Appendix: Roman Roads

Pages 116-123

An Inventory of the Historical Monuments in the County of Northamptonshire, Volume 1, Archaeological Sites in North-East Northamptonshire. Originally published by Her Majesty's Stationery Office, London, 1975.

This free content was digitised by double rekeying and sponsored by English Heritage. All rights reserved.



For ease of reference, all Roman roads in the area are described here together and not separately under the parishes through which they pass. The roads are identified by the numbers given to them by I. D. Margary in Roman Roads in Britain, I (1955). Roman roads not previously recognized are not numbered but are given descriptions. This section on Roman roads does not include trackways some of which are probably of Roman date; these are described in the appropriate parish inventory but are listed below (see p. 117).

570 WATER NEWTON to TITCHMARSH (Figs. 122 and 123)

This road enters the area from the N.E., at the N.E. side of Warmington parish (probably at TL 086927), but no trace of it now exists at this point. It can first be recognized where the modern A604 Peterborough-Oundle road turns onto it N.E. of Warmington village (TL 08319227). For the next 750 m. the Roman road is on the same line as the modern one; it is raised about 1 m. above the adjacent land and is intersected by a number of ditches and pit alignments (Warmington (9–14); Fig. 114). Immediately N. of Warmington village (TL 07979161) the modern road turns W.S.W., but there is no indication that the Roman road ran on, in a S.W. direction, from this point. The area, low-lying and formerly marshy, bears no trace of any agger in the present arable. It is likely that the Roman road also turned W.S.W. in order to avoid the low ground; then swinging S. at Eaglethorpe (TL 07649153), it probably passed across the W. side of Warmington village approximately on the line of the present lane.

Definite remains of the road appear S.W. of Warmington (TL 07439085) and can be traced as a mainly straight track which runs S.W., and passes alongside the large prehistoric and Roman settlement in Tansor parish (Tansor (1–3); Fig. 99). This track meets an old lane running E. from Ashton (TL 06138837) beyond which it cannot be traced. However another lane, which leaves the Ashton lane further W. is, to the S., on the alignment of the Roman road and is still slightly raised above the surrounding ground. After crossing the Oundle-Polebrook road (TL 05958744) the alignment follows the modern road to Barnwell and the slight bend in the latter, where it bridges a small brook close to the R. Nene (TL 05828689), is almost certainly where the Roman road also crossed. Further S. again, the modern road swings E. in a broad arc, and although the Roman road undoubtedly runs straight on at this point there is no trace of it in the present arable land. Thereafter the modern road and the parish boundary between Barnwell and Polebrook follow the alignment, and where the modern road turns W., at TL 05408612, the boundary runs on and has a well-marked ridge, or agger, on its E. side, 30 m. wide and 0.5 m. high. This ridge may be no more than a plough-headland of the medieval period although possibly the Roman road lies beneath it. To the W. of the Blind Spinney (TL 05188571) the road enters the grounds of Barnwell Castle and is then lost for a short distance. It reappears S.W. of the castle on the same alignment as the modern Well Lane (TL 04848518–04738501) and, after passing through the old station yard and crossing the railway, meets the present Thrapston-Oundle road (A604). In 1972 a sewerage trench across the line of the road, S.W. of the castle, revealed limestone cobbling nearly 4 m. wide (BNFAS, 8 (1973), 5).

For the next 6½ km. the line of the Roman road coincides with the almost straight modern road, and passes the two Roman and Iron Age settlements in Thorpe Achurch (2 and 3) (Fig. 101). Although there is a series of very slight bends in the road S. of Thorpe Waterville, it is probable that the Roman road also took this line over the marshy area on either side of Thorpe Brook (TL 02268126). Further S.W. there are large Roman settlements on either side of the road (Titchmarsh (16– 19); Fig. 107). The Roman road leaves the modern one (TL 00837960) E. of Titchmarsh, and is visible on air photographs as a white stony line running into the large Roman settlement N. of Springfield Cottage (Titchmarsh (22), TL 006794) where it meets the Roman road from Godmanchester to Leicester (57a) (Fig. 108).

