and tools, and pottery, thought prehistoric to late Roman (Dorset Procs. XXXVII (1916), xxxix-xl; XXIV (1903), xlv).

Portland

In addition to sites described below, a stone-lined well found 'a little to the south of the parish church' (St. George's, Reforne, SY 68657201) was said to have contained much Roman pottery (J.B.A.A. XXVIII (1872), 32–3); a coin of Marcus Aurelius (161–80) was found at Portland Bill (67836874; Dorset Procs. LXXXIV (1962), 112); Roman coins and other objects, possibly from wrecks, have been found on Chesil Beach one mile from Portland (about 675744; Num. Chron. II (1840), 255–6). Of several unlocated graves (Dorset Procs. XXXIII (1912), x; Dorset Album I, part I, in D.C.M.; P.P.S. XXIV (1958), 118), the most notable one contained a bronze collar, coarse ware jar, 2nd-century samian bowl and Late Bronze Age knife, in B.M. (Guide to Early Iron Age Antiquities (1925), 150–1; P.P.S. XXIV (1958), 118).

(99) North Common (SY 67 SE). Inhumation Burials and Occupation Debris have been found in North or Verne Common, Castletown, on Kimmeridge Clay.

About 20 skeletons 'in a cramped or sitting posture', with limpet shells and pottery believed Roman, were found in making the old reservoir c. 1855 (68777429; C. King Warry, Old Portland Traditions (c. 1908), 35). Several cists containing skeletons, some E.-W., were found in drain-laying c. 1860; one, with sides and cover stones of Kimmeridge shale, had nails suggesting a wooden coffin. Two iron ingots (in Portland Museum), (fn. 17) two 2nd-century coins, a handled bead-rimmed beaker and enamelled disc-brooch (D.C.M. 1923.3.4 and 7) were amongst objects found nearby. (Arch. J. XXIII (1866), 75; XXV (1868), 52–3, 56–7; Royal Engineers' List, copy 1882 by Major Peck, in D.C.M.; R. Damon, Geology of Weymouth (2nd ed. 1884), 240–2; Dorset Procs. XLIV (1923), 35–6, 38, 46, 48.)

Six apparently adult inhumations were found in laying water-mains in 1950 in Zigzag (now Verne Common) Road on the steep slope E. of Victoria Gardens. Two, probably of the first half of the 2nd century A.D., lying N.-S. with heads N., 230 ft. above sea level (68747380), had been compressed into short stone cists 3 ft. 4 ins. and 3 ft. 9 ins. long. The northernmost had a coarse ware jar and dish (cf. Arch. Ael. XXXV (1957), types 123, 307); the other had one coarse ware vessel (stolen). A third cist, some 40 yds. S.W. (68727377), orientated N.E.-S.W. with head to N., was 6 ft. 5 ins. long and had nails indicating a wooden coffin. The fourth and fifth graves, at the foot of the road about 100 ft. above sea level (68487366), were damaged and of doubtful orientation. The northernmost was apparently in a full-length cist, with a Trajanic samian dish (form 18) of Secundus and a coarse ware jar of the same type as in grave 1; the fifth, in an unlined grave, had Trajanic samian dishes (forms 18/31 and Curle II), and the base of a coarse ware jar. A sixth burial to N., E. of the Police Station (68517377), apparently without grave-goods, was in a diagonally tooled Portland stone sarcophagus with ridge-roofed lid (in Portland Museum), 6 ft. 4 ins. long by nearly 2 ft. wide overall. Objects are in D.C.M. and Portland Museum. (Dorset Procs. LXXII (1950), 83.)

(100) The Verne (SY 67 SE, 77 SW). Inhumation Burials and Occupation Debris have been found in and near the Verne Citadel, now H.M. Prison, on Portland Stone some 400–450 ft. above sea level.

According to Hutchins several inhumations in covered stone cists were found in the landslip of 1734 on the shore between the 'old pier', probably King's Pier, and Portland Castle, probably about 700735 towards N. end of East Wear. Four inhumations in cists, one of which was at least partly of shale, were found in building the Citadel c. 1860, 4 ft. deep within the destroyed enclosure (68987354) to S.W. (see Enclosures, Portland (98)). Three (one perhaps of a child with head W.) lay E.-W.; one to W. lay N.-S. with head N. Sherds were also found. In the same works, at 'top of Verne Hill', a cist of Portland stone slabs contained three burials, one with a large urn containing some charcoal.

Another cist, of beach-derived Portland stone slabs, laid 3 ft. deep in a deposit of 'glacial drift earth', (fn. 18) was found in 1882 on the S.E. Glacis (about 695734), and perhaps two burials c. 1878, one in a cist, the other with part of a Celtic bronze mirror (C. Fox, Archaeologia Cambrensis C (1949), 40). Two Romano-British coarse ware dishes (not traced) and a jar (D.C.M. 1923.3.5) also came from the S.E. Glacis according to the R.E. List. Remains of two burials 10 ft. apart were found in making a football ground here in 1933, about 5 ft. below surface (69487336). One was in a stone cist; the other had two 'Romano-British food-vessels' by the skull (C. H. Wood-house). A circular drystone wall some 6 ft. in diameter and 5 ft. high, perhaps a 'beehive' chamber (cf. Monuments (101) and (105) below), was found c. 1860 near the S.E. Demibastion (about 695735), containing an ox-skull, bones and ashes. An undated shell-midden found before 1882 in glacial drift beside the road below the S.W. Glacis (about 691734) contained flint scrapers.

Objects from these works include a waste core from a turned shale armlet, a spiral bronze ring from a finger-bone, and a gold coin of Allen's Gallo-Belgic type F. Other finds from the area include a lidless diagonally-tooled Roman sarcophagus of Portland stone, 6 ft. 2 ins. by 2 ft. wide overall, preserved at the Governor's house, and, in D.C.M. and Portland Museum, Iron Age 'C' or early Romano-British vessels and sherds, samian ware, fragments of 4th-century colour-coated ware, Roman glass beads and part of an incised circular shale plaque. (Hutchins II, 826; Arch. J. XXIII (1866), 75–6; XXV (1868), 47–9, 53–6; Royal Engineers' List, copy 1882 by Major Peck, in D.C.M.; R. Damon, Geology of Weymouth (2nd ed. 1884), 240–2; Dorset Procs. XLIV (1923), 35–7, 47; Brig. C. H. Wood-house, MS. note, 1933, in D.C.M.)

Sites (101–105) occupied Lower Purbeck Beds now largely removed in quarrying.

(101) King Barrow (SY 67 SE). 'Beehive' Chambers and Inhumation Burials have been found in quarrying in King Barrow and Withies Croft E. of the Verne YeatesEaston road some 400 ft. above sea level.

Between 1880 and 1884 over a dozen circular corbelled stone chambers or cists probably of Iron Age and Roman date, in two cases interconnected, were found within an area of ½ mile in Weston's and Steward's quarries at King Barrow, mostly together. One of these is recorded by the Geological Survey (6-in. MS. sheet, Dorset LVIII SE, 1890) at 69207285. Slightly divergent descriptions by Damon and Holmes derive from the superintendent, A. M. Wallis. The chambers were normally some 9 ft. in diameter and 8 ft. high with extremes of 6 ft. by 4 ft. and 12 ft. by 8 ft.; holes dug in limestone rubble to an underlying clay seam were lined with flat stones, corbelled towards the top to a man-hole some 16 ins. in diameter normally with a slab cover at the base of the overlying humus. Some were seemingly corbelled from the base. Most stood alone, some 10 or 15 yds. apart, but two were superimposed with the entrance to the lower centrally in the floor of the upper chamber; another pair was joined at the base either directly (see Fig., after Holmes) or by a short lined tunnel (Damon). Domestic animal bones, limpet shells (in D.C.M.), round pebbles ('corn-crushers') from Chesil Beach, 'slingstones', flint flakes, (fn. 19) pieces of Kimmeridge shale and a bronze coin were said to have been found in these chambers, and in one was a quantity of carbonized grain. (fn. 20) Oblong cists found nearby are supposed to have contained human and animal bones, ornaments, an iron sword, sherds and quern-stones, amongst which a rotary quern (fn. 21) and perforated whet-stone of micaceous schist, (fn. 22) in D.C.M., are probably recognizable.

Two similar chambers were found here in 1885 (fn. 23) and 1886, the former with stone 'spindle-whorl' and roughly-hollowed pebble 'mortar' outside. The other, containing a few 'sling-stones' and limpet shells, was near a cist with two superimposed extended skeletons, the uppermost on right side with head E.; 2 in. iron nails suggested a wooden coffin. Each had a pot; the lower had two metal finger-rings, one with 'zig-zag ornament'. Another chamber found c.1890 was near a cist containing human bones and sherds, with two querns and hammer-stones nearby.

Other burials probably in stone cists were noted in quarries superintended by Wallis: a grave, 1888, with iron knife, two iron rings about 1½ ins. in diameter, three flint hammer-stones, and three coarse ware vessels of the 3rd or 4th century (D.C.M. 0.208. 1–8; cf. Arch. Ael. XXXV (1957), types 65, 228)); three graves, c. 1890, two with an urn and one with a ring and beads; a grave with skeleton in 'crouching position' with two stone 'spindles', three large round stones, and a piece of worked shale. Two pear-shaped jars in Portland Museum, Iron Age 'C' or Roman, come from King Barrow Quarry, and a wheel-made jar of Romano-British grey ware and two 2nd-century dishes (D.C.M. 1889. 4. 1–3) are probably from the same site. (T. V. Holmes, Procs. Geologists' Association VIII (1883–4), 404–11; R. Damon, Geology of Weymouth (2nd ed. 1884), 164–6; MS. anon., 'Early Interments on Portland' (1886, in D.C.M.); Dorset Procs. XII (1891), xix, xxxii, II; XLIV (1923), 47; H. J. Moule, MS. List of Roman Pottery in D.C.M.)

In Withies Croft Quarries to S., a stone cist found in 1868 contained a coarse ware beaker and bowl, probably of the 2nd century, and in 1879 a grave with skeleton and two jars was discovered in the 'Maggot quarries' (691726). A hoard of eleven 3rd-century radiate coins, in Portland Museurn, came from the E. end of lawn 612 in Inmosthay Quarries (69007216) in 1896. (Arch. J. XXVII (1870), 217; J.B.A.A. XXVIII (1872), 34; D.C.C., 29 May 1879; Southern Times, 20 Feb. 1897.)

(102) The Grove (SY 67 SE). Inhumation Burials and Pits were found in 1851 in the Breakwater Quarries (now playing-fields) opposite the Clifton Hotel near the Convict Prison (now H.M. Borstal Institution) some 350 ft. above sea level (about 697725).

Nearly 200 graves are said to have been found, mostly stone cists but including two diagonally-tooled Portland stone sarcophagi (at Borstal Institution) with 3 in. walls and ridgeroofed lids about 1 ft. high. The smaller (see Fig.) is 6 ft. 3 ins. long overall; the larger (Plate 229), 6 ft. 10 ins. by 2 ft., and 1 ft. 6 ins. high, had remains of two skeletons head to foot, apparently without grave-goods. A stone plinth (Plate 229; at Borstal Institution), 7 ft. by 4 ft. and some 6 ins. thick, similarly tooled and perhaps for a sarcophagus, consisted of four pieces of unequal size with a semicircular channel at edge; it was laid in clay, or cement (G. Clifton). Some short cists with single cover-stones, in some cases with single slabs for sides, had skeletons with knees drawn up; others were extended in normal cists. A contemporary sketch by Poppleton indicates a width of 4 ft. 8 ins. for a 7 ft. 3 in. cist. The burials were from 1 yd. to 20 yds. apart, the orientation being unstated except that some in short cists had head to N. Primary sources state that vessels of black and dishes of red ware (in one case certainly samian with leaf-decorated rim) accompanied some of the skeletons; a skull of a calf and a quern-stone also came from graves. Roman coins and a pounding stone, and deep pits containing animal bones (some of birds), were found nearby. A later authority (Clifton) claims many double graves, metal equipment including javelin-head and shield fragment, round stones, ornaments and coins of Caligula (37–41), Hadrian (117–38), Constantine (306–37) and Julian (355–63). A Celtic mirror-handle (in Coburg Museum; Archaeologia Cambrensis c (1949), 29, 30) found in 1875, is probably from the area. (R. Poppleton and Jones, correspondence in D.C.M.; D.C.C., 9 Oct. 1851; Arch. J. X (1853), 60–1; G. Clifton, J.B.A.A. XXVIII (1872), 204–5.)

