THE BLACK MUSEUM

 The Black Museum

When maps go wrong

Throughout this work I have pointed out many examples of where maps have gone wrong, usually through omission of roads that ought to have been shown, or the inclusion of roads that never had been built or never would be built. The same applies to railways, and even canals, although it is roads that of course interest the cycling historian. To a large extent these errors are understandable and forgivable. Acts obtained for many roads in the mid-19th century were not followed through to construction owing to the economic threat from the railways, yet these roads found their way onto maps. Conversely, some roads that were built in this period did not appear on some maps for fifty years, as the attention of map-makers was focussed on railways. These deficiencies might have been acceptable at the time: what was not was their perpetuation on maps published well into the cycling era and indeed into the 20th century. One would have thought feedback from users would have put the publishers right, though such information does not seem to have been sought or acted upon. With the publication of an updated Ordnance Survey map in the 1890s there was no excuse for faulty maps from commercial publishers.

One example will encapsulate the situation. Several maps of the early cycling period show a number of roads for which an Act of Parliament had been obtained in 1831 by the Monmouth Turnpike Trust. These were reproduced (for example) on Bartholomew’s 1866 map for A. & C. Black, and on the subsequent ‘Pocket’ series maps taken from it (see map extract of this area on the Bartholomew page). These roads were never built, but contemporary  cartographers, wishing to seem up-to-date, had inserted the projected roads onto their maps. The local Ordnance Survey one-inch map published in 1832 correctly ignored them, but equally could not show the Usk – Chepstow turnpike of 1834 (the present B4235). Consequently a road actually built was omitted from Bartholomew’s (and many other) maps produced decades later. What is surprising is that both the inclusion of the first group of roads, and omission of the latter road, were only corrected in much later editions, well into the cycling era. As one moves north, the definitive one-inch ordnance map appeared later, and so such examples of missing main roads tend to be fewer.

Projected roads never built but finding their way onto cycle-era maps include those from Rochdale eastwards to join the Denshaw – Rishworth road; new sections between Ledbury and Cheltenham; from Dunsop Bridge up Whitendale to join the Slaidburn to Wray road over Salter Fell (a route today’s cyclists would greatly appreciate ), and Middleham to Wensley along the south of the River Ure. This last seems to have been a project abandoned after the building of Middleham Bridge across the river in 1830: nevertheless the unbuilt road was still appearing on twentieth century maps! Conversely, some cycling-era maps still showed the clifftop road south from Bridlington to Barmston. This was once the main road to Beverley, but lost its battle with coastal erosion  around the 1870s, its course now lying some way out at sea. This did not stop the road being listed as the Bridlington - Beverley route in the journal Cycling’s Pocket Road Book, 1907. Even more strangely, it was marked on Bartholomew’s 16m to an inch Motoring Map, which only saw publication in 1907, and which carried on indicating the vanished road on some (but not all) editions through to the 1960s. 

This section is therefore confined to cases where errors which should have long been corrected culpably escaped correction and lingered on cycling maps, plus some cases where maps were just plain wrong. 

Errors of Omission

The republication of the Ordnance Survey one-inch maps in the 1890s gave other map producers a solid foundation for new cycling maps, or updated revisions of their older maps, but for later road schemes commercial producers generally obtained their information from the highway authorities or other sources directly, so in the twentieth century these commercial mapmakers were often ahead of the OS in showing new roads. However, some roads, particularly those of moderate scale, were to slip the net, and not be shown until many years after completion. This was particularly true of maps ‘new’ in the 1890s, taken from contemporary Ordnance Survey sources, which were lax in updating for new road schemes. To take some examples from the W. & A. K. Johnston series, the transporter bridge at Newport (1906) did not appear in their maps and atlases until after WW2, while the transporter bridge at Middlesbrough (1911) never appeared at all. But the most glaring omission from the Johnston maps was Tower Bridge in London (1894), along with its approach roads, omitted from both the long-lived 3m to an inch series and the 1951 half-inch map. A more commonplace example arises in the Forest of Dean, where in the early 1900s a new north/south route was completed, linking Lydney on the Severn with Lydbrook, on the river Wye. This later became the B4234, but coming shortly after the local OS map was issued it was overlooked for some decades by various publishers. 