It has been said that the Roman road continued S. from this settlement, through Thrapston, Denford, Ringstead, Higham Ferrers and Rushden, and turned S.W. to the Roman town of Irchester (The Viatores, Roman Roads in the S.-E. Midlands, (1964), 339–341, 494–5). There is little evidence to support this suggested line: no road is visible on air photographs running S. from the settlement; moreover, the suggested line includes modern footpaths laid out by Parliamentary enclosures. A more likely interpretation is that the road turned N.W. at the Titchmarsh settlement, following the 57a road across the R. Nene as far as Lowick where it turned S. again and continued on the W. side of the river, via Woodford and Irthlingborough to Irchester (see below).

Continuation of 570, LOWICK to IRCHESTER (Fig. 124)

This road is ill-recorded, but such evidence as exists would suggest that it ran from the 57a near Lowick along the W. side of the R. Nene to Irchester. It is thus either the continuation of the Roman road 570, or an alternative route to Irchester if the suggestion of The Viatores (op. cit.) that 570 ran through Thrapston and Raunds is accepted.

The road appears to start on the S.W. side of Oxen Wood in Lowick parish from the known alignment of road 57a at SP 98328170. A little to the S.W. a well-marked line of limestone rubble, 7 m. wide and 0.3 m. high, is still visible running in a S.W. direction. If this is in fact the road, its projected alignment passes just S.E. of Lowick village and across Harper's Brook near SP 97708060. At this point there is a low ridge of limestone rubble on the same alignment (BNFAS, 7 (1972), 26). To the S.W. of the brook the same general alignment is followed by an almost straight hedge on the S.E. side of Drayton Park (SP 97668057–96747950). Thereafter the direction changes almost due S. and, although much distorted, the modern road to Woodford, which crosses the Cranford Brook, is probably on its general line (SP 96797907–96657697). The direction of the Roman road is probably preserved by the present Addington road out of Woodford for some 600 m. (SP 96677677–96497616). At the sharp bend in the Addington road it presumably continued S.S.W., as shown by a well-marked ridge existing 600 m. further on. The ridge is 12 m. wide, with side ditches, and is visible on air photographs (in NMR). It can be traced for a distance of 400 m. (SP 96377556–96227520) to a point where it is intersected by a ditched trackway almost at right angles (Great Addington (4); Fig. 13).

The Roman road then crossed a small brook and passed to the E. of Home Farm (SP 96127477). It is next visible immediately S. of the Little Addington parish boundary, as a low bank 8 m. wide running almost N.-S. for 200 m. (SP 96167428– 96197408). In 1967, a section cut through this bank (at SP 962742) revealed an agger, 8 m. wide and just under 1 m. high, composed of limestone blocks within clay and gravel. Except for the discovery of medieval pottery sherds from the top-soil and on the surface of the road, no dating evidence emerged (BNFAS, 2 (1967), 33–4).

Immediately to the S. of this section the modern road returns to the alignment of the Roman road which probably underlies the present one along the straight section to Irthlingborough (SP 94867085). Thereafter its exact route to Irchester is unknown.

57a GODMANCHESTER to LEICESTER (Figs. 125 and 126)

This road enters the county from the E., S.E. of Clopton village (TL 06707873). Its line is indicated by the almost straight modern road running to Titchmarsh village which, for part of its length, is also the Clopton-Titchmarsh parish boundary. At the S.E. end of Titchmarsh village (TL 02857928) the modern road turns N.W. and the line of the Roman road is lost, but it is certainly found on the original alignment, 2 km. further on. It is visible on air photographs as a broad, light-coloured strip crossing the modern A604 road and entering the Roman settlement in the S.W. corner of Titchmarsh parish at TL 00767947 (Titchmarsh (22); Fig. 108). A drainage ditch in 1965 revealed a section of the road, 1 m. below the present ground surface. It had a limestone core 8 cm. thick and 5.5 m. wide, capped by 23 cm. of gravel, 7 m. across. No side ditches were recorded (BNFAS, 1 (1966), 13). Immediately W. of this point the Roman road joins the 570 road from Water Newton, and can be seen on air photographs passing through the settlement (Titchmarsh (22)), but after crossing the abandoned railway, all trace has been destroyed by a gravel pit. The agger was noted here in 1961–2 during gravel-workings (BNFAS, 4 (1970), 43). Further N.W. the road crossed Harper's Brook which in Roman times was the main course of the R. Nene. Here (SP 99928011) a timber bridge which carried the road was discovered and excavated in 1967–8 (Aldwincle (11); Fig. 16). N.W. of the bridge 300 m. of the road was also examined. The agger had been removed by gravel-digging but the side ditches, up to 2.5 m. deep and 12 m. apart, were traced. Other pits, apparently dug to provide gravel for the roadway, were noted on both sides of the road (BNFAS, 4 (1970), 36–7). Close to the bridge, another road running N.E. from 57a was recorded.