(103) Broadcroft Quarries (SY 67 SE). Occupation Debris has been found in S.E. extensions to the quarries, near Easton, above the 300 ft. contour.

Limited excavation in 1950–1 exposed an occupation deposit 1 ft. to 1½ ft. below surface at the then E. edge of the quarries (69857193). Remains included a few samian sherds, much Romano-British coarse ware, some of the 4th century, a shale spindle-whorl, a rotary quern-stone of Wessex type, two pebble hammer-stones, some 50 sling-stones, winkle, limpet and oyster-shells, and a coin of Victorinus (268–70). Amongst earlier finds to W. were a coin of Trajan (98–117), and, according to quarrymen, stone cist burials and a small covered cist filled with limpets. (Dorset Procs. LXXII (1950), 87–8; LXXIII (1951), 93–4.)

Human remains found 350 yds. S.S.W. in 1951 in a fissure (69737165) N. of Bumper's Lane, with bones of domestic animals, several limestone and quartzite pebble hammer-stones (one perforated), and flakes and cores of Portland chert, have been referred to the Bronze Age or earlier (Dorset Procs. LXXIV (1952), 39–47; LXXXIX (1967), 120), but such objects are matched at Iron Age and Roman sites in the Isle of Portland. A 'beehive' chamber, some 10 ft. deep, is said to have been destroyed in quarrying N. of the railway c. 1920. (Information from Mr. G. H. Lynham, Director of The Stone Firms Ltd.)

(104) Bottom Coombe Quarries (SY 67 SE). An Inhumation Burial was found in the late 19th century some 250 ft. above sea level in the quarries (693715) W. of Wakeham Street, Easton. The burial was in a stone 'coffin', probably a cist. Several Roman coins, 'republican' to 4th-century A.D., are said to have been found at the same time. (Information from the late J. Pearce, J.P.; Dorset Procs. XXXVIII (1917), xxvii.)

(105) Coombefield Quarries (SY 67 SE). 'Beehive' Chambers, Inhumation Burials, and Occupation Debris have been found in the quarries (689706) at Weston some 200 ft. above sea level.

A corbelled stone 'beehive' chamber found in 1897, diameter about 5 ft., contained human remains according to B. M. Head. Some 2 yds. away, an oblong chamber with rounded corners, 7 ft. by 4½ ft. and 3 ft. high, lay N.N.W. to S.S.E., walled with Purbeck stone partly corbelled and closed with flat slabs; the filling, mainly of stones, included sherds, a crude stone 'mortar', a flint hammer-stone or 'pestle', flint flakes, 'slingstones', pebbles, charred wood, bones of ox and sheep, and shells of cockle and limpet. A covered cist of orthostats backed with horizontal slabs, similarly aligned but 50 yds. N.W., had contained a skeleton.

A stone-lined 'beehive' chamber, 4¼ ft. high, diameter 5 ft., examined by E. Cunnington in 1898, had an arched passage or 'entrance' on the N. side, 8 ft. long with a circular hole 4½ ft. deep in its floor. The chamber contained a rough mortar of grey stone and an elongated pebble 'pestle' (in D.C.M.), and three sherds of black ware, of which two surviving are rims of hand-made jars of Iron Age 'B' or 'C' tradition (in D.C.M.; see Fig.)

Two photographs in D.C.M. depict additions to the series. One, given by Head in 1906, is of a corbelled 'beehive', diameter about 4 ft., probably the empty structure referred to in V.C.H.; the other, taken by H. Colley March, was of a 'beehive' found in 1907, 4¾ ft. in diameter, with vertical walls surviving to 4½ ft. (Dorset Procs. XIX (1898), 128–9; XX (1899), 120–1; XXVIII (1907), lxxvii; XXIX (1908), cxvi; E. Cunnington, 'Memorandum of discovery of bee-hive building at Coombe Field Quarry' (MS. 1898, in D.C.M.); V.C.H., Dorset II (1908), 338.)

Sites (106–107) occupy Lower Purbeck Beds.

(106) Avalanche Road, Southwell (SY 67 SE). Occupation Debris, Iron Age 'A' to late Roman, comes from allotments and gardens adjoining St. Andrew's Church, about 175 ft. above sea level. The remains consist of sherds of Iron Age 'A' and Roman date, the latter including samian and 4th-century ware, mainly from immediately S.E. of the church (68707016) and from the garden of 'Coombe Ridge' to N. (68717022), with scrapers and flakes of Portland chert. A quern-stone of Wessex type, a flint hammer-stone, and a three-lugged mortar of Purbeck burr-stone (D.C.M. 0.230.1), are also recorded. Eliot noted Roman pottery which he believed to be associated with traces of a circular embankment in the vicinity. (Dorset Procs. LXXXV (1963), 101; G. Eliot, J.B.A.A. XXVIII (1872), 33.)

(107) Sweet Hill, Southwell (SY 66 NE). Inhumation Burials some 400 yds. S.W. of Monument (106) were found in road-making in 1950 below the 200 ft. contour. A diagonallytooled Portland stone sarcophagus with ridge-roofed lid (in Portland Museum), 7 ft. long by 2 ft. wide overall, freshly disturbed but containing hob-nails, was at 68386994, orientation unknown. Some 3 yds. S.E. an extended skeleton on axis of 140° with head S.E. and hob-nails at feet lay on loose stones over loamy subsoil, in a covered stone cist some 6 ft. long by 2 ft. wide overall, with slabs leaning outwards slightly. In the face of the cutting some 5 yds. N.W. of the sarcophagus, the corner of a destroyed cist, axis either 130° or 220°, had foot-bones, hob-nails, parts of a coarse ware jar of the 3rd or 4th century, and nails with adherent wood suggesting a wooden coffin. Objects are in D.C.M. (Dorset Procs. LXXII (1950), 83.)

Poxwell

Finds indicating an occupation site associated with Ancient Field Group (11) were made during excavation of strip lynchets in 1967 (SY 736840; Dorset Procs. LXXXIX (1967), 135–9).

Stafford, West

(25) Crofts (SY 78 NW). Inhumation Burials were found c. 1850 above the 200 ft. contour on Upper Chalk (724890). Of two burials found near the 'skeleton of a horse' and other animal bones, about 20 yds. from a barrow (Round Barrows, Stafford, West, (18)) in the field called Crofts S. of Parsonage Plantation, one had a samian bowl (form 22) of the 1st century A.D. apparently placed between the legs; the other was near five vessels of Durotrigian coarse ware. The vessels are in D.C.M. (Arch. J. IX (1852), 93; Hutchins II, 517.)

Steeple

Sites (28–29) are on sandy Bagshot Beds.

(28) West Creech (SY 88 SE). Occupation Debris comes from a low ridge about 150 ft. above sea level E. of North Hills Plantation (898827). The remains consist of a thin scatter of Romano-British pottery first noted in rabbit-holes in 1953, and subsequently after ploughing (Dorset Procs. LXXVI (1954), 86).

(29) Creech Grange (SY 98 SW). Occupation Debris and some traces of structures have been found at the house, which lies at the N. foot of the Chalk range (911822).

Much black and red Romano-British pottery, some figured in Purbeck Papers and perhaps mainly of the 2nd and 3rd centuries, and a samian sherd stamped Genitorf, were recovered from black mould in alterations to the kitchen garden in 1858. Quern-stones, pieces of shale and waste cores from turned armlets were also found. Amongst a quantity of stones, three stone pillars, 2½ ft. high with the lower 6 ins. left rough, were unburnt but thought to be supports for a drying floor, and there were two or three perforated fragments of baked clay 1 in. to 1½ ins. thick. Evidence for a pottery-kiln is inconclusive. A Durotrigian bowl of Brailsford's class 1 is in D.C.M. Coarse ware sherds found in 1958, one from a flanged bowl of the 3rd or 4th century, came from building work in the N. side of the kitchen garden (91098236). (Purbeck Papers I (1859–60), 215–20, 225; Arch. J. XV (1858), 87; Dorset Procs. LXXXVI (1964), 116.)

(30) Smedmore Hill (SY 98 SW). Inhumation Burials have been found at the W. end of Smedmore Hill in or near the upper quarry in Lower Purbeck limestone, some 400 ft. above sea level.

Nine or ten burials, probably all extended in covered cists, have been found since 1944. Two side by side in the quarry (about 91908002) were of a young adult and a child, extended approximately N.-S. with heads N.; the longer cist had shale ends. Most of the others were exposed in the E. face of the sunken approach. One, extended W. by S. with head W., was at 91938007 near the entrance. Three were side by side some 18 yds. further S. at 91928005. Two of these, containing adults with heads W.S.W., were almost wholly of shale slabs; one of them, 6 ft. 4 ins. long, had no end slabs. The third, almost destroyed by 1958 but apparently on the same alignment, may have been one of two unexamined cists seen in 1952. Human bones were also observed in the W. face of the approach. In 1967 another cist largely of shale, with an adult burial aligned E.-W. (head), was found at the edge of the field to N. (91968012). (Dorset Procs. LXXIV (1952), 49–50; LXXXI (1959), 94–7; LXXXIX (1967), 145.)

(31) Kimmeridge Bay (SY 97 NW). Inhumation Burials, Working-floors and Occupation Debris indicating shaleworking and salt-boiling from Iron Age 'A', have been found on Kimmeridge Clay in the cliff about 50 ft. above sea level W. of Gaulter Gap.

The cliff is being eroded and much material has been excavated from the site, by Miles, Austen and others, since 'coal-money' was first recorded here by Hutchins in 1768. The sequence of occupation has been determined by clearance of the cliff-face at seven points by Dr. Henrietta Davies and excavation of some 80 sq. yds. close to it by Mr. J. B. Calkin (90707917); the site extends some 240 yds. W. of the Gap (90787911 to 90607923) and perhaps for some 200 yds. inland.

An occupation stratum on virgin soil at a depth of about 3 ft. was associated with working-floors of a hand-cut shale armlet industry presumed to have obtained its raw material from the 'blackstone' or 'Kimmeridge coal' outcrop at Cuddle ½ mile S.E., and employing semi-standardized flint tools. Pottery refers it to an early phase of Iron Age 'A'. A sterile deposit above indicated a hiatus, followed by a lengthy further occupation in Iron Age 'A'; this was marked by haematitecoated pottery, in a loamy deposit some 15 ins. thick in part made up with beach-derived sand, and containing several thin lenses of burnt material, in two main horizons, including shale pebbles used as fuel. One of these burnt floors was at least 20 ft. by 12 ft. Short lengths of straight and curved drystone walls and pieces of burnt daub were interpreted as remains of huts at least once replaced, but similar structures at Gallows Gore (see Worth Matravers (40)) appear to be directly associated with shale-working. There was some evidence for continued production of shale armlets in this later phase. Pieces of small cylindrical jars of coarse reddish clay, apparently cut in half to form troughs with one open end and thought to relate to a salt-boiling industry, were mainly associated with the burnt layers. Pieces of small bowls of similar ware found in a heap of unburnt shale rarely occurred elsewhere in the deposit.

The upper level of some 12 ins. to 18 ins., below topsoil, contained Romano-British pottery of the 2nd to the 4th century, with some admixture of native types of perhaps earlier date. Beach-pebbles regarded as sling-stones and limpet shells occurred in all occupation phases. A drystone wall is recorded, but otherwise daub and groups of cobbles were the only evidence of structures in this phase. Abundant waste cores, mainly class C, often in small concentrated groups, and their associated debris including specialized flint tools knapped on the spot, point to a very considerable turned shale armlet industry, in which undecorated armlets of light-weight type were recognized.

