Milkieston Rings
Fort (Prehistoric), House Platform(S) (Prehistoric)
Site Name Milkieston Rings
Classification Fort (Prehistoric), House Platform(S) (Prehistoric)
Alternative Name(s) Milkieston Hill
Canmore ID 51433
Site Number NT24NW 4
NGR NT 24800 45950
Datum OSGB36 - NGR
Permalink http://canmore.org.uk/site/51433
- Council Scottish Borders, The
- Parish Eddleston
- Former Region Borders
- Former District Tweeddale
- Former County Peebles-shire
NT24NW 4 2480 4595.
(NT 2480 4595) Milkieston Rings (NAT) Fort (NR)
OS 6" map (1966)
On the summit of Milkieston Hill, a broad spur which projects NW from Cavarra Hill, there are the complex remains of a fort known as Milkieston Rings. Although it commands a wide sweep of the Eddleston Water from a height of 500' above it, the site does not possess any great natural strength: on the SE the ground rises gently away from it towards Cavarra Hill, while on the other three sides the flanks of the spur are only moderately inclined.
The fort defences clearly represent more than one period of construction, but the remains are difficult to interpret in detail owing to later disturbance of the site caused by quarrying and stone-robbing. At least three successive defensive systems can, however, be distinguished, two of which appear to be unfinished. In the first period it seems probable that the fort was a bivallate structure defended by the stone walls marked IA and IB on the plan.
Internally this fort measures about 160' by 120', and, although it is so severely mutilated that its original appearance is difficult to visualise, it should perhaps be classed with such forts as Black Meldon (NT24SW 3) and Cardrona (NT33NW 3). Both the walls are now reduced to low stony banks, which are spread to a width of as much as 30' in places, or to mere scarps, and no facings are visible at any point. There are two entrance, in the E and SW sides respectively, both of which could be original, although the SW entrance through the inner wall has been used in comparatively recent times as a means of access to a quarry. The quarry has wrecked almost a quarter of the interior of the fort, but in the remaining area there can be seen three platforms for timber houses, one of which abuts wall IA and is bounded on the other sides by a slight bank.
Subsequently, the refortification of the site was begun by the construction, on the N half, of a new series of defences, consisting of two pairs of ramparts (IIA-B and IIC-D): each pair of ramparts encloses a ditch, but no ditch appears to have been dug in the broad space between IIB and IIC. The ramparts seem to have been formed simply of upcast from their respective ditches and exhibit no trace of stonework. Where best preserved, the ditches are about 6' in depth while the ramparts stand to a maximum height of 4' above the unexcavated ground. These earthwork defences are clearly later than wall IB since the latter is overlaid by rampart IIA at one point, and it seems equally evident that they were never finished. At the W end the pairs of ramparts and the two ditches return and unite as though on one side of an entrances. But there is no sign of the opposite side of the entrance, while on the NE the system breaks off abruptly, the ends of the ramparts being left open at the point where the work ceased.
At a later date the work of re-fortification was apparently resumed, a third series of defences, designed to link up with those of the second period, being constructed round the S half of the original fort. This new system in its turn does not seem to have been completed. On the S, it consists of three ramparts (IIIA-C) with two intervening ditches, but, although the inner and outer ramparts continue round the W side to link up with the second period ramparts IIA and IID, the medial rampart dies away a few yards beyond an entrance in the SW side and does not appear to have been continiued beyond that point. Similarly on the E, the triple ramparts have apparently not been completed N of another entrance-gap. A wasted stretch of rampart which links with the inner ditch of the second period defences is probably a continuation of IIIA; but the medial rampart (IIIB) is again absent, while the outer rampart and ditch fade out on as comparatively steep slope, leaving a gap of some 50' between themselves and the second period defences. The two entrances also have an unfinished appearance, the wide gaps in the outer ditch contrasting sharply with the narrow gaps left in the inner and medial ramparts.
To the NW of the fort, a linear earthwork crosses the hillside in a straight line from ENE-WSW, approching to within 65' of the defences at its nearest point. It consists of a ditch measuring 15' wide by 3' deep, with a low bank on the lip nearest the fort and intermittent traces of another on the counterscarp. The ENE of the earthwork has been obliterated by cultivation. It seems reasonable to conclude that this is contemporary with the fort. (This earthwork has been re-numbered as NT24NW 35).
RCAHMS 1967, visited 1959
Whilst generally as described, the remains of the fort on Milkieston Hill appear to consist of only two constructional phases.
Phase 1 consists of two ramparts (IA and IB), together with a third, the W and NE portions of IIIA. This rampart is also preserved as a slight bank in the SE between IIIA and IIIB.