To the N.W. of the Nene flood-plain no trace of the road is visible, but it apparently passed under the present Aldwincle Lodge (SP 99308081). It reappears N.E. of Bullocks Wood (at SP 98908121) as an old hedge-line on a well-marked ridge 10 m. wide. This ridge is a former headland of the medieval common fields of Aldwincle, for traces of ridge-and-furrow can be seen running onto it. However, there can be no doubt that the Roman road was here used later as a headland. This ridge extends as far as the S. corner of Oxen Wood (SP 98448144). Thereafter the alignment is taken up by the Aldwincle-Lowick parish boundary running alongside Oxen Wood. N.W. of the wood (SP 98008200) the line is again lost, but 400 m. beyond, and cutting across the S.W. corner of Titchmarsh Wood (SP 97748225–97508247), is a broad hollow-way, 14 m. wide and 1.5 m. deep, cut into the valley side. For the next 630 m. there is no trace, but from SP 97058290 to 96208375 the line is visible on air photographs (held by the Monk's Wood Research Station, Huntingdon) as indistinct soil and crop-marks. S.E. of Cat's Head Wood, in the N. of Sudborough parish (at 96138382), is a short length of agger 10 m. wide and 0.3 m. high. Within the wood (SP 96108385–95958399) and continuing the alignment, the agger is replaced by a hollow way, 12 m. wide and 1 m. deep. N.W. of the wood, after a short break, an existing hedge is on the same alignment (SP 95858408– 95618429).

Beyond this point there is no trace of the road until it has passed the boundary of the area under review. It apparently continued N.W. through the S.W. side of Brigstock village and on towards Stanion.


This length of road branched off the Ermine Street, N. of Durobrivae, and its line is followed by the present A47 Peterborough-Leicester road as far as Wansford. Entering the area N.E. of Wansford (at TL 06819943) as an existing road to King's Cliffe, the Roman road runs almost straight for 5.5 km.; S.W. of Bedford Purlieus (TL 03609850) it is still raised 0.5–1 m. above the adjacent ground. Immediately N.E. of King's Cliffe it runs down the steep E. side of a small valley where it ends abruptly (at TL 01289781). No trace of its continuation has been found, but it probably ran down into the valley of Willow Brook at King's Cliffe, continuing S.W. to meet the 57a road S.W. of Corby.

From ALDWINCLE to ? (Fig. 125)

During excavations at Henslow Meadow, Aldwincle, a length of Roman road was uncovered, running N.E. from the 57a road just W. of where it crossed the R. Nene (Fig. 16). It was exposed for some 500 m. and only two parallel ditches, 10 m. apart, with traces of limestone and gravel metalling, were discovered. Where the road led is unknown (BNFAS, 7 (1972), 12; DOE Archaeological Excavations 1971, (1972), 21)

Ditched Trackways

These are described under their relevant parishes but are listed here for reference. The dates of the majority are unknown, but some may belong to the Roman period. Most have been recognized only from air photographs.

Addington, Great (4), (5)

Addington, Little (5)

Apethorpe (2)

Fotheringhay (13), (25), (31)

Islip (4), (5)

Nassington (23)

Oundle (7), (8)

Raunds (13)

Ringstead (2), (4)

Southwick (7), (10)

Stoke Doyle (1), (4)

Sudborough (5)

Tansor (5)

Thorpe Achurch (2), (3)

Warmington (5), (9)

Fig. 122 Roman Road 570 Elton-Barnwell

Fig. 123 Roman Road 570 Barnwell-Titchmarsh

Fig. 124 Roman Road 570(?) Lowick-Irthlingborough

Fig. 125 Roman Road 57a Clopton-Aldwincle

Fig. 126 Roman Road 57a Aldwincle-Stanion

Fig. 127 Roman Road 571 Wansford-King's Cliffe