|
THE ROADS OF WESTOVER Include details of turnpikes and stage coach times. See Journal article, "Turnpikes and Their Traffic: The Example of Southern Hampshire" Meyrick Archive Map: predates Taylor. Doesn't correspond in area of Iford Bridge. Very distorted - difficult to fit to any other contemporarily surveyed map. 18th century travellers' accounts of journeys across the Heath
The oldest surviving maps of Hampshire and Dorset identify an area of heathland between the Bourne and Stour rivers as “Sturfeld Heathe” (see Saxton's 1575 Dorset map left). Earlier mention of "Styrfeld" occurs in the charters of Christchurch Priory (1331), and the hunting ground of "Sturfeld" is mentioned in the patent rolls of King Henry VII (1486). Stourfield House and the Stourfield Estate are of later, eighteenth century origin, by which time most maps described the area as Poole Heath. Virtually the whole of the Heath has now disappeared beneath the expanding town of Bournemouth. The history of the roads across the Heath is of course intrinsically connected to the history of the settlements in the area. It is not until the 18th century that the layout of roads is indicated on maps of the area, and suprisingly few of the 18th century roads seem still to be in use at the beginning of the 21st century.
Poole did not develop as a major settlement until long after the Romans had withdrawn from Britain, and most of the roads and trackways that subsequently developed between Poole and Christchurch bypassed the plank over the Bourne Stream at what is now the centre of Bournemouth, which did not begin to develop as a town until the 19th century. When the Romans arrived in 43 AD, they found the area occupied by a tribe known as the Durotriges. Based on the location of archaeological finds, it has been suggested that the main inland track in this area at this time, a ridgeway, ran westwards from its Stour crossing between St. Catherine's Hill and Townsend, through Moordown towards West Howe. During the 1930s, when the Haddon Hill Estate north of Queen's Park was being developed, it seems that perhaps something was known of this trackway, because part of what is now Bradpole Road was named on the Development Plan as "Ridgeway Avenue". (Click Here to see the possible route of the neolithic track overlaid on a modern map) Numerous burial mounds indicate that there may well have been a network of other tracks across the Heath, but it is thought that the main trade route was the Stour itself and the port at Hengistbury. The Durotriges did not give up without a struggle, and so the Romans constructed a fortress at Lake Farm near Wimborne, from which they conducted a war of conquest that may have lasted as long as 10 years. The Romans also built a road, which has been well documented, to connect the fortress with the port at Hamworthy, west of modern-day Poole. It has long been assumed that a Roman road existed through what became the New Forest, but none has yet been identified. However, on his 1759 Hampshire map, Isaac Taylor makes an intriguing reference to a "Roman Way from Wareham to Christchurch" across Stourfield Heath. Confusingly, the map's dashed line seems to reach the Stour and then follow the river upstream to West Parley. Elsewhere on the map, Taylor uses dashed lines to denote forest perambulation boundaries, but this particular line follows no such known boundary, indeed it passes through Iford tything. It seems to indicate a physical feature that Taylor was able to plot in relation to tracks that were then in use.
