More information : (ST 969124 to SU 040192) Cursus (NR) Cursus (NR) (course of) (twice). (1)
A cursus circa 6.25 miles (10 kilometres) long which runs roughly southwest-northeast between Thickthorn Down and Martin Down. Narrow and roughly parallel-sided, it follows a slightly sinuous course across the chalk downland, crossing a river and several valleys. The monument in fact comprises two cursuses laid out end to end, the earlier southwestern (Gussage) portion terminating on Bottlebush Down. Both are defined by parallel banks with external ditches, though for much of their length these are visible only as cropmarks or soilmarks. The southwestern end survives as an earthwork. RCHME fieldwork in the (?) 1950s recorded the bank here as surviving to a height of more than 4 feet in places, and up to 30 feet in width. The ditch was also up to 30 feet across and up to 3.5 feet deep. The width of the cursus at this point was measured as circa 350 feet. This southwestern stretch is relatively straight though it changes course northeast of Gussage Hill before heading to its terminal on Bottlebush Down. The northeastern (Pentridge) Cursus then continues on a slightly different alignment, adjusting its course slightly on a few occasions before terminating on Martin Down. The cursus incorporates two long barrows. On Gussage Hill, ST 91 SE 9 lies across the interior of the cursus, while the Pentridge Cursus has SU 01 NW 47 actually incorporated into the bank of the northwestern side, although the barrow's alignment differs slightly. The full extent of the Dorset Cursus was not appreciated until the publication of Atkinson's fieldwork in 1955. The central portion, from Gussage Down to a point southwest of Pentridge, had been known as an open-ended earthwork since at least the early 19th century. The terminals survive as earthworks but hadn't been identified as parts of the same monument. The only published excavations were undertaken in 1982 by Barrett at al, who sought to interpret the cursus within a study of the prehistoric landscape of Cranborne Chase.
The following description of the cursus is derived primarily from work by RCHME (published in 1975; the fieldwork began some decades before, but details were checked in the field during 1971); the Ordnance Survey's Archaeology Division (mostly undertaken in 1954 - see the annotated 6" map in the NMR Linear Files); plus the accounts of Atkinson (1955) and Barrett et al (1991):
The cursus is a long, narrow parallel-sided enclosure just over 6.25 miles in length and varying between 300 and 400 feet in width. It has clearly been constructed in at least two phases. Orientated roughly southwest-northeast, it comprises two cursuses laid end to end. The southwestern portion, the Gussage Cursus, runs from Thickthorn Down northeast towards Bottlebush Down were it terminates at a transverse bank and ditch. The northeastern section, the Pentridge Cursus, runs from Bottlebush Down northeast towards Martin Down, terminating just southwest of the line of the later Bokerley Dyke. For much of its length, the Dorset Cursus has been levelled or severely reduced by ploughing, long stretches visible only as cropmarks or soilmarks. The southwestern terminal survives as an earthwork in a narrow, unploughed field on Thickthorn Down. RCHME recorded the bank as surviving to a height of more than 4 feet in places, and measures up to 30 feet in width. The corners, where the bank is noticeably higher than elsewhere, appear rounded. The external ditch also measures up to 30 feet across and up to 3.5 feet deep. Atkinson (1954) suggested that the ditch and bank may have been separated originally by a narrow berm. The width of the cursus at this southwestern end was 350 feet between the centre points of the ditches. For some distance northeast of the Thickthorn terminal the Cursus is largely visible only as a crop or soilmark.
In 1954, the Ordnance Survey recorded possible traces at ST 97671292 (northern side - a "distinct bump on hedge bank") and between ST 98021298 and ST 98071303 (southern side - "Hollow 5.5m wide, 0.7m deep with a low spread bank abutting it on the NW"). A well-preserved stretch is visible as earthworks on Gussage Hill, partly overlain by the earthworks of later prehistoric settlement and fields (see OS 6" map for details). Just northeast of Gussage Hill, the Cursus deviates slightly as it descends to cross the dry upper reaches of the Allen valley, before climbing beyond the point where the Cursus is crossed by the Roman road (at circa SU 010152). After a short ploughed-out stretch, the earthworks reappear in the vicinity of the Bottlebush terminal, which is visible as a transverse ditch and bank joining the northern and southern sides of the Cursus (at circa SU 01551565).
