The following is the text of an article which was published in the Proceedings of the Hampshire Field Club, by whose kind permission it is published here. It outlines the search for a road by our member David Weston and subsequent excavations under our auspices led by David Calow.

  “Upon reaching Winchester the southern road again divided …. forming an almost complete radial network from this centre, which would indeed be quite symmetrical if the short straight road to the east were included, but as yet no continuation of this road has been found”  I. D. MARGARY [i]

Since the earliest days of research into the roads of Roman Britain, there has been a belief that a road existed to the east of Winchester. It is presumed to have run north-eastwards towards Alton, along the valley of the river Wey, and onwards to London.  Yet its route is unknown, and its very existence is still not proven.  Excavations at Neatham (a small Roman town just east of Alton) suggested a crossroads with the south-westerly leg presumed to be the road to Winchester[ii], but its line whilst implied, has not been established.  Recently, the search has gained a new impetus, with a number of archaeological societies in the area undertaking excavations to locate it.  This article is designed to further the debate, by demonstrating that significant evidence (both circumstantial and actual) exists, indicating a possible route of the road between Winchester and Neatham.

Neatham has been tentatively identified as Vindomi, a town that appears in number XV of The Antonine Itinera.  It was probably compiled in the third century AD, using older sources. The relevant entry reads:-

“Likewise from Calleva to Isca Dumnoniorum 136,000 paces, thus written, Vindomi 15, Venta Belgarum.21…”.

Although it may seem strange that any route from Silchester to Winchester should go via Neatham, (given the existence of a direct Silchester to Winchester road), Professor Rivet and others have pointed out that few of the itinera do in fact take the most direct routes[iii].  Any conclusive proof that Neatham is Vindomi, would simultaneously prove that a road between Winchester and Neatham also existed, but sadly no such proof exists. However, if we accept the hypothesis, then Neatham should be 21 Roman miles from Winchester. Since a direct line between Winchester and Neatham is barely 19 Roman miles, it would suggest that if Neatham is Vindomi then the road did not take a direct route.  Any attempt to locate the road should bear this in mind, whilst  seeking to explain why.

The straight section of road (now the B3404) from Winchester’s east gate over Magdalen Hill Down is an obvious candidate for the western end of the route, as was pointed out by Margary in the opening quote.   At the eastern end of the route, a 3 mile straight alignment of roads running through Four Marks (Brislands & Blackberry Lanes), and extending along the A31 towards Alton also has merit.  These two potential eastern and western sections do not align.  Nevertheless, the often suggested route links both together by a straight line passing through the Ropley valley.  Blackberry Lane’s case for Roman origin is strengthened because it is known to be a very old road, and an extension of its alignment to the north-east, goes to the very centre of Roman Neatham.  On the other hand, its direction is not obvious for a road to Winchester, heading more to the south-west than would be expected. Intriguingly, if its alignment is extended further south-westwards, it goes to the centre of Roman Clausentum (Bitterne). It might be argued therefore that if Blackberry Lane is Roman in origin at all, it reflects a probably uncompleted road between Neatham and Bitterne, rather than one to Winchester.  Of course, this would not preclude a junction for Winchester coming off of it, but it seems illogical to suggest that Winchester was its originally intended destination.  Straight roads are not necessarily Roman in origin, and with regards to the road heading east out of Winchester, Professor C.F.C. Hawkes, writing in the 1920s, was convinced that it wasn’t Roman at all[iv]. He believed that the straightness of the road was the result of its construction as a turnpike in 1753. Insufficient space precludes a discussion of his case, but evidence discovered since his time suggests that he was wrong, and that this section of the road is indeed Roman.  In 1999 whilst cutting a trench to lay cables opposite Magdalen Hill Cemetery, a thick band of flint metalling was observed under the modern road.  It was considered that this may have formed part of the surface of a Roman road heading east out of Winchester[v]. Strong evidence also came from post-war aerial photographs in the area where the modern road ceases its eastward course, and bends sharply north-eastwards towards Alresford[vi].  These show the soil marks of an ancient double ditched, straight line feature, assumed to be a road running north eastwards from the bend.  It can also be distinguished physically on the ground as a faint terrace running towards the cottages northwest of Orr’s Meadow, and marks the parish boundary. The area around this bend undoubtedly holds the key to locating the subsequent route of the road to Neatham, but it is enigmatic.  To head directly towards Neatham or Blackberry Lane, the road would need to descend into the steep sided Itchen valley.  Although by no means insurmountable, the descent would have required some form of engineering (probably a zig-zag), but nothing is obvious on the ground.  Moreover, there is an abundance of good aerial photographs, clearly showing archaeological and geological features in the area between the bend and the Itchen.  None show any trace of a road.  If other ancient features show and the road doesn’t, the implication is that there was never a road there.  With no evidence of a crossing of the Itchen east of the bend, one has to believe the evidence indicating that the Roman road turned sharply north east at this point.  This raises a serious question. Why build a road due east from Winchester, only to abruptly change direction 3½ miles later, when a classic straight Roman road could have been constructed directly to the crossing point of the Itchen, if not Neatham itself? 