Phase 2 is incomplete but its ultimate form is clear in the N where it consits of two pairs of ramparts each with a medial ditch, IIA and IIB,and IIC and IID. Quarrying has destroyed the latter pair in the NE but their course can be traced until in the SE they run into IIIB and IIIC. Banks IIA and IIB cannot be traced in either the W or E, but in the S the bank IIIA on the plan clearly represents the remains of IIB. The linear earthwork to the N is as described and clearly provides an additional defence on the weakest approach to the fort.
Resurveyed at 1:2500.
Visited by OS (JP) 23 July 1971
Photographed by the RCAHMS in 1980.
Note (21 October 2015 - 20 October 2016)
This fort is situated on the summit of Milkieston Hill, which was taken into a series of decorative plantations in the early 19th century but is now under rough pasture. The fort itself, however, has been heavily quarried and robbed of stone, so much so that the evidently complex sequence has been rendered almost incomprehensible, and while the threefold sequence proposed by RCAHMS investigators in 1959 (RCAHMS 1967, 131-3, no.304, fig 118) cannot be entirely correct, its simplification by the OS into two phases has failed to observe several nuances in the surviving remains. What can be beyond doubt Is that there are two inner enclosures, though whether contemporary circuits as assumed by both the RCAHMS investigators and the OS is not clear, and two outer circuits. Contrary to the OS view, these represent at least two periods of construction, demonstrated at the E entrance and on the SSE, but the RCAHMS investigators are mistaken in their observation of the stratigraphical relationship between the innermost of the outer ramparts and the outer circuit of the inner enclosures, and indeed in showing the second phase ditch cutting on the same line as the inner of their phase three ramparts at the entrance on the NE. Furthermore, their suggestion that both their phase 2 and phase 3 schemes were left unfinished, fails to take into account not only the impact on the appearance of the surviving remains of relatively recent quarrying and stone-robbing, but also of demolition in antiquity. For the purposes of description, the fortifications are best separated into the two inner circuits, and the three or four outer ones. The innermost enclosure has been devastated by the later disruption, which has reduced its rampart to at best a grass-grown mound of rubble and round most of the circuit a ragged stony scarp enclosing an area measuring about 49m from NNW to SSE by 37m transversely (0.13ha). It now has one entrance on the E and a second on the SW, though the latter opens into an area of internal quarrying and is likely to be later, and in the surviving part of the interior there are two probably house-platforms and a low stony ring-bank, this last probably overlying the ruined rampart on the N. The second rampart, pursuing a roughly concentric course around the inner enclosure, has been equally heavily robbed, at best forming a low spread mound of rubble. There is no reason why this should not have been a free-standing enclosure taking in a much larger area measuring about 88m from NNW to SSE by 73m transversely (0.51ha); it too has entrances on the E and SW, and in this case both are likely to be original. This circuit was almost certainly reduced to its present state in antiquity, for in it robbed form that it is apparently overlain by the innermost rampart of the outer enclosure on the NW, which itself can be seen overriding an earlier configuration of these outer defences on the W. This scheme had no fewer than four ramparts around the N half, flanking two wide-spaced medial ditches some 4m in breadth by 1.8m in depth, the whole system returning and uniting on what must have been the N side of a major entrance on the W. Of the other side of this entrance, not a trace is visible, and though this led the RCAHMS investigators to suggest the scheme was unfinished, the area to the S shows every sign of heavy disturbance. A second entrance lay on the NE, but it is unclear if the ramparts and ditches ever returned and united here in the same way, despite the depiction by the first OS surveyors in 1856, and it is apparently blocked by another short length of rampart. The character of the defences on the S, nominally relating to a later scheme on the strength of the sequence where the innermost blocks the entrance on the W, is very different, comprising a belt of three close-set ramparts with intermediate ditches; traces of an earlier defence on this line can be seen between the inner and middle of these ramparts on the SSE. There are apparently entrances through the outer ditch on the ESE and SSW, at each of which there are traces of the outer rampart uniting with the middle rampart on one side of the entrance, but in each case the middle rampart carries on across the gap, and if original features the entrance way must have doglegged, on the ESE to expose the visitor's left side, and on the SSW the right side, at the latter making for a gap in the innermost of the three on the SW. Despite the evident complexity of these defences, the enclosure they form measures 110m from N to S by 85m transversely (0.74ha).
Information from An Atlas of Hillforts of Great Britain and Ireland – 20 October 2016. Atlas of Hillforts SC3663
Sbc Note
Visibility: This is an upstanding earthwork or monument.
Information from Scottish Borders Council