A cist burial, dated to the end of the 3rd century or later by the only coin recorded from the site, was cut through the workshop floor at 90707917. This was of an elderly arthritic female, with the severed head and detached mandible placed beside the shins, and a shale spindle-whorl beside the left hand; the cist of limestone and shale slabs was orientated 30° W. of N. and had been filled deliberately with soil and a deposit of limpets and snails (Helix aspersa). A coin of Carausius (286–93) was in the grave-pit. On the capstones a secondary burial in nearly the same alignment was of another elderly female with detached mandible near the knees and shale spindlewhorl by the left hand.

Miles' excavation yielded much material including a mortar of 'granite', probably Purbeck marble, and a supposed bullock sacrifice in a pentagonal cist of shale slabs; his pieces of handmoulded clay associated with burnt shale strongly suggest the props used in local saltings of the Roman period, not otherwise clearly attested here. Romano-British sherds and several armlet cores of classes A and C (in D.C.M.) were found by Dr. W. H. C. Frend some 200 yds. inland beside the stream N. of the cottages (90837930). Objects are in B.M., D.C.M., Christchurch Museum, and London University Institute of Archaeology. (Hutchins, in Gentleman's Magazine (1768), 111–2; Hutchins I, 556–7; W. A. Miles, The Deverel Barrow (1826), 39–44; J. H. Austen, Purbeck Papers I (1856–7), 82–97; Arch. J. XCIII (1936), 200–19; Dorset Procs. LXIX (1947), 33–41; LXX (1948), 29–59; LXXV (1953), 45–71; LXXXIV (1962), 139–41.)

Stoke, East

(51) St. Mary's Church (SY 88 NE). Occupation Debris comes from the graveyard extension beside the R. Frome on Valley Gravel below the 50 ft. contour (87288702). Romano-British sherds and a piece of shale core from a turned armlet, in D.C.M., were found at a depth of 2 ft. to 3 ft. in digging a grave in 1932. (Dorset Procs. LXXIII (1951), 92.)

(52) Rushton Common (SY 88 NE). Occupation Debris comes from Valley Gravel and Alluvium about 15 ft. above sea level beside the Frome some 350 yds. S.W. of Holme Bridge (88788650). Romano-British coarse pottery including 4th-century types and many worked flints, in D.C.M., were found in the meadows in 1947, apparently on the flood-plain. (Dorset Procs. LXXI (1949), 65–6.)

(53) West Holme (SY 88 NE). Occupation Debris comes from Bagshot Beds and Valley Gravel near the S. bank of the Frome, mainly below the 50 ft. contour. Romano-British coarse pottery, some in D.C.M., has been found on the surface and in test pits, on both sides of the road to Holme Bridge between about 88668600 and 89008637. (Dorset Procs. LXXIII (1951), 92.)

Studland

Romano-British pottery is said to have been found on Brownsea Island in Poole Harbour (Dorset Procs. XXVI (1905), 187).

Sites (44–47) occupy sandy Bagshot Beds.

(44) Kingswood Farm (SY 98 SE). A Floor and Occupation Debris have been found about 250 ft. above sea level on a S.E. slope at the foot of the Chalk downs. Deep ploughing c. 1953 disclosed remains of a rectangular floor of sandstone slabs at 99828201 with Romano-British sherds some 25 yds. to N.; no walls were noticed. (Information from Mr. Elford, Kingswood Farm.) Sherds (in D.C.M.) found in 1958 include one of the 3rd or 4th century.

(45) Godlingston Heath (SZ 08 SW). Occupation Debris including coarse 'briquetage' comes from an old sand or gravel pit E. of Thorny Barrow about 300 ft. above sea level. In the N. face of the pit, at a depth of about 2 ft., deposits of clay and sandy soil showing signs of burning were traced laterally for some 15 ft. in 1958 (01478212). Many fragments of reddish 'briquetage' of coarse sandy finger-marked clay, and Romano-British sherds, some oxidized, ranging from at least the 2nd to the 3rd century, come from these deposits and from the filling of old excavations in the sand some 20 yds. away in the S. face of the pit (01468210). Similar material occurs in rabbit-scrapes on the N. slope of the hill (01508213). The 'briquetage' consists mainly of short hand-moulded props with flat or grooved ends and pieces of large thin-walled containers, and compares closely with that from salt-boiling sites on the fringes of Poole Harbour. A large curved fragment of similar fired clay some 2 ins. thick may be part of an oven or kiln. (Dorset Procs. LXXXIV (1962), 141–2.)

(46) Woodhouse Hill (SZ 08 SW). Buildings and Huts in two groups, dating perhaps from just before the Roman conquest to the 4th century A.D., were excavated by Mr. N. H. Field in 1952–8 on the S. slopes of a spur of sandstone 200 ft. above sea level. Excavation between the groups was insufficient to disprove presence of further buildings.

In the N. group (03128220) slight lines of stones and traces of post-holes suggested two rounded huts (D) and (E) perhaps of the mid 1st century A.D.; the western (D) of 16 ft. by 14 ft. was succeeded on the same site by a roughly rectangular hut (C), measuring 18 ft. by 14 ft., with a central hearth, paved threshold to N.E. and a clay 'shelf' 6 ins. high on W. A ditch to E., 3 ft. wide and 1½ ft. deep at N., increasing in size to S., was filled in when, according to the excavator, hut C was extended by the addition of sheds on all sides save the W. to measure 26 ft. by 23 ft. Finds on the 'shelf' included a samian platter of c. 65–80, a brooch, and a coin of Vespasian of A.D. 73; the date of abandonment seemed to be c. 85. Pieces of crucibles for copper-working were associated with huts (D) and (E).

N. of these remains were two rectangular buildings or 'cottages', nearly end to end on slightly different alignments, with unmortared sandstone and flint footings some 2 ft. to 3 ft. wide, probably for cob walls. Each building, terraced into the westerly hill-slope, measured 36½ ft. to 37 ft. by 21½ ft., with an entrance 4½ ft. to 5½ ft. wide in the S. wall, and, in the W. building (B), possibly a corresponding N. entrance, perhaps an addition. Gaps in the E. walls were interpreted as double entrances. Floors were largely of clay and earth. The E. building (A), perhaps with an eccentric line of posts bisecting the long axis, for which, however, the evidence is insecure, had a hearth W. of centre, clay platforms in N.W. and S.W. corners, and an internal gully with round sump E. of the S. entrance. Finds from (A) included a chalk loom-weight and shale spindle-whorls, also iron slag and parts of an iron cauldron from the E. part of the building. Five coins of 259–96 and coarse pottery indicated occupation in the 3rd and perhaps the first half of the 4th century. The W. building (B) may have been divided into aisles by two rows of four or five posts, for which again the evidence is slight, and, by revetted changes of floor-level from W. to E., into three rooms, identified by the excavator as a dwelling-room with hearth in S.W. corner, byre and workshop respectively. A mass of stones 10 ft. long and 6 ft. wide, revetting the W. or dwelling-room, had two concavities in the E. side identified as heads of ox-stalls between N. and S. doorways. Evidence for an internal door between dwelling-room and 'byre' and a very low window in the adjacent S. wall is insecure. Most finds came from the E. end where there were a knife, a spike, and some slag. A coin of Allectus (293–6) below the floor and coarse pottery above it suggested construction and occupation in the 4th century.

In the S. group (03128217) five hearths and slight scatters of stones were attributed to seven huts of which the most certain (J) was probably oval about 30 ft. long by 20 ft. and perhaps of c. A.D. 25–75. The finds of pottery and brooches from the other 'huts' (fn. 24) were of similar date, except at hut (I) probably of the 2nd to 3rd centuries. Two parallel lines of walling 34 ft. long and 44 ft. apart were attributed to a barn-like building (F), not earlier than c. 180 and perhaps of the 3rd century, but no firm traces of E. and W. walls were found. N. of (F), a line of flints and sandstone 24 ft. long and 1 ft. wide running N.W. to S.E., and a gully 2 ft. to 3 ft. wide and 1 ft. deep running at right angles to it, were ascribed to a rectangular building (G) with internal partition, preceding (F) and of later 1st century to early 3rd century.

More substantial remains (H) to E. consisted initially of a rectangular building, 26 ft. E.-W. by 16 ft., apparently of the late 3rd century, divided into two rooms of which the eastern was interpreted as a byre, with two ox-stalls as in (B) and with a 5 ft. wide door in its S. wall. An easterly extension marked, however, by little more than a N. wall 23 ft. long, was apparently added subsequently when the 'byre' of phase I was possibly adapted for living quarters or storage; the extension contained a substantial central clay hearth on a stone base 4½ ft. by 3 ft. and a walled recess to N. An irregular line of walling running N. and then N.W. for 19 ft. from the W. end of the extension was interpreted as a new byre of phase II. Pottery, mostly from a rubbish-pit in the S.E. corner of the original building, suggested occupation in the 4th century; iron spikes and nails, two turned shale armlet cores, a baked clay sling-stone with the impression of a grain of Spelt wheat, and animal bones, were also found. The nearest remains of 'Celtic' fields (Ancient Field Group (29)) are ½ m. to S. on Ballard Down. (Dorset Procs. LXXXVII (1965), 142–207.)

(47) St. Nicholas' Church (SZ 80 SW). Inhumation Burials have been found about 100 ft. above sea level in rebuilding the church and in digging graves in the churchyard extension to N. A number of burials in stone cists and a rotary quern were recorded by W. M. Hardy during renovations in 1881. Indications of others in the churchyard, not examined, were followed by the discovery in 1951 at a depth of some 3½ ft. of a covered cist of weathered Purbeck marble slabs aligned N.W.-S.E. (03638254), partly floored with limestones and containing remains of an extended skeleton probably of a female aged 30–40 with the severed head and detached mandible placed near the left foot at the N.W. end of the cist. The grave was not wholly cleared but a turned shale spindle-whorl was found in the pelvic region. Another cist of Purbeck marble was found a few yards to N. in 1955. (Dorset Procs. XII (1891), 177–8; LXXIV (1952), 51–4; LXXVII (1955), 126.)

Sturminster Marshall

The Roman road from Badbury to Dorchester, to be described in Dorset V, crosses the N.W. part of the parish and survives as an earthwork in Little Almer Wood. Other remains of doubtful location consist of a grave with hand-cut shale ring, in a railway cutting, and some eight pits exposed in a chalk-pit in the 'common-field', containing Romano-British pottery, bones, etc. (Dorset Procs. LXXXV (1963), 105; LXXXVI (1964), 115).

(47) Bailey Gate (SY 99 NW). Occupation Debris has been found W. of Church House in gravel pits over Upper Chalk on the 100 ft. ridge S. of the river Winterborne. Several pits containing Romano-British pottery were observed in 1936, presumably in the workings traceable on the brow of the hill (944992). Pottery found in 1935 presumably in the same workings is in Poole Museum. (Dorset Procs. LXXXV (1963), 105; Bournemouth Daily Echo, 11 Sept. 1937.)

Swanage

A coin probably of Antoninus Pius (138–61) comes from a field N. of Caldron Barn Farm, sz 024795 (Swanage Times, 7 Oct. 1953), and an issue of A.D. 332 of Constantine I from Swanage beach, 03117896 (Dorset Procs. LXXXIV (1962), 116).

(91) Ballard Down (SZ 08 SW). Inhumation Burials were found in 1917 on Upper Chalk about 350 ft. above sea level on the S. slope about 100 yds. from the cliffs (03978123). A skeleton in a cist of Purbeck stone slabs was found, with remains of another nearby in a plain grave. (Procs. Society of Antiquaries XXX (1917–18), 236–7.)

(92) Ulwell Farm (SZ 08 SW). Inhumation Burials were found in 1949 about 150 ft. above sea level on Lower Greensand about 150 yds. N.N.W. of the farmhouse (02268092). Fragmentary remains of three skeletons close together were in cists of limestone slabs; one pierced slab was probably a Roman roof-tile. (Dorset Procs. LXXI (1949), 68; LXXXVIII (1966), 120–1.)