The cursus then continues beyond the Bottlebush terminal on a slightly different alignment, making slight changes in direction as it traverses a dry valley and then climbs a narrow spur as it runs through Salisbury Plantation, this stretch represented more by earthworks than soil or cropmarks. Beyond the spur, the Cursus runs in a relatively straight line towards its northeastern terminal, although only the terminal itself (at SU 04051920) survives as earthworks. A slight change of alignment occurs close to the terminal itself. In 1954, the Ordnance Survey described this terminal as a "wide and spread bank and outer ditch. Average measurements:- Bank 8.5m wide with inner height 0.3 metres. Outer height 0.7m. Ditch 12m wide, 0.3m deep." Probing by Atkinson during the course of his fieldwork (1948-53) apparently identified two narrow opposed causeways of unexcavated chalk across the ditches, located circa 1.25 miles from the terminal (specifically, according to Atkinson, 1300ft northeast of the approach road to Pentridge).
The full extent of the Cursus was not appreciated until the publication of Atkinson's fieldwork. The central portion, roughly the part from Gussage Down to a point southwest of Pentridge, appears to have been observed and described first by William Cunnington in the early 19th century. His description of the extant earthworks was used by Colt Hoare who included the site and associated earthworks in his "Ancient Wiltshire...", notwithstanding its Dorset location. Speculations on possible continuation of this apparently open-ended earthwork were presented by a number of individuals. In the later 19th century, Charles Warne for example offered a suggested course (which ultimately proved incorrect). In the early 20th century Heywood Sumner made some observations of the Gussage Down section, primarily in connection with the later settlement (ST 91 SE 3 and associated monuments). He was also the first to notice the incorporation of a long barrow (SU 01 NW 47) within the northwestern bank of the Cursus. Sumner also sketched and described the southwestern terminal, although he did not recognise it as such. He interpreted the earthworks as representing around half of a rectangular enclosure, a defensive "camp" which he suggested might be connected with the multiple ditch system running parallel to the northwestern cursus bank at this point (ST 91 SE 41). The first reported excavation within the cursus appears to have been undertaken by Alexander Keiller in the late 1920s or early 1930s in search of the missing eastern side of this proposed enclosure. The only published reference to this excavation occurred as a brief aside in Drew and Piggott's report on the nearby Thickthorn Down long barrow (ST 91 SE 21). Among monuments in the vicinity, they referred to "...the western half of a square earth-work of unknown date (the missing side was proved by a trial trench made by Mr A Keiller some years ago")...". Unfortunately, nothing more is known of this episode. Certainly no record, written or photographic, appears to exist among the Keiller archives held at the Alexander Keiller Museum, Avebury. A possible context is provided by a note in WEV Young's diary (quoted in Murray 1999, 66) referring to the fact that the Thickthorn Down barrow "was one visited by Keiller after observing it from the air on July 14th 1924". Indeed in Keiller and Crawford's "Wessex from the Air", the barrow is referred to in a list of sites "discovered but not photographed". The entry describes the barrow as being located "immediately SE of a mutilated earthwork [and] was observed to be a long barrow. A visit to the spot confirmed this diagnosis, and enabled measurements to be taken by Mr Keiller". Keiller may have taken the opportunity to examine the nearby "mutilated earthwork" as well, but this is purely speculation.
The nearest that Atkinson came to excavation was in the Spring of 1953 when the southeast ditch was partly exposed in section in a hole dug for a fence post against a hedge which ran close to and parallel with the cursus ditch at circa ST 978128. No finds were mentioned by Atkinson and it is not clear what record, if any, was made of the observed section.
The most substantial excavations occurred in 1982 and 1984 during a programme of work undertaken by John Barrett, Richard Bradley and Martin Green on Cranborne Chase. The 1984 excavation involved cutting a transect across the entire width of the cursus interior at a point where a surface artefact scatter could be seen to spread across the cursus interior and over and beyond the southeast bank. The grid reference for this excavation is not mentioned in the final report, but an interim (Bradley et al 1984) locates to SU 00601488, although the the accompanying plan suggests that the trench was closer to SU 00651485. The 1982 excavation occurred some 700 metres to the southwest. In addition, of course, work was also undertaken on sites in the vicinity, all of which have a bearing on the interpretation of the cursus. In both cursus excavations, the northwestern side was examined in most detail. Despite contrasting topography at each location, the northwestern ditch was of similar form at both - steep sided and flat bottomed. Where it crossed level ground, it was 3m wide at the top, 2 metres wide at the bottom and had been dug 1.2 metres into the chalk. At the 1984 trench, it abutted a Pleistocene river cliff. In both cases there were indications that the ditch had been accompanied by an internal bank. The proximity of the bank to the ditch, and the apparent narrowness of the original bank, prompted suggestions that the latter may have been revetted. The southeastern ditch was sectioned only in 1984. It was notably shallower, possibly because of the relative hardness of the chalk, and may have been recut at least twice. At least one of the recuts terminated in a causeway.
