The northern part of the A338: not a Roman road!

I have not blogged for a while!...

In this photo taken in February last year (2024), I am standing on a short section of the former A338 (realigned in the early 1970s?) just outside the Oxford Lane entrance to the village of Grove in Oxfordshire. The view is looking roughly northeast towards Oxford, some 15 miles away. The alignment of the present-day A338 can be seen beyond the petrol station. A number of sources — such as the Ordnance Survey map of Roman Britain — suggest that below my feet lie the remains of a Roman road, which appears to commence at this very spot and head towards Oxford.

For many decades, I accepted that the northern section of the A338 between the northern side of Grove and the junction with the A420 Swindon to Oxford road was Roman in origin, because it was shown as such on maps that I had seen in the early 1970s. Later in that decade, I became aware of the scholarly work of Ivan D Margary in identifying and mapping the Roman roads of Britain; and I believe it was he who first proposed that this alignment was Roman and identified it as route number 164.

In 1992 I moved to Grove and began working in Oxford, which usually meant a daily commute to work and back by bus. Needless to say, I became quite familiar with the northern part of the A338 over the period of 30 years and an excess of 12,000 journeys back and forth! There were plenty of opportunities to study the route from the top of a double-decker bus. Over time, I changed my mind that this part of the A338 was Roman, based on what I understood about Roman roads and how they were constructed — and what I could actually see on the ground.

A well-engineered Roman road is built on an agger (a slightly raised embankment) and normally has ditches on both sides for drainage. Its alignment is usually as straight as an arrow, and only changes to avoid natural obstacles, which may be achieved using a number of short straight sections. The northern A338 does have a number of straight sections which gives it a Roman feel; but most of the current route shows no clear signs of an agger, the ditches (where present) appear to be too close to the metalling, and there are numerous inexplicable (un-Roman) changes of alignment and obvious curves!

Subsequently, further reading confirmed my suspicion that assumptions (“conjectures”) had been made about the northern course of the A338 in the mid-twentieth century, and are still being perpetuated today in recent archaeological reports…

The origins of the section of the A338 north of Wantage actually lie in the construction of the Besselsleigh Turnpike in 1771. At its southern end between Venn Mill and Wantage, the turnpike runs to the east of Letcombe Brook. This became the main route from Wantage to Abingdon and Oxford. (In the latter part of the nineteenth century, the section from Wantage Road railway station to the town of Wantage was used by the Wantage Tramway until it closed in 1945.) 

It is recorded that prior to this, the main road to Abingdon and Oxford (the Roman “trade route”) north from Wantage ran from the modern day Denchworth Road in Belmont (where Roman remains have been found); and kept to the west bank of Letcombe Brook, crossing the River Ock near Garford, before heading over Foxcombe Hill to the River Thames at Hinksey. 

Bill Fuller in his series of histories about Grove and Wantage noted that the old road to Abingdon and Oxford passed to the west of old Grove along the track now known as Cow Lane. It surely then ran northerly on the connecting bridleway (also a parish boundary between East Hanney and West Hanney), and crossed what is now the road linking East Hanney to West Hanney. A little beyond this, the route to Garford has been lost as sections were sold to local landowners soon after the new turnpike was built. Similarly, the route from Wantage to Grove cannot now be accurately traced; as the construction of Grove Airfield in the early 1940s, and subsequent housing development since the 1960s have between them obliterated most of the old tracks and roads in the area.

Given that so little evidence remains today of the Roman route from Wantage northwards, then it seems to follow that it was not engineered to be a military road and main highway in Roman times. (The building of the Besselsleigh Turnpike used a mostly new and straighter alignment, in preference to upgrading the old road.) Perhaps the Roman route north was actually a prehistoric trackway, improved and still in use in Roman times (and later) — like The Ridgeway over the Downs south of Wantage?