Sites (93–94) lie on Wealden Beds.

(93) Godlingston Farm (SZ 07 NW). Occupation Debris comes from an E.-W. ridge about 50 ft. above sea level some 600 yds. S.E. of the farmhouse. Three nearly complete Durotrigian jars were found in 1961 in trenching for drains near the stream (01917979). Excavation by Mr. J. B. Calkin revealed sherds of similar type at a depth of 5 ft. to 6 ft., and 3rd or 4th-century coarse ware has since been collected after ploughing. An area of dark earth with sherds and sling-stones was seen on the crest 185 yds. W. (01747975), and drain-trenching revealed widespread sherds, limestone fragments and peat deposits. Vessels and sherds are at Godlingston Farm and Christchurch Museum. (Dorset Procs. LXXXVI (1964), 118–9.)

(94) Swanage Bay (SZ 07 SW). Inhumation Burials were exposed in 1877 in a cliff fall about ½ m. N. of Swanage (about 031798). Of two covered and partly floored cists of stone slabs, about 3 ft. apart at a depth of 18 ins., one about 2½ ft. long contained fragmentary bones of a child, the other, some 3 ft. long, probably the same. Large nails in both graves, one with oak (?) fibres, suggested wooden coffins perhaps of boards 2 ins. thick. One cist contained one or two sherds of black ware. The remains were near the 'North Ledge', presumably the Tanville Ledges offshore near Highcliffe Road, New Swanage. Objects, untraced, were formerly in Mowlem Institute, Swanage. (The Times, 23 Aug. 1877; C. E. Robinson, A Royal Warren: Picturesque Rambles in the Isle of Purbeck (1882), 101.)

Sites (95–98) lie on Middle Purbeck limestone.

(95) Herston (SZ 07 NW). Inhumation Burials, and Occupation Debris from Iron Age 'A' to late Roman times, have been found in quarrying the N. slopes about 250 ft. above sea level (about 021783).

The site, perhaps all now destroyed, yielded haematitecoated and other Iron Age 'A' sherds, Iron Age 'C' and Roman sherds, a few hand-cut shale armlet roughs and class A and C cores from turned armlets. Of two pits examined in 1931, one contained Iron Age 'C' pottery. Several extended burials in stone cists side by side are referable to the Roman period. Some pottery is in Poole Museum. A cist with adult inhumation was found in this area c. 1860. (Dorset Procs. LXX (1948), 42, 50; LXXV (1953), 69; Swanage Times, 7 Oct. 1953, 17 Feb. 1954; W. M. Hardy, Old Swanage (1908), 12.)

(96) Atlantic Road (SZ 07 NW). An Inhumation Burial was found in 1953 below the 200 ft. contour in building a house about 100 yds. S.S.W. of the Cottage Hospital (02807834).

The extended skeleton, evidently, from some 3 dozen 2½ in. nails, in a wooden coffin, lay with head S.E. at a depth of 4 ft. in an unusual cist consisting of two 1½ ft. high drystone walls nearly 2½ ft. apart and over 6 ft. long, capped with a double layer of local 'blue rag' slabs. The remains were of a male of about 50 years with exceptionally thick skull; the filling contained a brooch of the mid 2nd century, in private hands. The cranium is in Christchurch Museum. (Swanage Times, 30 Sept. 1953; Dorset Procs. LXXVI (1954), 76–7.) A circle of burials, undated, was reported 500 yds. to N. (ibid., LXXXVII (1965), 112–13).

(97) Bon Accord Road (SZ 07 NW). Occupation Debris was found c. 1904 above the 200 ft. contour in levelling the Durlston Court School playing-field (02907810). Sherds identified as Romano-British, flints, and coins, one certainly Roman, were found with indications of hearths. (W. M. Hardy, Old Swanage (1908), 11–13.)

(98) Durlston Cliff (SZ 07 NW). Inhumation Burials were found in 1904 some 200 ft. above sea level in building 'The View', now 'Durlston Cliff', in La Belle Vue Road a few yards from the cliff edge (03437813). Of eight skeletons found close together by the builder, W. M. Hardy, four were extended with feet to E. in plain graves; four were in covered stone cists, one of these, about 2½ ft. by 2 ft., containing two skeletons 'in a bent attitude' with feet to S. Two coarse ware sherds, one with handle, were apparently associated with cist graves. (W. M. Hardy, Old Swanage (1908), 11–12; Swanage Times, 7 Oct. 1953.)

Tyneham

Sites (39–40) occupy Wealden Beds.

(39) Tyneham Village (SY 88 SE). Occupation Debris and Inhumation Burials have been found near the head of Tyneham Gwyle about 150 ft. above sea level. Bones, samian ware and sherds thought Romano-British were found in 1860 in the diverted stream bed (88148031). A human skeleton was found in a nearby garden, and burials in cists of shale slabs were found in removing the S. porch of St. Mary's Church (88148039). Seven interments and a separately buried skull found in a sandy mound (87948030) near the Rectory, one with a vessel of 'black unbaked clay' near the head, were probably of Iron Age 'C' or Roman date. (Hutchins I, 628; locations from O.S. 6 in. map.) See also p. 454.

(40) Worbarrow Bay (SY 87 NE, 88 SE). Inhumation Burials, and Occupation Debris perhaps Iron Age to late Roman, with evidence for shale-working and perhaps Purbeck marble and salt industries, have been found on the cliffs N. of Tyneham Gwyle above the 100 ft. contour and near the Purbeck marble outcrop at Worbarrow Tout.

Remains, consisting of a layer of large stones at a depth of about 1½ ft. with darker soil above and many sherds, shale fragments and flint flakes, are visible in the cliff face for some 125 yds. (86938003 to 87017995) and may extend for some 30 yds. inland. Excavation in 1937 by Dr. W. H. C. Frend in an area over 20 yds. N.-S. (apparently about 87008000) revealed the layer of stones, including Purbeck marble from the beach, at a depth of 1½ ft. to 2 ft., in one place near the cliff edge acting as foundation for a floor of chalk blocks. Little was found on or above this layer, but black earth up to 1½ ft. thick below it contained charcoal, clay daub, much pottery, sling-stones, spindle-whorls of pot and shale, and a bronze plaque with repoussé decoration near a depression or pit in the cliff face. Other finds indicated shale-working on the site and perhaps also Purbeck marble and salt-boiling industries nearby. Evidence for the shale industry were pieces of shale, flint flakes with worn points, an armlet fragment and turned cores of classes A, B, C and D, and a deposit of 72 cores beneath a stone slab with an iron knife and arrow-head. Amongst stones used for the rough flooring were a quern-stone, a loom-weight of Purbeck marble, and part of an unfinished lugged mortar of Purbeck marble 'burr-stone'. A few pieces of salt-boiling 'briquetage' come from the site, and a coin of Commodus (180–92) from the cliffs.

The pottery including Durotrigian types is mainly of the 1st and 2nd centuries with some strays of Iron Age 'B' and perhaps 'A', but 2nd-century samian sherds were found in several places on or above natural subsoil. Common later types were rare, but an otherwise similar range of pottery collected by the excavator from the combe some 130 yds. E. (87118001) included conical flanged bowls of the 3rd or 4th century and a piece of salting 'briquetage'. Colour-coated pottery of this date came from the stream about 250 yds. S.E. (87207990).

Warne describes similar features in excavations in 1842, noting charred wheat in several places, mainly below the stone layer. The earliest accounts, by Miles and Pennie, substantiate but add few acceptable details except an inhumation in a covered stone cist. Another, all or partly of shale (D.C.M. Accessions Book, 20 Feb. 1908) was found in 1908. Objects are in D.C.M. (Dorset Procs. LXXI (1949), 51–2; LXXV (1953), 69; LXXXIX (1967), 146; W.H. C. Frend, 'Iron Age and Roman Sites in Purbeck' (MS. in D.C.M., 1936), 8–12; W. A. Miles, The Deverel Barrow (1826), 34–9; J. F. Pennie, Tale of a Modern Genius II (1827), 336–41; J. H. Austen, Purbeck Papers I (1856–7), 82; Warne, Ancient Dorset (1872), 327–9.)

Sites (41–42) lie on Kimmeridge Clay.

(41) Hobarrow Bay (SY 87 NE). An Inhumation Burial and Occupation Debris from a salt-boiling industry come from a mound at the cliff edge below the 100 ft. contour ½ m. S.S.W. of South Egliston (89547909), and about 300 yds. S.E. of the 'blackstone' outcrop of the Kimmeridge Shale at cliff-top in Brandy Bay.

The mound, on an E. slope and partly sectioned by the cliff, is now some 75 ft. in diameter and 8 ft. high, with indications that the deposits composing it fill a hollow of unknown depth. A roughly circular dished area of some 120 ft. diameter extending from its N. and W. base and yielding pottery and salting 'briquetage', may indicate its extent before exploitation for use as marl.

Excavation by Miles to a stated depth of 20 ft. to 30 ft. revealed pottery, 'briquetage', objects of iron, shale, bone and horn, animal bones, shells, clinker, and beds of clay, ashes and burnt shale to a depth of 20 ft. if not throughout. A covered stone cist at 5 ft. contained an inhumation; another at 12 ft., below human bones, had ashes and snail shells; a third of shale, slightly lower, contained shale ash. In the centre, at unspecified depth, a stone wall about 3 ft. thick, with 'divisional slabs' set off it, ran N.E. A lathe-turned shale armlet core of Calkin's class A occurred at 12 ft. and another apparently of class B at 20 ft.

There have been small excavations in the mound by the Rev. A. J. Watson, Dr. W. H. C. Frend and Mr. J. B. Calkin, and objects have been collected by Dr. Henrietta Davies and R.C.H.M. Frend found a deep layer of shale ash, clinker and 'briquetage' below 1½ ft. of shaly clay topsoil, and observed heaps of 'briquetage' alternating with layers of shale ash in the cliff section, the thickest of which was 10 ft.

The material from the above sources (in D.C.M. and London University Institute of Archaeology) consists mainly of 'briquetage', some pottery, a broken hand-cut shale armlet rough and several turned armlet cores of classes A and C. The clinker may derive from high combustion of shale, here clearly used as fuel in salt-boiling. The 'briquetage' of this industry consists of many roughly hand-moulded clay supports with flat or grooved ends, complete examples ranging from 3 ins. to 7½ ins. long; a few small irregular squeezed lumps; many fragments of thick straight-sided containers of pebbly clay with rounded corners; a few pieces of thinnerwalled vessels of doubtful shape, some with knife-cut rims or edges, and several bun-shaped pads. The pottery, mainly Durotrigian coarse ware, suggests occupation beginning in Iron Age 'C' or early Roman times but includes a few Iron Age 'A' and late Roman sherds. (W. A. Miles, The Deverel Barrow (1826), 51–3; Purbeck Papers I (1859–60), 217; Hutchins I, 627; W. H. C. Frend, 'Iron Age and Roman Sites in Purbeck' (MS. in D.C.M., 1936), 5–6; Dorset Procs. LXX (1948), 56, 58; LXXI (1949), 52–3; LXXV (1953), 69; LXXXIV (1962), 140–2.)

(42) Charnel (SY 87 NE). Occupation Debris, probably from a salt-boiling industry, comes from two mounds at the cliff edge about 50 ft. above sea level ½ m. S. of South Egliston.

A roughly circular mound some 12 ft. high and 30 yds. in diameter, with a lower extension some 15 yds. wide running N. for about 15 yds., stands at the cliff edge (89867893) some 50 yds. E. of the Egliston Gwyle stream. It seems to consist largely of unburnt shale, but construction of a concrete pillbox has thrown up burnt shale and several fragments of pebbly 'briquetage' (in D.C.M.). About 45 yds. N.E. (89897897) a mound some 5 ft. high and 19 yds. N.-S. by 15 yds. E.-W. seems to consist of unburnt shale. (Dorset Procs. LXXXIX (1967), 146.)