Roads that Once Were

The earliest cycling maps to be advertised as such were simply the publishers’ usual products, rebranded, and sometimes with a smattering of cycling-related detail added. They all suffered from the same deficiency: the road network of the 1840s (or earlier) was shown unchanged. But in a few cases, former main roads had fallen into such disuse and disrepair that they could not be considered suitable cycling routes. Nevertheless, it took a long time for some map-makers, and publishers of road books, to conform to reality. 

The assumption that any main road that existed before the railway age would require marking as such on a cycling map is best encapsulated by the ‘Stokesley to Whitby High Road’, across the moors of North-east Yorkshire. Virtually all the old Georgian-era maps showed this as a main road, though it was little more than  a packhorse route, albeit the shortest route from much of inland Yorkshire to the then important port of Whitby. Turnpiking of the road via Guisborough, and the advent of railways, robbed it of any importance. There were various tales of woe from cyclists who had been misled into using it, and journalists fulminated against ‘cycling’ maps that marked it. Rather embarrassingly, it somehow found its way into the 1897 CTC Road Book, though with the caveat that parts were being lost to the heather. In fact this only applied to the two-mile section over Danby Moor. Some improvement was subsequently undertaken, as it was marked throughout as a motorable route in the 1920s and 1930s, but this central section, never being tarred, was gradually dropped from motoring maps. The Danby Moor section is still a public road, signed as ‘unsuitable for motor vehicles, though of course cyclists can still get through. 

Another case of a road fallen unto disuse is that of the Sanquhar to Muirkirk road, in the Scottish lowlands. The ironworks at Muirkirk flourished towards the end of the eighteenth century and a new turnpike road was built from Strathaven to improve links to Glasgow. To improve access from the south, and also hopefully bring some paying custom to the Strathaven – Muirkirk section, a new road was constructed over the lonely moors to Fingland and Sanquhar, on the Dumfries road. This was built c. 1800 and advertised as forming the shortest route between Glasgow and Dumfries, which technically it was, but others were unarguably better, especially after Telford's rebuilding of the Carlisle road.

One of the industrialists with interests at Muirkirk was John Loudon Macadam, who owned a tar works and carried out some experiments on road construction here. The Sanquhar road may therefore have a wider historical interest.

By 1850 the Muirkirk to Sanquhar road was effectively obsolete, as railways had taken all the traffic. For the cycling era, it therefore falls within the category of ‘Roads that Once Were’. The Ayrshire section was formally dropped as a publicly-maintained highway in 1884 but had long been abandoned. The early Contour Road Books describe it as ‘grass-grown and disused’ and, if shown on maps at all, was indicated as a track or footpath, the only clue to its former status being its careful winding course along the hillsides to ease gradients. However, it is included here as it still appeared on two small-scale maps aimed specifically at cyclists – G. W. Bacon’s Cycling Map of Southern Scotland, (and one version of the Bacon Cycling Road Map of England), and the Bazaar Tourist’s Route Map of England & Wales. The latter had begun life as John Cary’s Reduction of his Large Map of England & Wales, the former could trace its origins back to its inclusion in Paterson’s Roads, rival road book to Cary’s New Itinerary. Mutual claims of plagiarism between the two parties had led to indecisive court action in 1800, so the competition may be said to have continued posthumously a century later.

What I have termed the Bazaar map was an elderly map adopted, but not adapted, as a small-scale cycling map of England & Wales, together with southern Scotland  It first appeared in this form - “especially prepared for the use of Bicyclists” -  in 1876 and passed through various hands, eventually being published by Bazaar, the proprietors of what was to become better known as Exchange & Mart. Apart from a handful of railways added in the 1830s the map was unaltered from its first appearance in the 1790s. Spare space around the map was taken up with about thirty ‘Tours’, actually routes radiating from London, including one to Glasgow via Dumfries, passing through Muirkirk. The routes were compiled, and allegedly revised, by H. Hewitt Griffin, but were simply cribbed from old Georgian route books and distances were similarly lifted, not always accurately. The map shared publishers with H. H. Griffin’s Bicycles of the Year published throughout the 1880s and which could be relied upon to give the map a plug – “indispensable”.

Incredibly, the Muirkirk to Sanquhar section was still included up to the demise of the map.  The road also lingered on the equivalent Bacon maps and route itineraries. If I had a category for Forgettable Fossils both maps would be in it.

