I am mostly concerned in this post, with the start of the branch at the Speen end. There is indecision about road 41 to begin with. Even on the latest 25,000 series O/S map, there are two competing alignments for the Roman road (see just under Wickham). Further, there is no indication on the modern O/S map of either the point of the junction or the direction taken after this (Indeed no O/S map information exists until west of Marlborough). To me, it is unclear how this road from Marlborough connects to the ‘main’ Roman road 41 at or near Speen. Did it go through Hungerford or is there a cross country route? Margary thought it was cross-country – and he had a reference to depend on.
This, the main western Roman road, branched from the Speen-Gloucester road (41) at a point 3/4 mile to the south of Wickham, but for the first mile there are no visible traces although the alignment is certain.
His confidence is due to the reference at this point to Berks, Bucks & Oxon Archaeology Journal vol 29, page 232. Remarkably, this page is now on-line and can be seen here: Berks, Bucks & Oxon Arch J. vol 29 page 232. (Thank you, all at ADT (University of York)). It turns out to be a paper written in 1925 by P. Williams entitled ‘Roman Roads of Berkshire’.
But there are difficulties with Margary’s description. How does one measure “3/4 mile south of Wickham”? And where does such a measurement start anyway? At the junction, the church, or another central unit? But we need to see if it is certain the alignment exists.
So what did Mr Williams write about this road junction?
At Wickham the branch Road to Bath must have diverged at about an angle of forty-five degrees, but its beginnings have not been discovered. The Ridge has been found in Orpenham Copse and at Elgar’s Farm: it becomes unmistakable through Three Gate Copse, and in the next dip, across tillage, it is plain.
Elgar’s Farm cannot be located on the 25,000 O/S map. The first point which is identifiable is Orpenham Copse. It should also be noted that embankments in the woods mentioned by Williams are currently labelled as ‘Roman road’ by the modern O/S map. So working backwards from these, an extension of this alignment would join road 41 about the site of the modern B4000 (which is here shown on the map as being yards to the side of 41 at this point) and a side track going south-west. (Not “3/4 mile south of Wickham”). The grid reference for this point being SU402713. Interestingly a road to the north east – to Easton – is aligned to this fork in the road, and could be a continuation of 53, so this could be a crossroads. A note on place names: Easton’s name is because it is east of Wickham, which would have been the main focus here. To Roman road hunters, the name Wickham is important: its name is from vicus, a small administrative land unit (or town). In the present age, Wickham is but a village of few houses. Its church (St Swithins) stands 200 yards from the modern road junction, away from modern habitations, but very close to the old route of Roman road 41. Significantly, it has an Anglo-Saxon tower with re-used fabric – including Roman columns. In my experience such ancient churches are constantly placed next to a Roman road.
As there is no clear evidence of route 53 closer than Orpenham Copse, should we look for other possible routes? The junction may not exist where Margary and Williams expect it to be. Are there any other possible ways 53 could connect with 41? There is an unclassified road leading north in the neighbourhood of Orpenham Copse, that takes one to the centre of Wickham, via the church. Along some of its length is a parish boundary. It’s not straight by any means, but the junction with the main road would be dignified by the old Roman settlement of Wickham, again, close to the site of the old church. I think this possibility should be considered.
(Both these routes are shown as dashed lines in the map at the head of the post).
So we have seen where the junction may be placed, and where 53 starts its course. Where does Mr Williams see it continuing after Radley Farm?
The Road evidently passed through Radley Farm House; it is clear through Stibb’s Wood, on emerging from which, on the West side, there is a steep drop on to the lane in Radley Bottom. The Ridge is here very clear, sloping diagonally downwards on untouched grass, and where it crosses the valley there is a plain hump in the road surface. The Ridge can be seen in Heath Hanger Copse, and it continues as the North fence of Oaken Copse. Then it is carried on as the South fence of Great Hidden Farm, and by Peaked Lot, to the main road at Memorial Cross.
From this point no traces have been found, but it re-appears again near Marlborough.
But Margary has been able to add more information. He says
From this point the course goes across cultivated land, and little trace remains for some miles, but signs of the flattened ridge are perhaps visible just west of the Chilton-Foliat road and west of Cake Wood.
We can see on a map that this proposed route takes the road directly down the steep sided valley of the Kennet, across a marshy floor (where no bridge or ford exists) and back up a steep slope the other side, then to Cake Wood on the summit. I don’t see any evidence of this route now. Could we look elsewhere? And would it be possible to backtrack instead, from the known route of the Roman road along this stretch?
The nearest identifiable traces towards Marlborough are in Hen Wood. But this wood is large, and Margary isn’t clear about the exact route through it. So we are stuck for now, with more work to do in defining the exact route.
So in conclusion, Margary can be found to be inaccurate, sometimes vague. However, there is an opportunity to clarify and amend in detail the routes that he first collected together for us to explore.