(43) Brandy Bay (SY 87 NE). Occupation Debris, some probably from salt-boiling, comes from two adjacent sites on the cliffs below Tyneham Cap, near the junction of Portland Sands with Kimmeridge Clay.

Dr. W. H. C. Frend's note of Iron Age 'C' and Romano-British pottery at a site (89397961) comparable with that at Hobarrow Bay (41) is confirmed by a few scraps of burnt shale and reddish flint-gritted ware of fabric resembling salting 'briquetage' found in 1958 about 70 yds. S.W. some 300 ft. above sea level (89347956). Romano-British sherds and one probably of Iron Age 'A' come from the cliff edge 220 yds. S.W. (89217941). Stone cobbles and Iron Age 'A' sherds noted by Frend in the cliff face between Tyneham Cap and Broad Bench are unlocated. Objects are in D.C.M. (W. H. C. Frend, 'Iron Age and Roman Sites in Purbeck' (MS. in D.C.M., 1936), 4, 6; Dorset Procs. LXXI (1949), 53; LXXXIX (1967), 146.)

Sites (44–45) are unlocated.

(44) Povington Farm (SY 88 SE). Occupation Debris from a lathe-turned shale armlet industry has been found on the farm N. of the Chalk range, presumably on London Clay or Bagshot Beds.

Following an earlier discovery of large quantities of waste cores from turned shale armlets in draining an irregular undulating field said to have foundations below surface, excavation by J. H. Austen in 1856 and 1859 produced from an area of some 2 sq. yds. of black earth about 700 cores of five varieties including Calkin's classes A, B, C and D, some with incompletely detached rings, and many broken armlets. Circular blanks, shale chips and flint tools were found, and a hand-cut shale dish or lamp-stand (in D.C.M.); 'black Roman pottery' occurred. Finds were made over an area of about ½ acre. (Purbeck Papers I (1856–60), 90–1, 221–3; Arch. J. XVI (1859), 299–300; Hutchins I, 560–2; Dorset Procs. LXXV (1953), 67, 69.)

(45) Egliston Mead (SY 88 SE). Occupation Debris from a lathe-turned shale armlet industry has been found in a field called Egliston Mead near Tyneham village, probably on Wealden Beds. In 1856 J. H. Austen examined part of a site thought to cover an acre; many turned armlet cores of Calkin's class A were found scattered in black earth, with a piece of an armlet and a few sherds. The site, about a mile inland from Worbarrow Bay, is possibly Chapel Close (89508050, 'Egglestone Little Mead' on the Tithe Map, 1840) or its neighbour 'Egglestone Great Mead' to W. (Purbeck Papers I (1856–7), 91; Dorset Procs. LXXV (1953), 69.)

Wareham Lady St. Mary

(85) Wareham (SY 98 NW). Occupation Debris, Iron Age 'A' to late Roman has been found at several points within or near the post-Roman town walls, notably in the N.W. and S.E. quarters. The site is Valley Gravel some 25 ft. above sea level between the rivers Frome and Piddle (Fig. p. 323).

Two Iron Age pits about 5¾ ft. in diameter and 5½ ft. deep were found in the old ground surface under the W. walls in excavations in 1952 (91978747; a on Fig. p. 323); they contained Iron Age 'A', 'B' and perhaps 'C' pottery and were overlaid with deposits including burnt material. Similar sherds and much Romano-British pottery, some of the late 4th century, were found in the rampart, derived from the old ground surface; Romano-British sherds were found in a like situation to S. in 1940 (92048731; b).

At 20 West Walls in 1953 H. G. Burr found traces of occupation probably of the 4th century, and below it part of a clay floor and two post-holes. Underneath this floor a pit contained early Romano-British sherds (92058748; c). Another pit or ditch was found opposite (92078749; d) with Iron Age 'C' or early Romano-British pottery and worked shale. A complete 4th-century jar was found c. 1936 at c, and in 1940 two similar vessels, a quern and a stone bowl or mortar, were found nearby at 9 Tinker's Lane (92028751; e). A group of small pots reported from the electricity transformer station adjoining these discoveries (92028743; f) may have been Roman or Saxon. Numerous Roman coins are reported from gardens inside the N.W. angle of the walls (g), whence came Romano-British sherds in 1947. Two coins of Constantius II (324–61) were found c. 1935 at the junction of Mill Lane and Cow Lane (92208746; h), and pottery, some certainly Romano-British, had been found in Mill Lane in 1898–9; sherds from here, including Durotrigian ware, are in D.C.M. Finds made in 1896 on the site of All Saints' Church (92288749; i) included samian and glass beads.

In the N.E. quarter the chief find was a group of three almost intact but worn samian bowls, in the Royal Albert Museum, Exeter, from the Folly (about 92258778; j); one was by the Antonine potter Mainacnus. Some 40 coins, probably a hoard, were found c. 1850 in lowering the hill at the N. end of North Street (k); Romano-British sherds have also been found in the garden of 45 East Street (92528746; l).

Most of the finds still preserved were made c. 1889–90 in the cemetery extension E. of St. Mary's Church (92588718; m). They include Claudian to Antonine samian, much coarse ware, and coins of Augustus (issue of 29–27 B.C.) and Constantine I (306–37). Large stones, a well, burnt human bones and charcoal were also found, not necessarily Roman. A coin of Constantine, of A.D. 330–5, comes from the Priory grounds (92508715; n), and at 21 South Street (92388727; o) were found four coins unusual in that, with a worn, probably Constantinian issue, were three Persian coins of the 2nd and 3rd centuries A.D. Finds from unspecified gardens near the E. rampart included 3rd and 4th-century coins and a lead statuette of a negro athlete, lost but of doubtful authenticity.

The only certain Roman finds from the S.W. quarter were a coin of Antoninus Pius (138–61) from West Street, sherds and a coin of Constantine I from S. of Castle Close (92178711; p), and sherds, possibly Roman, from gardens at West Port, perhaps extra-mural. Unprovenanced discoveries include a bronze patera handle, a Purbeck stone mortar, and part of an incised shale plaque.

These finds indicate occupation of relatively humble character, established in Iron Age 'A' and continuing throughout the Roman period, over much of the area within the later walls. The inscriptions in St. Mary's Church, cut on Roman architectural fragments of unknown derivation (see p. xlix), may imply continuity of occupation into post-Roman times. Objects are in St. Mary's Church, D.C.M., and Exeter Museum. (Dorset Procs. XX (1899), 148–60; LXXVI (1954), 82–5; Med. Arch. III (1959), 120–38.)

Wareham St. Martin

(17) Swineham (SY 98 NW). Occupation Debris comes from a field 600 yds. E. of Swineham Farm on Valley Gravel at the neck of an alluvial peninsula in Poole Harbour. Coarse pottery ranging from perhaps the 1st to the 4th century was found at least down to the water-table at a depth of about 3 ft., mostly in what may have been a shallow ditch cut into natural sand (94228799). Many sherds occur in upcast from dyke-scouring between 94238801 and 94208808. (Dorset Procs. LXXVII (1955), 127.)

Weymouth

A 4th-century coin hoard believed to come from the area is described in Num. Chron. 6th ser. IX (1949), 252–3. Several objects in the British Museum (Durden Collection), believed to come from Greenhill (SY 683802), include a Celtic bronze collar and tankard handle, and a Roman legionary belt-plate (Arch. J. cxv (1958), 95); there is, however, another Green Hill W. of Monument (449) below. Three Romano-British vessels come from an unspecified site or sites in the Dorchester road and a coin of Trajan (98–117) from 'Monksdene' therein (68038203; Dorset Procs. XLIV (1923), 40, 41). Shale objects and part of a quern, in B.M. (Durden Collection) and D.C.M., come from Ferry Bridge, Wyke Regis (666762), but there is no confirmation of a Roman building said to have been found there c. 1900 (Dorset Procs. XII (1891), 33; XXXI (1910), xxix).

Sites (437–438) lie on Cornbrash.

(437) Radipole, Spa Hill (SY 68 SE). Occupation Debris and Inhumation Burials have been found at about 100 ft. on the ridge at the N. end of Radipole Lake.

Part of a concrete floor was found in 1937 by the N.E. corner of 112 Spa Road (66938128), with a 1st-century brooch, sherds of samian and coarse ware, some of the 4th century, and three coins including an issue of 367–75 of Gratian. Sherds and an imitation as of Claudius I (41–54) come from gardens to E.N.E. (66978142), and a bronze key from Icen Road to E. Objects are in D.C.M. and private possession.

Over seven skeletons, some at random, three extended E.-W. and one N.-S., were found nearby by J. Medhurst c. 1844, near the road at the crest of the hill (about 66918127), accompanied by black ware urns and samian. Remains of a Roman 'causeway', pitched stones, sherds, ashes, bones and shells were found in an extensive area of made soil and, in adjoining fields, debris including Roman tile and coins. A shale plaque, samian dish and grey ware jar, said to have been found at Radipole in 1869, are in B.M. (Medhurst Collection). Further burials on the crest of Spa Hill, found with a samian dish and black ware jar, were destroyed in 1886.

A cylindrical amphora stamped P. towards the base, in D.C.M. and dating probably before c. A.D. 50 (M. H. Callender, Roman Amphorae (1965), 199–200), was found c. 1890 in mid-channel silt between the Backwater dam and the Gasworks tunnel (about 67687906). 'Roman ware, Roman bricks, ridge and flat tiles, remnants of Kimmeridge Coal urns, coins, pottery, etc.' had been found earlier in the Backwater near the Gasworks (Notes by E. Jackson Brown, 1835–45, extracted by V. L. Oliver, in D.C.M.), and a flanged coarse ware bowl of the 3rd or 4th century, in D.C.M., came from the N. end of Radipole Lake, about 670805 (A. D. Passmore, letter 24 Apr. 1934, in D.C.M.). Two amphorae were recovered from Weymouth Bay (1880, in D.C.M.) and Portland Harbour (1914–18, D.C.M. correspondence, 1920).

The evidence is insufficient to determine the site of the port installations served by the Roman road from Dorchester (p. 528), although the remains at Spa Hill lie on its projected alignment. Some remains of the road were claimed by Ellis at the foot of the hill. (Gentleman's Magazine (1844), pt. ii, 636; (1845), pt. i, 79; Dorset Procs. XI (1890), 88–90; XLIV (1923), 41–3, 54–5; LXXIII (1951), 96–9; LXXVI (1954), 97–8; Warne, Ancient Dorset (1872), 187.)

(438) Radipole Lane (SY 68 SE). An Inhumation Burial and Occupation Debris have been found about 100 ft. above sea level in the South Hill Estate at the E. end of the Chickerell ridge. A contracted inhumation probably not earlier than the 3rd or 4th century A.D. was found in 1935 near the hill-crest (66338100), lying E.-W. with head W. in a scoop at the base of a shallow pit. A stone cist, apparently empty, was found earlier about 100 ft. to N. near a coarse ware jar probably of the 2nd century. Romano-British sherds were seen in builders' excavations here, and in the field to E., on the crest and S. slope. Objects are in D.C.M. (Dorset Procs. LXXIII (1951), 95–6.)

Sites (439–445) lie on Corallian Beds.

(439) Newberry Road, Weymouth (SY 67 NE). A Villa is probably indicated by a mosaic or mosaics found in 1835 and 1902 on the lower S.-facing slope of a re-entrant in Nothe Clay about 25 ft. above the sea which lies 200 yds. to E.

The corner of a mosaic floor removed to D.C.M. in 1902 was found E. of the road, about 67987839, and was probably distinct from a 'Roman tessellated pavement' found nearby in 1835 perhaps under Spring Terrace, 67977837. The surviving pattern of five colours, 10½ ft. by 5 ft., is unusual. A coarse border of two red lines on a white ground encloses a finer triple border consisting of black and white chevrons, guilloche in black, white, brown and red, and chequers in white, red, grey and brown. Remains of a centre panel consist of a white ground divided apparently into triangular spaces by three oblique lines of guilloche. A small grey-edged square at the base of the best preserved triangle was of variously coloured tesserae set lozenge-wise, of which only the white are present. The adjoining triangle, in which the white ground tesserae are set lozenge-wise, may have contained a curvilinear pattern. The only find recorded was a Saxon sceatta allegedly on the floor. (Dorset Procs. XLIV (1923), 39–40; LXXXV (1963), 96–8.)