Roads that Never Were

Until the 1850s the turnpike road system was still expanding, but not all proposed routes made it to completion. As a result, many ‘ghost’ roads were to appear on maps of the period, at the time a pardonable sin, but the retention of many such roads on maps sold decades later was less forgivable.

A glaring example of a mapmaker not correcting such an error is the Redditch – Feckenham - Pershore turnpike road proposal of 1826, in Worcestershire. This road was never built, though the southernmost section, from Flyford Flavell to Pinvin, was shown on the 1831 Ordnance map, which suggests some land purchase or preliminary work must have taken place. At that stage of the Ordnance Survey there was no expectation that the map would require updating for many years, nor financial allowance for such revision, so the advance inclusion of nascent schemes would be justified. An isolated piece of hedge near Flavell on the intended road alignment seems to be the only trace of anything actually carried out.This southern section was still shown on Bacon’s County Map of Worcestershire in the 1920s, not bad for a road proposed but never built a century before The whole road through from Redditch to Pinvin was, however, shown on the “new” 1866 Bartholomew maps for A. & C. Black, and was still to be found on various of their Pocket Series maps on sale thirty years later – by which time it was duly coloured red to reflect its main road status! Oddly, the one bit of the road covered by the Act that was built – Pershore Road, Birmingham – never made it onto the Bartholomew maps (or Pershore). 

North Devon seems to have been a hotspot for aborted road schemes and the following map extract shows a number of such projects. It’s taken from W. H. Smith’s North Devon, revised to about 1890, being another extract of Bartholomew’s quarter-inch mapping – not that Bartholomew were any worse than other publishers; just that their maps are handiest for illustrating examples. Of roads (not to mention railways) shown on it but never built are

• Okehampton – Monk Okehampton – Torrington – only North Road, the exit out of Okehampton seems to have been completed, superseding the old ridge route (now Tarka Trail)
• Hatherleigh – Beaworthy and on to Launceston – possibly in conjunction with land enclosure, as zero prospects as a turnpike. Shown on a variety of early maps: the early OS shows a through route of sorts, but parts are omitted from the modern map.
• New road to Chulmleigh, bypassing Chawleigh – Cary’s map of 1794 shows Chulmleigh as being on the only road from Exeter to South Molton, Barnstaple, Bideford etc, but later turnpike development left it largely sidelined.

These abandoned proposals were all shown coloured as main roads suitable for cyclists. Another fictitious ‘main’ road shown on the map but not this extract, was a new road between Wiveliscombe and Bampton, on the Somerset border. Other uncoloured roads, presumably still regarded as projected rather than completed, but never proceeded with, include

• Realignment of a section of the Petrockstow to Torrington road.
• A new route from Coleridge and Winkleigh northwestwards to Torrington – probably an early draft of the route of the present-day A3124, which takes the only practical line. This opened c. 1829.
• Cutting of the bend near Portsmouth Arms railway station (unbuildable!)

All the above errors were not eliminated until the reissuing of the Bartholomew quarter-inch maps in 1897.

Roads That Nearly Were

A variant on the above category are those paths and tracks which at various times have been suggested as worth upgrading to roads for wheeled traffic, specifically those which have erroneously appeared as such on maps. One example in the Lake District is the Sty (or Stye) Head route, linking Borrowdale with Wasdale. Bacon’s Lake District map ‘for cyclists and Tourists’, circa 1893, shows this as a main road, and although from the earliest days of cycling bicycles have been dragged over it, it has never been more than a bridle path. 

It should be noted that there is no riding or driving track over Stye Head, and we hope there never will be. We once met a man carrying a bicycle over the pass. His map had betrayed him into the belief that there was a goodly road. He soon found his mistake, and bitterly bewailed the warmth of the day, whilst we were left to bewail the extreme warmth of his language. But even his case was not quite so bad as that of the young husband who spent his honeymoon cycling with his bride in the district, and had to carry his wife's machine as well as his own over the pass. That was, indeed, a severe introduction to matrimonial felicity.                        -  Leeds Mercury, 28 May 1913

It is likely that the Bacon map was the cause of the above misconceptions. There were campaigns to have the route made into a proper road at various times, most notably in the 1890s (as one for horse-drawn traffic), 1910s and the 1930s, in fact the idea has never completely gone away. The earlier proposals received a fair bit of support from cyclists, though once it became plain that motor traffic would mar the tranquillity of the route beyond repair that stance was to change. Quite how the path became a road on the Bacon map is uncertain: I can’t imagine any contemporary cyclist branding it as a recommendable route other than very tongue in cheek. Perhaps the publisher misread contemporary proposals for its transformation as having been carried out. Whatever the reason, it was a first-class blunder which might have had serious consequences. This four-mile section of bridle path has always attracted cyclists as the road alternative is extremely circuitous, but involves two or three hours of hard and hazardous work, indeed dangerous for the unsuspecting.