The Heath was in Hampshire until the county border was redrawn to the east of Christchurch in 1974. Taylor's Dorset map of 1765 does not indicate any continuation of the supposed Roman Way, but on John Bayly's 1773 map (see below) it can nevertheless be picked out from among the other roads by reason of its characteristic series of straight sections. From this and later maps it is evident that large parts of the Dorset section of the "Roman Way" do seem still to have been in use as late as the early nineteenth century. It has to be questioned whether what Taylor saw was really a road of Roman origin. Clearly he saw something resembling a road along a clearly defined route when he made his survey, but we have no way of knowing what led him to believe that it was of Roman origin. However, we may reasonably assume that Taylor knew something about Roman roads, as he identified a number of other such roads on his 1759 map, including the Ickneild Way and the Port Way uninterrupted between Grateley and Hannington. Both of these roads are accepted as of Roman origin. However, no evidence of Roman occupation of Westover has been identified, except the port at at Hengistbury. Christchurch and Holdenhurst were Saxon settlements, so it is difficult to see why a Roman Way from Wareham would lead to Christchurch rather than to Hengistbury. In their book "The Christchurch Commons", Sue Newman and Mike Tizzard quote an anonymous 19th century memorandum in the Druitt Collection which describes the Fee Bank in Stocker's Mead as a "road from Priory to Iford: from a quantity of rubble and loose stones found in the centre supposed to be originally a Roman road". The book also refers to the Bank's description in a 1796 map schedule (Malmesbury's Survey) as "a Bank on which the Monks used to walk". This might be the key to the true origin of Taylor's "Roman Way" across Stourfield Heath. It was perhaps constructed in Saxon times or later by the monks of the Priory as an overland route between the Priory and Wareham, and perhaps incorporated at its western end surviving portions of Roman military roads. Since the last Ice Age, the land mass of southern Britain has been gradually lowering. Since Saxon times, Westover has lowered by about a metre. As a consequence, the shape of the coastline around Hengistbury Head has changed significantly, and the best crossing points on the River Stour are unlikely to have been the same in Saxon times as those in use today. Taylor indicates on his map that the "Roman Way" crossed the Stour between Pokesdown and Tuckton. This could have been by ferry or by bridge. We might ask why no evidence of Taylor’s “Roman Way” was recorded during the development of Bournemouth. An answer might be deduced from this extract from a local archaeological journal published in the 1930s, which illustrates the attitudes to archaeology prevalent in Bournemouth before the First World War:
The Way is not as straight as might be expected of a Roman road. The route avoids all steep gradients, suggesting that the road was to be used chiefly as a trade route by heavy carts. A steep hill could seriously reduce the load that a carter could carry, and thereby significantly increase haulage charges. The route of the road was probably also affected by the strategic location of small forts, including one shown on the 1870 OS map near the turn in the "Roman Way" at Pokesdown. If this fort was of Roman or earlier origin, perhaps it also marked the junction of a branch road to the port at Hengistbury. A Speculative Route It is tempting to think that the straight road (the A351) leading north-east from Wareham is Roman in origin, but in fact Taylor shows a road avoiding Lytchett Minster which ran to the other side of Black Hill, close to the line of the railway. Cyril Cochrane states in his book “The Lost Roads of Wessex” that a section of the Roman Way can be seen near Wareham "joining the main [A351] by the bridge at Sandford from the direction of Keysworth Farm". He goes on to say that the Way followed closely the shore of Poole Harbour, and that it was the route followed by the mapmaker Leland in about 1540, when he noted a short cut to Poole by using a ferry across Lytchett Bay. Remember that, in Roman times, sea level was lower relative to the land and so the ferry used by Leland might have been filling in a section of the Roman Way that was by that time submerged. Cochrane goes on to say that a short section of the Way still exists, unmade up, to the west of Upton House, approaching the A35 at an acute angle. He notes that at the time of writing his book in 1969, the short stretch was still known as "the Roman road". From here it appears to have headed slightly south of the hill at Hillbourne, identified on the large scale 1869 OS map as “Mount Pleasant Clump”.
John Bayly’s 1773 map of Dorset (right) is of even more use than Taylor’s map in tracing the western end of the "Roman Way". It indicates that the Roman Way changed bearing roughly at what is now the junction of Waterloo Road and Denison Road. From here it headed straight across Wallisdown to Westover, its next change of bearing being in Charminster, at or near the junction of Green Road and Murley Road. It then seems to have headed across the south-west corner of what is now Queen’s Park to the junction of Holdenhurst Road and Hayes Avenue. The next stretch took it to the junction of Ashley Road and Shelley Road East, from where it headed eastwards towards the aforementioned fort at Pokesdown, at the junction of Oxford Avenue and Alexandra Road. The final section of road shown by Taylor takes it to the Stour near the junction of Iford Lane and Exton Road. Cochrane suggests two sections of the "Roman Way" that can still clearly be seen, but it might yet be rediscovered at two other points. The first is at Holton Heath, between Wareham and Upton, where the land may not have been ploughed in modern times. The second is at Talbot Village, although as part of Slade’s Farm all trace of the Way may have been ploughed away during the nineteenth century. There is no obvious trace of the Way at Queen’s Park among the other bank and ditch remnants. It is sure to have been quarried away from the clay pits that covered Queen’s Park West, but as the main Park was for centuries common land and for the last century has been a golf course, the road may still exist below the first and eighteenth fairways and in the woodland in the south-west corner of the Park. Following the development of Poole, which began in earnest in the 13th century, travellers heading eastwards and to Wimborne from Wareham and the west might still have made use of the Roman road, but most travellers would then probably have turned southwards to Poole. From here the most direct route to Christchurch would have been by way of the coastal route which we now know as Christchurch Road, which leads through the centre of Bournemouth and crosses the Stour at Iford. John Speed's 1611 map of Hampshire shows "Allom house" and "Bascomb copperashouse", which were the alum and copperas mines at Alum Chine and Boscombe Chine respectively. Both products were used for dyeing, and were exported through Poole by the owner of the mines, the Lord of the Manor of Canford. The products may have been taken to Poole by sea rather than by the coastal land route, but the existence of the mines surely contributed to the development of the roads and tracks through their use by labourers.