(440) Clearmount, Rodwell (SY 67 NE). Inhumation Burials were found c. 1864 in the railway cutting about 100 ft. above sea level.

A 'Roman burial place' was cut through at Rodwell in constructing the Weymouth and Portland Railway, opened in 1865. Two vessels in B.M., not closely dated, were found 'in a railway cutting at Clearmount', presumably E. of Clearmount Road (about 675779), while Lock's Guide states that 'during the excavation for Rodwell station' (674783) 'pieces of Roman pottery and some Roman coins were unearthed, as well as the bones of a man, horse and dog.' A Roman bronze spoon came from the garden of 'Faircross' (669783). (Dorset Procs. XLIV (1923), 42–3; Procs. Society of Antiquaries III (1865), 93; Pictorial and Descriptive Guide to Weymouth (Ward, Lock & Co., 10th edn.), 53.)

(441) Wyke Regis Reservoir (SY 67 NE). Inhumation Burials have been found above the 200 ft. contour E. of All Saints' Church.

Of two burials, each with a pot, found in 1858 in the glebe allotments about 30 yds. E. of the graveyard (about 66307784), one was contracted E.-W., head E., in a 4 ft. long cist with single stone slabs for sides and a shale(?) cover, and had a Durotrigian bowl of Brailsford's class I by the shoulder; there were remains probably of two additional stone cists. A beadrimmed jar containing a samian cup (form 27) of c. 60–80 and a coin, untraced, allegedly of Faustina I (c. 139–45), came from construction c. 1900 of the older part of the reservoir, about 66367788. In 1936, following discovery in the adjoining new reservoir of three or four burials, one at least in a stone cist, a skeleton in a shallow grave containing six or seven iron nails was excavated some 20 yds. from the roadside (66417791); it was extended E.-W., head E., on its right side with knees slightly flexed. A handled beaker of black coarse ware of the later 3rd or 4th century was recovered in fragments from the grave. Two Roman coins and several sherds are recorded from the allotment gardens to W. Objects are in D.C.M. (Arch. J. CXVI (1959), 201–2; J.B.A.A. XV (1859), 283; Dorset Procs. XXXI (1910), xxix; LXXI (1949), 70–2; Dorset Daily Echo 28 Oct. 1936.)

(442) Overlands, Wyke Regis (SY 67 NE). An Inhumation Burial and Occupation Debris were found in 1937 in drainage trenches on the housing estate near Camp Road, about ¼ m. W. of Monument (441) and some 200 ft. above sea level.

A cist 4 ft. by 2 ft., with single stone slabs for sides and cover, contained a contracted adult skeleton on its left side, head W., with a coarse ware jar probably of the 3rd century in front of the face and a few sherds and winkle shells in the filling. The site, 'Overlands' in the local newspaper and North Road in D.C.M. records, was probably on the knoll where the latter joins Overlands Road, about 65837792. A Durotrigian bowl of Brailsford's class 1a and a small handled cup of black ware, both whole, probably indicate other burials. A 'hearth', limpet and winkle shells, including a thick deposit of the latter, suggest domestic occupation, and many Roman coins were said to have been found in ploughing. Vessels and bones are in D.C.M. (D.C.M. correspondence, 17 July, 20 Aug. 1937; Dorset Daily Echo 3 July 1937.)

Roman remains including bricks and burnt stones are said to have been found at Barrow (658776) to S. (E. Jackson Brown, MS. notes 1835–45, extracted by V. L. Oliver, in D.C.M.).

(443) Broadwey (SY 68 SE). An Inhumation Burial was found beside the Weymouth road c. 1844. The remains, apparently of a young person, were found 'in the Roman grave by the side of the Roman road, four miles south of Dorchester' (F. T. Buckland, Curiosities of Natural History 2nd ser. (1860), 112). 'Portions of a Roman cinerary urn' were found at St. Nicholas' Church in 1862 in digging Dr. Puckett's grave (66808354; Hutchins II, 490), and a Durotrigian bowl, in D.C.M., comes from the 'lower end of the railway cutting'. (Dorset Procs. XLIV (1923), 43.)

(444) South Down Farm, Lodmoor (SY 68 SE). An Inhumation Burial and Occupation Debris were found in 1934 in making a road E. of the farmhouse about 100 ft. above sea level (about 68738189). A skeleton, found at a depth of about 7 ft. by the builder, Mr. A. C. Whettam, but removed before examination, was believed by V. L. Oliver and C. D. Drew to have been buried at the bottom of a shallow domestic pit containing burnt material, oyster shells, sherds and iron nails. (V. L. Oliver, MS. notes in D.C.M., 1934; information from Mr. A. C. Whettam.)

(445) Jordan Hill, Preston (SY 68 SE). A Building (Temple?) and Cemetery have been found on the hillcrest about 150 ft. above sea level overlooking Weymouth Bay. The building was excavated by J. Medhurst in 1843, and by C. S. Prideaux and C. D. Drew in 1931–2; the exposed foundations (69898207) are in the care of the Ministry of Public Building and Works. The cemetery to N. and N.E. was investigated by Medhurst in 1845–6.

The Building was some 25 ft. square internally with walls 3 ft. 8 ins. wide on foundations 9½ ft. wide. Both stone and clay roofing tiles were found. A thin concrete surround on all sides was at least 11 ft. wide on S. and 9 ft. on W. Observers in 1843, perhaps misled by two offsets in the foundations, noted traces of steps at the centre of the S. wall, and bases of four small columns in situ 4 ft. from it; (fn. 25) a limestone base and a Purbeck marble Tuscan capital (Fig.) for a column or columns about 5 ft. high, in D.C.M., were found loose near the N. wall. An oblong shaft 4 ft. by 3 ft. and 12 ft. deep lay within the S.E. angle, largely under the line of the wall footings; the sides were lined with roofing slabs set in clay, and the filling consisted of 16 layers of ash and charcoal alternating with double layers of roofing slabs arranged in superimposed pairs, between each of which were bones of a single bird and a small bronze coin. At the bottom a stone cist contained two urns, a sword, spear-head, knife, steelyard, bucket-handle, crook and other iron objects; a similar cist half-way down enclosed urns, a sword and a spear-head. The bird bones were identified as of buzzard, raven, starling and crow, and one coin as of Theodosius I (379–95). There were also bones of a hare.

The building was set, apparently centrally, in an outer square enclosure, not confirmed in 1931–2, with walls 5 ft. thick and dimensions variously given as 280 ft. square, 110 ft. or 37 yds. square. Within this were found many animal remains, especially horns of oxen, pottery, and about 300 coins all apparently of the late Empire including one of Arcadius (383–408). In later excavations and consolidation 177 coins were found on the site, 61 of them certainly of the period 388–95, amongst them a bronze Durotrigian coin and examples from Vespasian (69–79) to Arcadius. Of two coin hoards discovered on the hill, one in 1812 consisted of several hundred silver coins, mostly it seems of 244–68; the other (69808201), found in 1928, was of over 4,400 bronze coins ranging from Postumus (259–68) to Honorius (393–423) but mostly issued after 379.

The Cemetery, perhaps pre-Roman in origin, lay N. and N.E. and extended at least 300 yds. over an area where Roman material occurs on the surface. The most reliable account is probably that of T. W. Wake Smart in Warne's Ancient Dorset. About 80 inhumation burials of adults and children were found in an area of about one acre, variously orientated and often flexed, and sometimes in groups of up to six individuals. Some were in stone cists and one grave was paved with chalk tesserae; nails indicated that some had been in wooden coffins. Low drystone walls, one of crescent-shape 21 ft. long, apparently demarcated burial plots and sometimes had burials in their structure (Shipp (in Hutchins) states that the cemetery was within a parallelogram 500 ft. across with a low thick wall). Near the burials were (i) several floors of white clay, one seemingly of 18 ft. by 12 ft. with stone walls; (ii) clay-lined hollows containing ashes, animal bones and sherds, several apparently provided with stone-lined drains; (iii) two stone cists containing burnt shale and calcined animal bones; (iv) several stone piles on which rested animal bones or vessels containing them. Some burials had single pots or groups; in one group of nine vessels, three (a samian dish, a black ware imitation of samian form 37, and a handled cup of black ware) stood on an engraved oblong plaque or tray of shale placed at the shoulder, with five black ware bowls ranged around it, and a bottle of yellowish ware at the knees. Another imitation of samian form 37 was made of shale. Some 80 vessels survive (mostly in D.C.M. and B.M.) out of perhaps 125 listed as from Jordan Hill in sale catalogues of the Medhurst Collection (Sotheby's, 1 July 1879; C. T. Jefferies, Bristol, 1893), including many Durotrigian jars, bowls and handled mugs, some samian ware, imitation samian in black fabrics, a Gallo-Belgic terra rubra bowl and terra nigra platters, and two lead-glazed beakers. Although sherds of late Romano-British ware were found, most of the whole vessels may be attributed to the cemetery and assigned to the second half of the 1st century A.D., with some few of the 2nd century or later. A bronze armlet, fingerrings and sandal-nails accompanied burials, also iron arrow-heads, an iron sword, styli, and bone weaving-combs.

Other objects from the cemetery area, not necessarily with burials and in some cases suggesting domestic or industrial occupation, included iron spear-heads, saddle and rotary quern-stones, chert and flint balls, a shale armlet and latheturned armlet cores, a Durotrigian silver coin and three of the 3rd and 4th centuries, sherds including painted ware of New Forest type, and many pieces of angular supports resembling salt-boiling 'briquetage'. Less precisely attributed objects from Jordan Hill are four more whole or partial shale plaques or trays with engraved decoration, a carved shale slab, a shale tablet with a stylized lion in relief from a site yielding bronze coins and sherds, Iron Age swan-necked and ring-headed pins (Arch. J. XCI (1934), 288–9), La Tène I and III and Roman brooches, and a bronze Roman mirror-handle (Archaeologia Cambrensis C (1949), 32).

The date and purpose of the Jordan Hill building are uncertain. The coins and pottery point to intensive occupation in the 4th and early 5th century but do not exclude an earlier date of construction; although it appears likely that the Theodosian shaft predated the building, or was a foundation deposit, their relationship cannot now be established. The building has usually been regarded as the cella of a temple perhaps with portico, surrounded by a temenos. The columns, and the shaft which seems clearly to be of ritual character as at Ashill (V.C.H. Norfolk I (1901), 295–6), would support this ascription, although use as a signal station has been suggested (Dorset Procs. LVII (1935), 140). Objects from the site are in B.M., D.C.M., Pitt-Rivers Museum (Oxford), Taunton, Bristol, and Warrington Museums. (Gentleman's Magazine (1844), pt. i, 185–6; (1844), pt. ii, 413, 635–6; H. C. Harford, Essay on Roman Remains near Weymouth, read to Soc. Antiquaries, 21 Mar. 1844 (privately printed); Abstracts of Procs. Ashmolean Society II (1843–55), 55; Hutchins I, 562; II, 838–9; Warne, Ancient Dorset (1872), 224–34, Appendix 297–8; William Shipp MS. II (in D.C.M.), 59–66; Warne, MS. 'Illustrations to the History of Dorset' (in D.C.M.), 179, 248, 256, 272, 392; Dorset Procs. LIII (1931), 265–76; LIV (1932), 15–21. Coins: G. A. Ellis, History and Antiquities of Weymouth and Melcombe Regis (1829), 3, 253–4; Warne, C.T.D., mopr, 60; Dorset Procs. LI (1929), 158–82; LVII (1935), 140–2; Num. Chron. XI (1931), 14–27. Other objects: Purbeck Papers I (1859–60), 228–9; Dorset Procs. XLIV (1923), 48–53; Catalogue of Roman Pottery in the B.M. (1908), 10, 127, 418, 420; Guide to the Antiquities of Roman Britain (B.M., 1922), 70–1, 114–5; (1951), 29–30, 36, 56); Later Prehistoric Antiquities of the British Isles (B.M., 1953), 56, 72.)