Early Neolithic pottery was recovered from the primary silts, as was a quantity of knapping debris and some small fragments of animal bone. An antler pick was also found lying on the surface of the primary fill. The bone and antler produced radiocarbon dates spanning the second half of the 4th millennium BC (calibrated), though none of the samples are securely associated with the primary construction phase. Two further samples produced dates comfortably in the Mesolithic. Peterborough Ware and Beaker sherds were recovered from secondary levels in the ditch fills (see SU 01 SW 143 for details of the later activity encountered in the 1984 excavation). However, it is clear that even in the Late Neolithic and beyond, the presence (and presumably significance) of the cursus continued to influence activity in the vicinity, including the construction and use of further monuments (see associated monument records).
As for the manner of its construction, Atkinson had argued that the width was a fairly constant 300 ft, though variation is clearly evident throughout the full length of the monument. He suggested that "all obviously local irregularities occur in the NW ditch. From this it may be conjectured that the SE side was the one to be ranged out and marked initially on the ground, the position of the outer ditch being determined by offsets taken from the former at intervals, and evidently not always measured with equal accuracy". However, Barrett et al noted that the cursus was clearly built in stages, and argued that while Atkinson's suggestion seems to work at some points along the cursus, there are other places where the situation is reversed, notably "on either side of the Allen Valley...This may be reflected by [the ditches'] contrasting character on excavation. The same interpretation is supported by air photographs, which show that at one point the east ditch of the Cursus was dug continuously whilst its west ditch ran in a seried of slight arcs, separated by wide causeways. At this point, the main earthwork seems to have replaced a slighter ditch on the same line; this can be observed again about a kilometre north of the Bottlebush terminal."
The Cursus follows an irregular course, changing direction on a number of occasions, and all changes of course are located at breaks in the surface topography, underlining the difficulties involved in establishing lengthy alignments across such a landscape. This, and the environmental data obtained from the excavations, highlight the fact that the Cursus need not have been constructed in an open landscape. One notable change of direction occurs just after the Gussage portion of the cursus crosses Gussage Hill, where it changes course twice in just over 200 metres. Barrett et al argue that this is because "the length of the Cursus running south from the Bottlebush terminal was aligned exactly on a long barrow [ST 91 SE 9] on the crest of Gussage Cow Down, but seems to have drifted off course in the lee of the hill where that landmark could not be observed - the change of direction was needed to enclose the barrow within the monument. Its importance as a skyline feature is emphasised by the positioning of the Bottlebush terminal below the crest of the hill; it seems to have been located on rather lower ground so that the barrow would appear on the horizon." Indeed, the stretch from the Bottlebush terminal to the long barrow is the longest part of the Cursus that can be observed from a single vantage point. The barrow will have been visible from both terminals, but the terminals of this first phase cursus are not themselves intervisible. It has been suggested that at least part of the reason for the visual relationship between the Bottlebush terminal and the long barrow was to enable observation of the setting of the midwinter sun behind the mound. Of the various astronomical (and other) alignments read into the cursus and its neighbouring monuments, this remains the most convincing.
The Martin Down terminal, meanwhile, is not located at a particularly prominent spot, and like much of the cursus would have had little visual impact outside the monument. However, like the Bottlebush terminal it would have been possible to view a considerable length of the monument and also surrounding monuments from it.
As well as incorporating the two long barrows, several round barrows were later constructed within the cursus (in addition ot the numerous round barrows built in its immediate environs). Those within the interior include ST 91 SE 122 and 123, and SU 01 NW 65 and 72. During the later Bronze Age or Iron Age, the Cursus, or at least part of it, was incorporated into a complex of linear earthworks and broadly contemporary field systems. This is most clearly demonstrated at the Cursus' southwestern end where the multiple ditch system ST 91 SE 41 approaches the Cursus from the southwest before changing direction slightly and running parallel to, and partially overlying, the northern side of the Cursus for a few hundred metres. Excavation across the multiple ditch system close to the Cursus terminal recovered residual Beaker pottery. [NB Barrett et al claim that the Down farm pond barrow (SU 01 SW 83) shares the same axis as the Cursus, which passes circa 35 metres to the south. However, the published illustrations suggest that this is incorrect - the cursus runs northeast-southwest, whereas the pond barrow's axis is presented as northwest-southeast]
The history of the cursus and its relationship to neighbouring monuments is fully discussed by Barrett et al (1991), and has been developed by other subsequently (eg Gosden 1994; Tilley 1994). Some aspects are also noted in individual monument records (see associated monuments). |