There were alternatives to this coastal road, particularly for travellers heading towards Holdenhurst. In fact, Taylor’s Dorset map suggests that in his day the inland routes were more important than the coastal road, which although it appears on some earlier maps is omitted altogether from others. These alternative routes may have been trackways of prehistoric origin, or they might have developed after Poole’s rise in prominence. Either way, they ran parallel and close to the supposed Roman Way over much of the Heath, suggesting that the Roman Way had already been out of use for some time. The upgrading to turnpike status in 1759 of the road from Poole to Ringwood via Longham and the introduction of tolls probably had a significant effect on the use of tracks across the Heath. While the turnpike undoubtedly took much of the commercial traffic, travellers wishing to avoid tolls could no longer use the road with its Stour river crossing at Longham, and may have sought routes over the Heath to alternative crossings at Throop, Holdenhurst and Iford. Taylor’s Hampshire map is of great importance to any study of the roads of Westover, as it shows in more detail than any of its predecessors the layout of roads before the Christchurch Inclosure of 1805, which led directly to the closure of many of the ancient tracks and to the development of Bournemouth. Just as the Poole’s development seems to have led to the demise of the supposed Roman Way, so the development of Bournemouth rendered largely redundant many of the tracks across the Heath that bypassed the new town.
THE ROADS OF THE SECOND MILLENNIUM Let us now take a closer look at the routes followed by the roads and tracks through Westover shown on 18th century maps, beginning with the roads that have survived to the present day. As discussed before, during the 18th century the main routes to the east out of Poole were:
A Road from Poole to Holdenhurst Taylor’s Hampshire map shows another route running between these two. His 1765 map of Dorset shows that this route branched off the Ringwood road just north of what is now known as Constitution Hill, and passed through Wallisdown. From there it led across Stourfield Heath via Littledown to Holdenhurst. Taylor’s Hampshire map also shows a road joining this road from the west, just inside the old Hampshire border. One of the striking features of the south-east corner of Taylor's Dorset Map is the apparent insignificance of the coastal road – it is shown on the map to be less significant than the more northerly route from Poole to Holdenhurst. Bayly's 1773 Dorset map labels the route via Wallisdown as "To Christchurch" while leaving the coastal route unlabelled. Morden's Miniature 1720 map of Dorset shows the Poole-Kinson-Ringwood road, but no roads to the south-east of it, and neither do Bowen's 1749 nor Osborne's 1748 Dorset maps show a coastal road. However, Moll's 1724 and Seale's 1732 Dorset maps do show the coastal road, and not the Poole-Holdenhurst road. Conder's 1784 maps shows one of the two - it's not entirely clear, but seems to be the coastal road. So we have no evidence of the Ringwood Road branches of the Poole-Holdenhurst Road before 1759, but it does seem to have been well-established by then. Perhaps its rise in prominence was prompted by the introduction in that year of tolls on the Ringwood turnpike? Alternative and toll-free routes to Ringwood were to be found via Throop and Holdenhurst, and it may be significant that the Poole-Holdenhurst Road branched off the Ringwood Road just before the toll section.