(446) Bowleaze Cove, Preston (SY 78 SW). Occupation Debris, Floors, and an Inhumation Burial have been found near the mouth of the River Jordan on Early Holocene Loam over Oxford Clay and in landslips on Furzy Cliff to the W. on Nothe Grit over Oxford Clay.

An occupation layer probably of the 4th century A.D. some 6 ins. to 8 ins. thick, containing animal bones, Romano-British tile fragments and sherds ranging from at least the 2nd century onwards, is sectioned some 9 ins. below surface in a low scarp above foreshore W. of the stream; it extends for some 60 yds. from 70198195 to 70248194, and perhaps eastwards towards the stream (J. N. Carreck). A floor of fine pebbles on a double course of large beach stones, and fragments of amphorae, were observed in 1935, and in 1945–9 traces of ditches running seawards. Romano-British sherds come from the stream 100 yds. N.E. (70328201). Piles of a supposed Roman landing stage are said to have been found by J. Medhurst c. 1845.

About 1 ft. below surface in the upper face of Furzy Cliff to W.N.W., some 60 ft. above sea level (70148196), a roughly mortared 'floor' some 9 ins. thick, of oolite, chert and shingle, extends in section for about 22 ft. between rough settings or curbs of heavier stones. The feature slopes gently eastwards in conformity with ground surface; a thin stony layer continues its line to E. for about 24 ft., the last 12 ft. or so overlying a mass of 'cannon-ball' concretions which, though occurring naturally in bands in the Nothe Grit, appear here to have been deliberately pitched. Romano-British sherds and tile fragments occur above and immediately below the 'floor'; a Durotrigian silver coin of late 1st century B.C. was also found. Objects are in D.C.M. (Dorset Procs. LXX (1948), 63–4; LXXIII (1951), 97; LXXXIV (1962), 112; J. N. Carreck, Procs. Geologists' Association LXVI (1955), 74–100; K. C. C. Selby, D.C.M. Correspondence, No. 609, 17 May 1935.)

A grave found 'in the cliff near Jordan gate' between 1835 and 1845, probably in the landslips at Furzy Cliff, contained a skeleton with a 'hollow gold' finger-ring with glass intaglio, a model bronze battle-axe, and a silver coin of Ptolemy of c. 285 B.C. The coin, now lost, was evidently of Ptolemy I or II, but the reset intaglio is Graeco-Roman of the 1st century B.C. or later. Objects are in private possession. (E. Jackson Brown, MS. notes, 1835–45, extracted by V. L. Oliver, in D.C.M.)

Sites (447–448) lie on Kimmeridge Clay.

(447) Preston (SY 78 SW). A Villa discovered by J. Medhurst in 1844 below the 50 ft. contour on the W. slope of the Jordan valley, ½ m. N.N.E. of Monument (445), was partly explored by him and T. Baker following diversion of the stream through the site in 1852, and again by Baker in 1871, and by C. D. Drew in 1932. The principal remains, a mosaic scheduled as an Ancient Monument, were concealed beneath a concrete floor in 1946 following extensive dilapidation (70298270).

The mosaic (Plate 225), occupying a room 21 ft. square, appears to have been at the N.W. angle of two ranges running due E. and S. respectively. A border of coarse white tesserae with red band, and an inner border of finer white tesserae with black band, enclosed an oblong guilloche-framed panel in black, white and red, of perhaps 18 ft. by 12 ft. apparently with long axis E.-W. This panel consisted of two end-panels with running leaf scrolls flanking a square panel cut off at the corners to form an octagon containing a rosette in a central square; the spandrels were filled with frets and triangles. The design was apparently symmetrical except perhaps for maladjustment of central square to octagon shown in the D.C.M. version of Crickmay's drawing on which the illustration is based. The materials were said to be Kimmeridge shale, Purbeck stone, and tile.

The floor, some parts of which including a corner were damaged, was not completely exposed until 1871, when some of the surrounding area was examined. There were no remains to N., but Baker found 'a room at the south-west with very rough tesserae, the court paved with stone at the centre, and a room to the south-east about twelve feet square, also roughly paved with tesserae'. His account refers to a wall 63 ft. 8 ins. long, and perhaps to three other rooms, one 12 ft. by 11½ ft. and two of unascertained dimensions. Small tiles 8 ins. square of chalk and shale (some of the latter in D.C.M.) were said to come from the villa.

Drew disclosed S. and E. wall foundations of the mosaic chamber partly under those of the protective building erected by Baker, and two parallel walls running E. in alignment with this chamber, forming a N. range at least 98 ft. long until cut by the stream. Of Drew's parallel walls, the N., consisting of a single diagonally pitched limestone course, lay throughout beneath the modern yard wall and was probably Baker's long wall; the S. wall, 2 ft. 4 ins. wide at base, was of two courses pitched in clay with a single mortared course set back above. The area between had a well-worn floor of limestone slabs of Roman date, in part over-riding and hence later than the wall foundations, occasional patches of ash where paving was missing, and at the E. end (70318270) a stokehole and rectangular oven or furnace, 2½ ft. by 1¾ ft., apparently contemporary with the floor. A table-leg of Kimmeridge shale (further discussed in Antiquity XXIV (1950), 25–8) was found below a floor-slab. Foundations and loose red and white tesserae were found in a test pit across the stream to S. (70328269), and three test pits 10 yds. to 20 yds. S.W. of the mosaic suggested a W. range running N.-S. Finds included painted wall-plaster, clay and stone roof-tiles, coins of Carausius (286–93) and Constans (issue of 335–41), and pottery mainly of the 3rd or 4th century including ware of New Forest type, in D.C.M. Part of an antefix of red clay (in D.C.M.), similar to those from Dorchester but not from the same mould, may come from the site (see p. 538).

Medhurst is believed to have found traces of structures elsewhere in the meadows between Jordan Hill and the village of Preston, and Romano-British pottery is said to have been found in the churchyard (70578296). (Gentleman's Magazine (1844), pt. i, 186–7; (1844), pt. ii, 636; Warne, Ancient Dorset (1872), 224, 235, Appendix, 297; T. Baker, J.B.A.A. XXVIII (1872), 94–6, plate facing 292; D.C.C., 24 Aug. 1872, 13; Dorset Procs. X (1889), xxviii-xxx; XXI (1900), 205–6; XLIV (1923), 49–50; LIV (1932), 21–34.)

(448) Sutton Poyntz, Waterworks (SY 78 SW). Inhumation Burials and Occupation Debris were found E. of the Waterworks in 1939 in building a house 'Watermeadow', on a S.-facing slope below the 100 ft. contour. Two burials lying E.-W. on right side with legs flexed, were found at 70678393 in a drain-trench running S. from the S.W. corner of the garage forming the W. end of the house. One, 34 ft. from the corner with head E., had a bead-rimmed bowl of Brailsford's class 2 near the head, with sherds and shale fragments and two bronze brooches, one a plate brooch of the second half of the 1st century A.D., in the grave filling. The other, at 30 ft., with head W., had a bead-rimmed vessel by the head and hob-nails at the feet; the grave filling contained sherds and a bronze needle. A third burial, unexcavated, was found in the inspection pit at the S.W. corner of the garage. Romano-British sherds, some of the 3rd or 4th century, come from the garden N. of the house. A bronze brooch, a coin of Antoninus Pius (138–61) and 4th-century sherds came from the house foundations. Objects and sketch plan are in D.C.M.

A handled jug was found before 1898 at the Waterworks, and a black ware tazza at the Rimbury reservoir ½ m. S.W. (699833; Dorset Procs. XLIV (1923), 53).

(449) Sutton Poyntz, Plaisters Lane (SY 78 SW). Occupation Debris, a Hut, and Inhumation Burials were found in 1931–6 in the garden of 'Wyndings' at the junction of Upper Greensand and Kimmeridge Clay on a S.-facing slope below the 200 ft. contour.

Rough limestone paving was found in excavations N. of the house by J. F. C. Kimber in 1935–6, over an area at least 60 ft. E.-W. by 40 ft. (70198424), with traces at E. of an oblong hut some 17 ft. by 15 ft. overall with drystone wall, enclosing an approximately central stone-lined oven 2 ft. wide internally. Limestone slabs suggesting a structure had been found previously under an extension on the N. side of the house, and a pit, a rotary quern-stone and pottery including a complete native imitation of a samian form 30 bowl (in D.C.M.) on the site of the tennis court to N.W. Finds from the site range from 1st to 4th century, including samian and New Forest type ware, several complete Durotrigian vessels and later cooking-pots, fragments of two shale plaques or trays (one engraved) and of three lugged mortars of Purbeck marble. In addition to two human skeletons (one of a child), found earlier in the garden, a contracted inhumation was excavated in 1936 W. of the house (70178423); it lay E.-W. with head E. beside which was a pedestalled Durotrigian bowl of Brailsford's class 2. (Finds and unpublished notes in possession of Mrs. J. F. C. Kimber; J.R.S. XXI (1931), 241; XXVII (1938), 194; Dorset Procs. LXXXVI (1964), 114–5.)

Whitcombe

(26) Whitcombe Hill (SY 78 NW). Building Remains, Pits, Ditches, Inhumation Burials and a stone Relief (Plate 228) have been found E. of Cole Hill Wood near the crest of a ridge of Upper Chalk 250 ft. above sea level.

Surface debris seen after the relief was ploughed up in 1963 consisted of pottery sherds ranging from the 1st or 2nd to the 4th century A.D., clay roof-tiles and large worn paving slabs of limestone. Excavation in 1965 revealed (i) remains of a rectangular building (71138807) approximately 27 ft. by 22 ft. on a north-easterly alignment, with a floor of chalk and earth and robbed walls with traces of internal painted wall-plaster in situ; (ii) some 25 yds. to W. (71118807) flint foundations 1¾ ft. wide of a wall running on a different north-easterly alignment, flanked on both sides by remains of a chalk and earth floor with three infant burials; (iii) 45 yds. S.W. of (i), three contracted inhumations, one male and two female, buried on the right side in hollows in the chalk (71008803). These last were variously aligned N.W.-S.E. (head), S.W.-N.E. (head), and E.-W. (head); the man was buried with two Durotrigian vessels, possibly pre-Roman, the women with joints of pig, sheep and horse, and in one case an amber-coloured glass ring bead.

Further excavation in 1966–7 indicated a native Durotrigian settlement of pre-Roman origin with chalk-cut storage pits and ditches, some of which had been made up with chalk at the sides where they had cut through older pit-fillings. Of nine more inhumations, mostly in shallow scooped graves, at least seven were contracted on the right side with heads in easterly directions. One adult male wore an iron bracelet; two others had a pair of Durotrigian pots, while two seem to have been unaccompanied except, in one case, by bones indicating food offerings such as occurred in several of the graves. A girl aged about 16, one of two burials in rectangular graves, had a necklace of ten glass, one paste and two wooden beads, and two Durotrigian and two samian vessels (forms Déch. 67 and Ritt. 12) dating the interment to c. A.D. 90 at earliest. A male of about 27, with indications perhaps of a wound in the right shoulder, was unique in the region in being buried with iron weapons as well as ornaments. These had consisted of a sword 2½ ft. long, lying by the right side in a bronze-mounted wooden scabbard probably attached by two rings to a belt or baldric, a spear perhaps grasped in the right hand, and a hammer-like weapon with chalk pommel in the left. A bronze brooch of La Tène II type lay above the right shoulder and a bronze ring with stud on the breast probably to fasten clothing or a baldric. This burial must be presumed prior to the Roman conquest.