One possibility is that the “bend” was not a bend at all, and that this section was not built with Neatham in mind, but rather as a road in its own right.  It is entirely feasible to postulate a pre-Roman route or trackways from northeast and central Hampshire to the tidal reaches of the Itchen, utilising the spur on the eastern side of Cheesefoot Head as a means of crossing the east-west  line of the downs. For the new Roman city of Venta, it would have made good sense to build a short, straight road, at an early date, due east to link to this route, to ensure that traffic did not bypass the civitas capital on its way to or from the coast.  Once constructed, the track over Cheesefoot Head would have become much less important and more difficult than the metalled road through the city, and the junction would have rapidly become a bend.  At some later stage one or other of the routes heading north east would have been built as a fully engineered Roman road. Unfortunately, the ancient road ceases to be visible within a few hundred metres of its turning north eastwards, and its subsequent course is far from clear.  One possibility is that its route is reflected in the parish boundary running towards Ovington.  More likely is that its initial direction is a result of very local geography (a steep slope immediately to the east); and that having negotiated that, it followed one of the postulated pre-Roman routes, along the higher ground, and down the spur into the Itchen/Arle valley near Seward’s Bridge, just west of Alresford.  This is the same route that was used by the 17^th century coach road, the turnpike, and the present day A31. 

Although a medieval road/river crossing at Seward’s bridge was probably built in the 12th century, and was certainly in existence by the early 13th century, it is likely that the medieval builders would have chosen to cross the river at an established crossing point.  Given its location at the confluence of the Itchen and Alre, it is quite conceivable that Seward’s bridge was the Roman crossing point.  The strong, straight, north-east/south-west alignment of roads and tracks on both sides of the river around the bridge, clearly visible in aerial photographs and maps from Ogilby to the present day, would support this proposition.  An extension of those alignments to the  northeast of the river, correlates closely with the old ridgeway track, now called the Oxdrove Way, and perhaps more significantly, runs within a few hundred metres of the known villa site at Pinglestone Farm. 

If at some later date it had been decided to build a road from Winchester to Neatham and further eastwards, then the most obvious starting point would have been here, at the Itchen crossing, and not the city itself or “the bend”.  A straight line from Seward’s Bridge to Neatham would take it to the north of previously postulated routes, and would run through New Alresford and Bighton, crossing the Itchen/Wey watershed near Medstead and then running on through Alton to Neatham.  There are historical reports of a Roman road between Alresford and Alton.

“From this place [Alresford] to Alton there goes all along a Roman Highway, part of which makes a Head to an extraordinary great Pond here at Alresford;”  (William Camden 1607).
“All along, a perfect Roman way from Aulton to Alresford” (John Aubrey 1690).
“Here [Alresford] is a very large pond, or lake of water, kept up to a head, by a strong batterd'eau, or dam, which the people tell us was made by the Romans; and that it is to this day part of the great Roman highway, which leads from Winchester to Alton, and, as 'tis supposed, went on to London, tho' we no where see any remains of it, except between Winchester and Alton, and chiefly between this town and Alton.”  (Daniel Defoe 1724).
“From this Place to Alton there goes a Roman Highway, of which part serves for an Head or Stank to a great Pond, or little Lake, at this Town”: (Thomas Cox 1738).

There is insufficient space to discuss why people linked Alresford dam (built at the very beginning of the 13^th century) with a Roman road.  The important fact is that reliable witnesses claimed to have seen a Roman road between Alresford and Alton.  The critical significance of their reports is that the road they themselves would most likely have been on, (and from which they saw sections of a Roman road), was the road through Bighton.  This was the main highway throughout the 17^th century, and had been since the middle ages; whereas the current route to Alton via the Ropley valley, did not come into being until the building of the turnpikes.

Bighton parish is unusual in many ways.  It has almost parallel boundaries along its northern and southern sides, and a greater number of extant and destroyed field boundaries running parallel or at 90 degrees to them, than might be expected.  Geography may play a part in these alignments, but cannot explain everything, as a number of the boundaries seem to relate to Roman units of measurement. There is a clear suggestion that the boundaries and lanes are mirroring something within the area.  The possibilities are that they reflect a Roman road, or perhaps even an ancient estate system, the ghost of which can be seen in the modern parish.  There is a known villa site to the north east of Bighton. 