Ogilby 1675 and Bowen 1720 (shown above) both show a branch to Holdenhurst from the coastal road, just east of Parkstone. It seems likely that the later Ringwood Road branches met this Parkstone branch road somewhere near the old county border, and that this is the oldest of the Poole-Holdenhurst routes. The exact route followed by the roads to Holdenhurst is the subject of some debate, fuelled in part by Cyril Cochrane’s book “The Lost Roads of Wessex”, in which he states categorically that the main road followed the route of present-day Talbot Road, Alma Road, and Richmond Park Road. Examination of map evidence does not support this theory. In “The Growth of Winton”, Sue Lands is more circumspect, stating that “the Medieval track had cut across from [Poole] following the high ridge of the present Wallisdown Road, Talbot Road, and on across Wallis Down, Little Down to Iford and eventually to Christchurch”. In doing so, she acknowledges a certain vagueness between Talbot Road and Littledown, but in other respects this theory too is discredited by the evidence. The map evidence suggests that this was primarily a route to Holdenhurst, and that to reach Iford and Christchurch travellers would have had to turn onto the Wimborne-Christchurch road, known even in the mid-17th century as Castle Lane, (see 1647 survey) after passing Littledown.
In any event, the Holdenhurst routes seem to have fallen into disuse and all but vanished in the very early years of the nineteenth century. None seems to have been still in use by the time of the first Ordnance Survey in 1811. They appear on Cary’s 1804 map of Hampshire (see right), but not on the 1805 Ordnance Survey sketch (see above). By 1826 a new route existed along the line of Wallisdown Road. Dorset Inclosure Acts might shed some light on the reasons for this. We can speculate that the decline of these roads was due mainly to the following factors:
So what route did Taylor’s roads from Poole to Holdenhurst follow? Because of the distorted scale of Taylor’s map, answering this question is crucial to the interpretation of the route he indicates for the Roman Way. Let’s look at each section of the route in turn. An accurate indication of where the main road branched off the Ringwood Road is given by Paterson’s 1785/93 strip map (above), which marks each milepost. Between the 105th and 106th mileposts is a branch road (marked as leading to Christchurch) which corresponds exactly to the present junction of Constitution Hill Road and Ashley Road – at the top of the hill as indicated by Taylor. The coastal road to Christchurch (via Parkstone) is also indicated, along with further routes at Bear Cross (which leads to Christchurch via what is now Castle Lane) and at West Howe. The West Howe branch is shown joining the main Constitution Hill branch just inside the old Hampshire county boundary. These four routes are clearly shown on Milne’s 1791 Hampshire Map, and Bayly’s 1773 Dorset Map shows that there were other tracks too.
Neither the Constitution Hill nor West Howe branches are shown on the 1811 Ordnance Survey map. But some of the more minor tracks shown by Bayly in 1773 are recorded by the 1811 Survey, notably two north-south routes near the old Hampshire-Dorset border, one of which corresponds to the present Alder Road. This suggests that the routes that were omitted in 1811 had perhaps by then fallen into disuse.
Greenwoods' 1826 Dorset Map (above) suggests that Wallisdown Road might have been the route followed by the West Howe branch, but Rowe’s 1813 map of Dorset (right) paints a different picture and suggests that Wallisdown Road was fairly new in 1826. Here the old branch road is shown meeting the Turnpike at “How Corner”, as indeed it seems to on Taylor’s 1765 Dorset map. As there was a settlement at How before 1765, this seems a far more likely scenario than a slightly more southerly road leading from nowhere in particular.
Taylor's Hampshire map suggests that the meeting point of the various Poole-Holdenhurst branch routes was between Quamp Corner and Fern Barrow (although his distorted scale then shows the Constitution Hill branch passing south of Fern Barrow, which would take it implausibly close to Christchurch Road). Midway between these points today is the junction of Boundary Road and Wallisdown Road. Brannon's 1860 map (left) shows a track running south-west from this junction, which stops when it meets a track running south from Talbot Village. Could this be a remnant of the Constitution Hill branch? If so, then it probably ran over land now occupied by Bournemouth University.
On Greenwood’s 1826 Hampshire map (right), the area between the county border and Charminster Road is simply marked as “Wallisdown Plantation” – parts of which were later to be known as “Malmesbury Plantation” and “Talbot Woods”. No roads are shown through the plantation, but there is, however, a short stub road leading westwards from harminster Road at what is now Fiveways, and it seems highly likely that this had been part of the Holdenhurst route. Indeed, the 1870 Ordnance Survey maps show a track linking this stub road to the junction of Wimborne Road and Talbot Road. From Fiveways, Taylor shows the Poole-Holdenhurst Road joining and following eastwards the line of Great Dean Bottom, which is the valley now occupied by Queen’s Park Avenue.