The relief, apparently found near wall (ii), is of local Portland limestone, 8 ins. thick and 2 ft. 3½ ins. wide with a 3 in. plain border, and is now 1 ft. 11 ins. high having lost top and bottom portions. It depicts a bearded horseman with bare head turned to face the onlooker, wearing a flowing cloak curving in folds behind him over a belted tunic. The rider carries in his right hand a thick lance held at a forward slant and bears a round shield on his crooked left arm. A baldric, presumably to support a sword (not shown) at the left side, passes over a pad on the right shoulder. The horse, a stallion with short mane, drooping tail and prominent eye and nostril, walks to the right; it is harnessed with a neck rein, breast and crupper straps with two pendant straps or tassels, and has a rectangular saddle cloth and a saddle with raised front or saddle tree curving over the rider's thigh.

The sculpture, perhaps of the later 2nd or 3rd century, may be either a tombstone of an auxiliary cavalryman or a dedication possibly to a Thraco-Danubian hero-god regarded as a protector of hunters and a saviour in the struggle against forces of evil. Its probable date, and the lack of other evidence for a military site, favours the second interpretation. Coins suggest occupation of the site into the late 4th century at least. The relief and other finds are in D.C.M. (J.R.S. LIV (1964), 172; Dorset Procs. LXXXVI (1964), 103–4; LXXXVII (1965), 96; LXXXVIII (1966), 113; LXXXIX (1967), 126–7; cf. E. Will, Le Relief Cultuel Greco-Romain (1955), 56 ff.; G. I. Kazarow, Die Denkmäler des Thrakischen Reitergottes in Bulgarien, 2 vols. (1938); D. Tudor, 'I cavalieri Danubiani', Ephemeris Dacoromana VII (1937), 189–356.)

An inhumation burial, apparently contracted but of unknown date, was found at a depth of 3 ft. about ½ m. to N.E. in making the loop road round Whitcombe Hill (D.C.C., 19 May 1836).

Winfrith Newburgh

Two bronze penannular brooches, one of Iron Age and one of Romano-British type, and other objects, in D.C.M., come from an unlocated site in Sleight Bottom, N. of Randall's Farm (SY 805819).

Winterborne Monkton

An inhumation burial in a stone cist probably of this period was found in 1862 near the top of the large Bronze Age bell barrow N. of Maiden Castle (SY 66698928; see Round Barrows, Winterborne Monkton (6)). A coin of Constans (333–50) was found near the Roman road (680880; note in Haverfield Library, Ashmolean Museum).

Winterborne St. Martin

For Roman occupation at Maiden Castle, including a late 4th-century temple, see Hill-forts, Winterborne St. Martin (142). Two intrusive inhumation burials in stone cists probably of Roman date were found in Clandon Barrow, 65648900, and about ten without cists in a barrow to S.E., 65748894 (see Round Barrows, Winterborne St. Martin (134–5)).

(146) Manor Farm (SY 68 NW). An Inhumation Burial was found in 1950 in ploughing Middle Eweleaze below the crest of a spur of Upper Chalk above the 400 ft. contour (64658724). The burial lay in a stone cist about 7 ft. long E.-W. Some 20 round flint nodules, a sherd of Iron Age 'C' and two of mediaeval type were found on the site, and some Roman and mediaeval sherds occur in ploughsoil 100 yds. N. where traces of scarps may belong to destroyed 'Celtic' fields peripheral to Ancient Field Group (6). (Dorset Procs. LXXII (1950), 90.)

Woodsford

Two Romano-British shale spindle-whorls and a flint tool come from the area of the mediaeval settlement, Woodsford (4), SY 760905 (Warne, Ancient Dorset (1872), 19).

Wool

A bronze coin of the 2nd century A.D. (in D.C.M.) comes from a garden in Wool village beside the Lulworth road (SY 846862).

(50) Burton Cross (SY 88 NW). Occupation Debris was found in 1959 in laying an effluent pipe-line for the Atomic Energy Authority on a gentle S.-facing slope on Reading Beds about 200 ft. above sea level.

A layer of large flints 20 ft. across, and traces probably of horizontal beams about 6 ins. wide, lay about 30 yds. apart on natural sandy subsoil (83628562). An extensive layer of dark soil above contained pottery ranging from late 1st to 4th century, two 2nd-century coins, clay daub and cement-like material, possibly derived from a main building or buildings to N.E. where surface finds include fragments of clay roof-tile. A crop-mark adjoining the latter apparently indicates a rectangular ditched enclosure some 100 ft. by over 120 ft. long cut by the Winfrith Newburgh parish boundary (83728577). Worked flints and sherds, one probably of Iron Age 'A', were also found in the pipe-line trench to W. (83378578). (Dorset Procs. LXXXIV (1962), 125–30; see also Coombe Keynes (21) above.)

(51) Bovington Camp (SY 88 NW). A Durotrigian pot, probably from an Inhumation Burial, was found in 1962 at about 150 ft. on the crest of the low ridge of Bagshot Beds occupied by the permanent military camp (83218911). The vessel, a bowl of Brailsford's Durotrigian class 2 (in D.C.M.), was found in pieces about 3 ft. below the surface in laying electric cable to the junior ranks' club. (Dorset Procs. LXXXVII (1965), 110–11.)

Worth Matravers

Romano-British sherds have been reported from an unconfirmed site in a quarry on the Upper Purbeck marble of Primrose Hill about SY 98827948 (Dorset Procs. LXXXV (1952), 93). Burials in stone cists are said to have been found in the churchyard, 97287744 (Purbeck Papers I (1855), 34). A recently discovered site at Dunshay (98347974) near the Purbeck marble outcrop is described in Dorset Procs. LXXXVIII (1966), 120.

Sites (39–41) lie on the centre of a ridge of Middle Purbeck limestone about 450 ft. above sea level. The first two, wholly or largely destroyed, are about ½ m. W. of surviving 'Celtic' fields (Ancient Field Group (28)).

(39) Gallows Gore, West (SY 97 NE). Inhumation Burials, Pits, Hut-floors, and Occupation Debris, Iron Age 'A' to late Roman, with waste from shale armlet manufacture, have been found since 1931 over an area of 220 yds. by 100 yds. in Lander's quarry S. of Downshay Wood (97827901).

A child's skeleton perhaps in a wooden coffin, with objects including a brooch of c. 50–80, was enclosed in a covered cist of two slabs in the N.E. sector of the site. An unlined grave of two females extended E.-W. with heads at opposite ends was found 12 yds. to S., one elderly with three Durotrigian bowls of Brailsford's class 1 by the body and a native copy of a Gallo-Belgic butt-beaker of perhaps the mid 1st century A.D. in fragments over the head; nearby were two inhumations, one extended in a cist E.-W. Of two more to W. in plain graves, one had legs flexed.

Of 14 pits examined, six were of Iron Age 'A' and two of 'C' or early Roman date; one of the latter, stone-lined, was possibly a well. The others were Roman or undated. Two areas, one oval with central hearth, sling-stones and a crude flint industry, were probably hut-floors of an early phase of Iron Age 'A'; a third was a working-floor with debris of a hand-cut shale armlet industry and 'A' pottery. Thereafter occupation was apparently continuous into the late 4th century A.D., although Iron Age 'B' elements were confined to stray sherds. The Roman phase was marked by storage-pits, the well (?), fragments of burnt clay daub, and a few short lengths of drystone wall thought to be of the 2nd century or later near which were floor-slabs and stone roof-tiles; there were also two rectangular storage cists, one of reused roof-tiles containing three jugs of perhaps the mid 2nd century, and three shale-working floors with ribbed and light-weight plain armlets, class C armlet cores and flint lathe-tools. Miscellaneous finds included Iron Age loomweights, weaving combs, spindle-whorls, Romano-British brooches of the 1st and 2nd centuries, coins of the 3rd and 4th centuries, a pierced shale slab (in D.C.M., possibly a window-shutter), and fragments of crude pottery containers perhaps connected with trade in salt. Objects are in B.M., D.C.M., Christchurch Museum and Poole Museum. (Dorset Procs. LX (1938), 66–72; LXIX (1947), 42–4; LXX (1948), 29–59; LXXV (1953), 52, 69; LXXXIV (1962), 115; Swanage Times 20 Jan., 1954; information from Mr. J. B. Calkin.)

(40) Gallows Gore, East (SY 97 NE). An Inhumation Burial, Walls, and Occupation Debris, Iron Age 'A' to late Roman, with waste from shale armlet manufacture, have been found since 1931 in Lander's quarry (98177895) N.W. of Gallows Gore Cottages, some 400 yds. E. of (39).

Excavations in the small area remaining unquarried revealed an identical sequence of occupation to that of (39), but without evidence for a pre-Roman shale workshop. No structural features of Iron Age 'A' or 'C' were found, although burnt clay daub and evidence perhaps for leather-working occurred in the 'C' or early Roman level. A thin scatter of flagstones marked the base of Romano-British occupation of perhaps the 2nd to the 4th century. At one point five stone blocks some 6 ft. apart, the largest 2 ft. square and 5 ins. high, were associated with concentrations of shale-workers' debris including worked flints and lathe-tools, shale armlet fragments, and turned cores of class C with a few of class A. These were possibly bases for individual lathes belonging to the primary industry here. Two short lengths of drystone wall here may have served a similar purpose; one of them overlay shale debris of the primary phase. Elsewhere a small storage cist with a paved area adjoining a short length of drystone wall was probably of the 3rd or 4th century. The inhumation, S. of the road (98227875), was of a child in a covered cist of two slabs with a hand-made vessel probably of Iron Age 'A'. Objects are in Christchurch Museum, B.M., and in the possession of Mr. J. B. Calkin. (Dorset Procs. LXIX (1947), 42; LXX (1948), 40–1; LXXV (1953), 45–71; Trans. S.E. Union of Scientific Societies XL (1935), 30; information from Mr. J. B. Calkin.)

(41) Compact Farm (SY 97 NE). Occupation Debris, Iron Age 'A' to late Roman, with waste from shale armlet manufacture, comes from about ¼ m. N.E. of St. Nicholas' Church. Excavation for a silage pit (97427782) revealed a black layer containing fragments of turned shale armlets and waste cores of class C, specialized flint lathe-tools and other objects, with Iron Age sherds on natural limestone beneath. Unstratified finds here and for about 100 yds. N. included Iron Age 'A' sherds, rough-outs for hand-cut shale armlets, part of a turned shale bowl, a bronze signet-ring, and sherds of at least the 2nd to the 4th century A.D. Some objects are in D.C.M. and Christchurch Museum. (Dorset Procs. LXXV (1953), 54, 59, 69; LXXVI (1954), 77; LXXXII (1960), 86; LXXXVI (1964), 118.)

Sites (42–43) lie on Lower Purbeck limestone.

(42) Weston Farm (SY 97 NE). Occupation Debris was found in 1958 on a S.-facing ridge about 350 ft. above sea level (96977730), following examination of soil from the site dumped near (41). The objects (in D.C.M.) were sherds of hand-made pottery probably of Iron Age 'A' and pieces of worked shale including a waste disc for a hand-cut armlet, and Romano-British sherds of 1st to 4th century. (Dorset Procs. LXXXVI (1964), 118.)

(43) St. Alban's or St. Aldhelm's Head (SY 97 NE). Occupation Debris comes from the ridge about 350 ft. above sea level E. of the quarry (96587609).

A thin scatter of debris over an area of about ½ acre included sherds of hand-made pottery probably of Iron Age 'A', samian and Romano-British coarse pottery of perhaps the 1st to the 4th century, a coin of Gallienus (253–68) and five turned shale armlet cores of class C (in possession of Mr. P. A. Brown). A barrow excavated in 1850, 300 yds. S.W. at the angle of a 'Celtic' field of Ancient Field Group (24), had near the top coarse pottery, fragments of shale including a waste core from a turned armlet, and a samian sherd with five coins ranging from c. 98–273 (see Round Barrows, Worth Matravers (34)). Stone settings possibly indicating cist burials have been noted in the area. (Dorset Procs. LXXVIII (1956), 76–7; Purbeck Papers I (1855), 34, 37–8.)