Also shown as a ‘main road’ on the full map is the Gatescarth Pass, between Haweswater and Longsleddale: this was once practicable for light carriages but was never a through route for wheels. Both this and Sty Head were left uncoloured on later editions of the map.

While commentating on Lake District roads, the two main road books at the start of the 19th century – Cary’s and Paterson’s – both included a route from London to Whitehaven via Cartmel, Hawkshead, ‘Wastdale’ and Egremont. No other intermediate points were given between Hawkshead and Wastdale, but coincidentally these bitter rivals both gave the distance as 12½ miles, only fractionally over the crow-fly distance. I will leave it to the reader to speculate on how this route found its way through the fells. Fortunately, this was one route not carried over to later cycling road books.

The Blind Leading the Blind

In the 1880s and 1890s reducing costs for introducing colour to maps resulted in most publishers attempting to highlight those roads best suited to cycling. There tended to be two approaches: highlight in colour (usually red) either all main roads (predominantly the old turnpike network), or just pick out those roads identified by correspondents, or from published articles. The trouble with the latter system was that the highlighted roads tended to be, at least initially, very few in number, and were based on perhaps one or two rider’s experiences. Moreover, I can only conclude that these riders were sometimes themselves lost, probably without realising it, and the routes as supplied to the publishers in good faith were not those actually followed. This is the only explanation I can give for many of the routes highlighted, some parts of which simply didn’t exist. Some of those in the Yorkshire Dales are described in the G. W. Bacon section. But the quarter-inch maps of John Bartholomew & Co., a firm whose products are traditionally regarded as the acme of cycling maps, included some shocking errors.

Probably the oddest route ever to be shown as a ‘cycling and driving’ road is that illustrated below, between Wensleydale and Swaledale over Abbotside Common, in what is now North Yorkshire. It appeared, fortunately only briefly, on the Bartholomew quarter-inch map of c.1897 (the new 12-sheet series).

 The likely provenance of the route arose as follows:

1857: OS issues 6” to a mile (1:10,560) County map of the area. Various tracks, annotated as footpath, marked across the high ground.

1860: One-inch map issued, derived from the 6” map. The above paths indicated by faint ‘scratches’ through the heavy hill hachuring covering all the route.

1866:  John Bartholomew produces his quarter-inch map of England & Wales for A. & C. Black. This was based on the OS map, with nearly all the above tracks retained and indicated as roads. This level of retention was itself unusual, as in reducing from an inch to mile to a quarter inch a considerable amount of detail must of necessity be discarded. Such paths would never appear on a map of this scale today, but maps of this scale were used by walkers, anglers etc well into the 20^th century.

1895: The OS 6” map is revised. The path on the Swaledale side is now largely omitted.

1897: Like most commercial map producers, Bartholomew are eager to identify a network of cycling routes. Somehow, a most improbable route is identified over Abbotside Common, whilst the regular road routes between the two valleys, such as the Buttertubs Pass from Hawes, or those from Askrigg, are ignored.

1898:  New Series One-inch map published, reflecting the 1895 resurvey and  omitting most of the paths covering the route.

1910:  Revision of 6” map published. Even less of the old path network indicated.

 I don’t believe any cyclist ever had actually taken the route indicated prior to the map’s publication. I doubt if any cyclist, attempting to follow the highlighted route, got more than a few yards before realising the wrongness of the map. For one thing, the old paths crossed private grouse moors, streams and peat-hags and have never been rights-of-way (and of course are not marked on the modern Landranger or Explorer maps). As can be seen from the above history, they had fallen out of use by the time this map, with its improbable and impracticable ‘cycling’ route indicated, appeared.

How, then, did it find its way onto the Bartholomew map? There are a number of possibilities, the two most likely being:

• A lone cyclist who was an acquaintance or trusted correspondent of John Bartholomew, and who was using an older copy of the map, crossed from Swaledale to Wensleydale via the Oxnop Ghyll road to Askrigg, a ‘proper’ road and one now tarred. He sent details of the route he thought he had taken, possibly as a sketch, to the publisher, who eagerly coloured it up and issued the revised map.