A low terrace runs behind the properties that front the medieval road through the village.  This terrace has in part been accentuated by the levelling of ground to build the houses, but aerial photographs and maps show that it is older than the buildings. 

This feature is within metres of the straight line between Seward’s bridge and Neatham.  Approximately a mile to the east of the village, a 1946 aerial photograph[vii] shows a very strong feature with all the hallmarks of a Roman road on exactly the same alignment.  The feature is also visible in photographs taken in 1970 and 1982[viii].  In each of the subsequent photographs however, increasing damage is evident, with the lighter deposits spreading down the slope.  Something similar, but much weaker, can be seen on the same alignment in a field to the west of the village.  Field walking demonstrated a greater density of flints on the surface over the feature than elsewhere in the field.  The feature was resistance surveyed, and excavated under the direction of David Calow of Surrey Archaeological Society during the summer of 2007.  On the basis of the excavations and aerial photography, it was interpreted as a heavily damaged, engineered Roman road, about 5m wide, made from sorted and prepared flints, and cut into the chalk hillside. It is quite conceivable that this was what Camden, Aubrey and Defoe saw, prior to its subsequent destruction and disappearance.

If this object is the Winchester to Neatham road then the question of where it crossed the watershed into the Wey valley arises. Difficult terrain particularly on the eastern slopes of the high ground needed to be negotiated.  The position and alignment of the feature in the field east of Bighton, precludes it from taking the route followed by the subsequent medieval road; and besides such a routing would take it well away from the direct line between Seward’s Bridge and Neatham.  However, a very small (approx 3º), change in direction, would take it to a highpoint just west of Medstead at SU652367.  At this point tracks from 5 different directions meet, all of these paths appear on the oldest maps, and have clearly been in existence for many years.  The one running south westwards, points directly at the eastern extension of the Bighton feature, and is itself a continuation of a small straight section of the otherwise winding modern road between Bighton and Medstead.  The views from this highpoint are expansive, and it is the sort of location that Roman road builders might have used as a sighting point.  It is approximately 200 metres north of the direct line between Seward’s bridge and Neatham.   A small directional change at this high point (southwards by 6º), would not only have returned the road to the direct line, but also would have kept it on a ridge, avoiding the steep ends and sides of the valleys at Beech and Chawton Park Wood.  Angular changes of this order at high points are quite consistent with Roman road building practice, and the proposed alignment is coincident with the lower section of the eastern track from the high point.  It also coincides with another old track and Holloway (Foul Lane) just east of Medstead, the rear boundaries of properties along the Beech road, and most significantly a straight line feature, invisible on the ground, but clearly showing in a 1984 aerial photograph[ix] in the field immediately south of Abbey Road at Old Park farm.  This feature, which extends for approximately 600 metres on the aerial photograph is also partially visible on a 1967 aerial photograph, and was excavated during the summer of 2007 under the auspices of Surrey Archaeological Society[x].  It was interpreted as an engineered flint, or flint and clay road, approximately 5 – 7m wide which had been damaged to a greater or lesser extent over its length.  If this feature is extended eastwards, it runs into Bushy Leaze wood, at a point where an ancient park boundary ditch and bank (probably medieval), comes from the south and makes a sharp turn onto a similar alignment.  The park pale may very roughly reflect the line of the road, but further eastwards in an area made difficult by modern forestry and old brick works, the course is unclear.  The most obvious route would have been to continue above the southern edge of the Beech valley (where a terrace is visible) and descend into the Wey valley in the area of Wyard’s farm.  Here, the straight line from Seward’s bridge would have been regained and the road could have continued on dry ground, passing just to the north of the source of the Wey, directly to the crossroads at Neatham.  There are however other plausible routes through Alton that space here precludes describing, and further research is needed.

In this brief article, I hope to have established that there is sufficient evidence to suggest that a major, Roman road east of Winchester existed, and to indicate why it may not have pursued a direct course to Neatham.  I also hope to have demonstrated that there is good evidence to indicate that the route may have been via Alresford, Bighton and Medstead.  It must now fall to archaeologists to prove or disprove the hypothesis.  As was stated at the beginning, there has been an upsurge of interest in this road in recent years, resulting in a number of excavations.  Some of these excavations add support to the theory, as I have demonstrated.  Others may support different routes.  Of course, a road through Bighton does not preclude the existence of other minor, or estate roads in the area.  Nor does it preclude the possibility that different routes existed at different times during the Roman occupation.  However, it seems extremely unlikely that there could have been more than one major, Roman public road linking Winchester with Neatham at any one time.  It may be that after years of finding no road at all, we are about to find evidence of more than one.  If so, it is the discussion as to which is the “main road” that this article seeks to encourage.