It is impossible to tell whether it followed the line of the present road, but certainly there was a track along this line in 1869, but which stopped when it reached the track from Strouden Farm to Pokesdown. The 1805 Inclosure Award map suggests that the track might once have continued along the side of Hadden’s Plantation, and this could very well have been the route taken by the Poole-Holdenhurst road. This is supported by Milne’s 1791 map (left), which suggests that the road then passed not as close to Littledown House as Taylor showed, but straight on to Holdenhurst, passing the Manor House at Townsend. Cemetery Junction / Rush Corner Acceptance that Taylor’s 1759 map does not show the route of the present Alma Road and Richmond Park Road begs the question, “does the distinctive 'Y' junction shown on the map represent what is now known as Cemetery Junction?” Probably not. The road shown branching north-west off Charminster Road is not what is now (and in this article) known as Wimborne Road - the northern end of which is the road shown further west. It seems that in 1759 Wimborne Road did not continue far south of the Poole-Holdenhurst Road. However, Taylor does show a track leading north-west from where it met the Poole-Holdenhurst Road to Quamp Corner, and the remnants of just such a track can be seen on the 1869 Ordnance Survey map. With no southerly continuation of Wimborne Road, there could be no Cemetery Junction, or Rush Corner as it was known previously.
The 1869 OS map also reveals the disjointed remnants of a track running roughly parallel to and to the east of Wimborne Road, which turns towards and joins Charminster Road at what is now its junction with Alma Road. This track is shown more clearly on the smaller scale 1870 OS map (right), running southwards from Moordown and east of Winton before veering eastwards to join Charminster Road at Heath Farm, and could be the track plotted by Taylor, forming the north-west branch of the “Y” junction on Charminster Road. There can be little doubting the appearance of Cemetery Junction / Rush Corner on Greenwood’s 1826 map, and the 1805 Award Map seems to show it too – but even this is some 46 years after Taylor’s survey, and the relative importance of Wimborne Road could certainly have increased in that time. Maps such as Milne's drawn from surveys in the intervening years show no sign either of the southern section of Wimborne Road nor of the track between it and Charminster Road. Examination of a detailed topographical map might provide further evidence for the theory that Taylor did not show Cemetery Junction, because he clearly marks Boscombe Bottom and a stream leading to it from Charminster Road which might help to pinpoint the location of the junction marked on his map. Modern surveying techniques, including satellite photography and positioning systems, allow the latest Ordnance Survey maps to be plotted with great accuracy. When Taylor made his maps in the eighteenth century, he probably relied on lines of sight, a compass and chains of standard length when making his survey. Inevitably, this means that in order to trace on a modern map the roads and tracks shown on early maps, a simple overlay is not possible. The interactive map which may be accessed by clicking here is an attempt to translate the early maps onto a modern grid. This was done by working backwards in time, initially through several editions of the Ordnance Survey maps, at each stage picking up and adjusting the route of recognisable but poorly plotted tracks, and eliminating tracks that had developed between the stages. There is a certain amount of speculation involved, but in nearly all cases the tracks marked on early maps can be related to tracks shown on the large scale 1869 OS maps. Bibliography: "Dorset Maps", David Beaton, The Dovecote Press 2001 "Old Hampshire Mapped", Jean & Martin Norgate, 2003 "The Lost Roads of Wessex", Cyril Cochrane "Pokesdown and Iford Yesterday", J.A. Young, Bournemouth Local Studies Publications 1990 "The Growth of Winton", S.J. Lands, Bournemouth Local Studies Publications 1976 "The Makers of Christchurch", Michael Stannard, Natula 1999 "The Romans", Bill Putnam, The Dovecote Press 2000 "Roman Roads in Britain", David E. Johnston, Spurbooks Ltd., 1979 "The Christchurch Commons", Sue Newman & Mike Tizzard, Natula 2007 Acknowledgements: Thanks to Bill Putnam for his extremely helpful comments on an earlier draft of this document, and to Michael Stead for introducing me to this subject and providing various maps. © Stephen Gadd 2003 | ||||||||||||||