• Someone at Bartholomew read an article in one of the numerous cycling journals of the period describing a crossing between the two valleys, and, misinterpreting the route actually followed (most probably the Oxnop Ghyll one), added this wholly impractical one to the network of ‘Driving & Cycling Roads’.

It might seem incredible that a road might be deemed suitable for cycling (let alone carriage-driving) on the basis of just one rather vague input, but at the time competition between map producers in the cycling market was high, and the addition of a new route to your map would be regarded as much in the way as a newspaper ‘scoop’ would be.

Another possibility is that the route was deliberately added by Bartholomew, safe in the knowledge that no one would actually try to use it, as a test to identify plagiarism by other map producers.

All the foregoing might seem rather excessive to cover one route on one short-lived edition of a map, but I think it useful to give a mechanism of the way in which a general-purpose map had to be rapidly adapted for the needs of cyclists. In 1901 the Harrogate sheet of the Bartholomew half-inch map was to appear, which despite being at twice the numeric scale swept away much of the misleading detail.

 

Just Plain Wrong - Bridges, Fords and Ferries

On scales of one inch to the mile and greater it is generally possible to indicate a bridge by a conventional symbol, and to indicate fords and ferries by naming them as such. On smaller scale maps this was not always possible, and map-makers, cribbing from earlier maps, often had to guess what form a water crossing took. The big map producers were based in cities such as London and Edinburgh, so would not have local knowledge to rely on. This led to errors being perpetuated, or maps being rather vague on what form a crossing actually took. I have mentioned elsewhere the Shian Ferry, over Loch Creran in Appin, which remained wrongly-placed on A. & C. Black’s map of Scotland for over sixty years.

One good example of Victorian mapping mistakes that must have misled cyclists is at Wroxeter, in Shropshire. This is the site of a Roman town and lies where a branch of Watling Street crossed the nearby River Severn and led southwest towards Wales. Although there may have been a bridge here in Roman times, there has only been a ford ever since, and even that would have been impassable for many months of the year. Certainly, the crossing has never been practical for cyclists. This didn’t stop a succession of mapmakers showing, or at least hinting at, the availability of a bridge and marking the old Roman road first as a main road and later as a cycling route. The old Ordnance Survey one-inch map, if studied closely, shows a ford; but the 1890s quarter-inch map derived from it shows a bridge. The extract below is taken from Savory’s County Map of Shropshire, published in 1895, using a Gall & Inglis base map originating with John Cary about sixty years previously – sufficient time, one would have thought, for someone to have pointed out the road led nowhere.

The reverse error was shown on the same map at Cressage, just a few miles downriver. A toll bridge had been erected here in 1801, but the map doesn’t specifically indicate this, implying only a ford. The bridge here held a minor place in cycling history, as there was a test case in 1903 on whether the proprietors could charge sixpence for a cycle, assessing it as a vehicle. Like many tolls which had been inaugurated before the cycling era no specific toll had been specified for cycles: indeed some cyclists claimed this to mean their cycles were toll-free. Victory went to the CTC who had taken up the case and the cyclist again paid the same penny toll as a pedestrian. The verdict was dependent on the wording of the Act authorising the bridge and so was not applicable to all such cases. The present bridge is toll-free.

As well as wrongly indicating bridges, the presence and status of ferries was often poorly indicated on maps. Here I am not dealing with the major ones, such as Greenwich on the Thames or Queensferry, on the Forth, but the innumerable minor ones once commonplace on nearly all our rivers. In the nineteenth century ferries greatly outnumbered bridges on commercial waterways such as the Severn and Trent, ranging from ad-hoc ‘one man and a boat’ services to ones established by Royal Charter or Act of Parliament which stipulated a certain level of service. For mapmakers, keeping track of the constantly-changing availability of ferries was impossible and in most cases they seem not to have tried. The Contour Road Book of Scotland provided some scant details of the mainland ferries, but struggled to obtain details. Nowadays, of course, you just find the app!

An example of the difficulties encountered by mapmakers and users is provided by looking at the River Severn through Gloucestershire. In 1800 the only road bridges were at Upton and Gloucester itself. Turnpike road improvements led to the construction of Mythe Bridge, Tewkesbury, and Haw Bridge, on the Cheltenham – Ledbury route. But into the 20th century there were still about 15 ferries, mostly of only local importance, operating across the river within the county. All of these have withered away. From a map-user’s point of view, the showing of roads – quite appropriately - up to each riverbank at these points gave no clue to what form of crossing might exist, but on many maps the road was taken straight across the river. It may be supposed that, in most cases, there would be a ferry of some sort, though it might involve ringing a bell, or waving a flag, or shouting, to attract the ferryman from the opposite bank. At times the river might be too high, too low, or the ferryman too drunk, to cross, and the unfortunate cyclist would have to pedal round to where would hopefully be a bridge. 

On the Severn above Gloucester, the 1927 Bartholomew quarter-inch maps clearly show bridges at Ashleworth and Lower Lode, Tewkesbury, despite this map being derived from the firm’s half-inch map, whereon the crossings were clearly named as ferries. These errors were quickly corrected. The Philips’ 3m to an inch series and derivatives showed a bridge at Ashleworth, and some versions even marked this as a ‘secondary motoring road’ despite the ferry not carrying motor vehicles.

Near the mouth of the Severn the two ferries – the Old Passage, from Aust to Beachley, and the New Passage, from Pilning to Portskewett, were killed off by the opening of the Severn railway tunnel in 1886. This led to some vagueness in the new road books appearing for cyclists, as they re-parroted the old descriptions and distances from earlier works. Not until 1926 was the Old Passage ferry reinstated and this lasted until the opening of the Severn Road Bridge in 1966. This carries a cycle path.

Water Water Everywhere…

 

I have received Sheet No. 9 of the new Ordnance Survey ½-inoh map, large sheet series. It is called “Leeds and Bradford,” and shows these Siamese twin cities almost in the centre of the sheet. It is more useful, therefore, to the cyclists and tramps of those places even than Bartholomew's beautiful maps, which show Leeds and Bradford either the northern or southern limit of two sheets . . . It is a particularly fine map, and costs 2s on cloth. Bradford people will rightly object however that it not up to date in Upper Nidderdale. Where is their great Gowthwaite reservoir, that has diverted roads and drowned the river? Where is the municipal railway to Lofthouse? Bartholomew’s map, published six or seven years ago. shows great lakes higher up the valley that will not be there for many years to come. The new Ordnance Survey ignores one over 2 miles long, which was completed 7 years ago. Which is the greater sinner? 

'Kuklos', Daily News, London, 14 Sept 1907

The above encapsulates the different approaches of the Ordnance Survey and commercial map producers to the updating of the maps, at least in the earlier years of the 20th century when such updates generally entailed re-engraving. The former revised the one-inch map at pre-set intervals: changes were then carried forward to their smaller-scale maps when these were due for revision. These meant that OS updates were accurate, but delayed. Gowthwaite (Gouthwaite) reservoir had been completed in 1901 after eight years abuilding, so shouldn’t have come as a surprise to the Ordnance Survey. Their commercial rivals were more gung-ho, preferring to get new features quickly onto their maps, and if these were perhaps a little inaccurate, then so what? As seen in the section on OS quarter-inch maps, the 1921 edition still omitted any indication of the City of Bradford’s Nidderdale reservoir schemes though the further reservoirs at the head of the valley were then well underway, and the map was still showing the road between Middlesmoor and Horsehouse crossing the valley at Haden Carr. By contrast, the Bartholomew map initially showed just one long reservoir, severing the old road, with no indication of any alternative crossing. Under their 1890 Act, Bradford Corporation were empowered to close or divert rights of way, but in the latter case replacements had to be provided beforehand. By 1910 Barts was showing only the upper (Angram) reservoir as completed, but the road route severed by the other reservoirs under construction (a never-to-be-built reservoir had meanwhile appeared on their map in the adjacent Colsterdale). Later editions show, correctly, the separate Scar House and Angram reservoirs (there were initially to be three), but no continuous road route across the valley. This is surprising, given that the route had appeared as one of the few cycleable roads marked on Bartholomew maps in the 1890s. In fact, the right of way now crosses the Scar House dam and cyclists of a determined nature have always been able to traverse the old road over to Horsehouse. Construction work was wound up in the 1930s, but the modified route never appeared explicitly on the Barts map, nor for that matter those of other publishers.

___________________________________________________