Roads, Railways, and Canals: Technical Choices in 19th century Britain
<center>Roads, Railways, and Canals: Technical Choices in 19th-Century Britain</center>

by Francis T. Evans from Technology and Culture, Vol.22, 1981. Scanned by Dannielle Hall for ENAM 484

Between 1760 and 1840 Britain passed from a state of local economies. with poor to middling transport, into a nation with the promise of a national railway system superimposed on a network of good canals and roads. This change has often been examined from an economic point of view, and there are some excellent studies of individual technical elements in the transport system.' The excuse for this further essay lies in the lack of an overall technical comparison of the roads, railways, and canals as they competed during the Industrial Revolution. Here the aim will be to concentrate on the technical factors, comparing the advantages and disadvantages as they were understood by contemporaries. The question will usually concern how the methods of transport were expected to behave rather than how individual railways or canals performed. Once the choice among roads, railways, and canals appeared, investors, engineers, and managers needed criteria upon which to base decisions; and these criteria, however crude, were often technical.

When railways were young, there was not a great difference between advanced technical thought and the information digestible by the interested outsider; transport engineering did not require much specialized terminology or mathematics. Consequently, it is reasonable to base this comparison of popular technical expectations upon commonly available sources, particularly encyclopaedias. Successive editions of Encyclopaedia Britannica and Chambers's, Rees's, Tomlinson's, and Penny encyclopaedias reflected very faithfully the swings of advantage and disadvantage which characterized the alternative transport systems.2 This study examines some crucial points pursued further in Mechanics Magazine, Priestley's Navigable Rivers and Canals, and other sources, but even these are representative of the broad climate rather than of a specialized elite.3 It is felt that these sources reflect opinion more accurately than generalizations based, for instance, on the best contemporary practice or on the multitudinous patents which throw little more than an amusing sidelight on events. The better encyclopacdias gave clear accounts which were kept up to date in successive editions, and their comparisons illuminate the changing fortunes of roads, railways, and canals, particularly in the interesting period when the canals were in technical trouble, amounting almost to a crisis, and while railways were still working toward a coherent technical structure.


In the mid-18th century, the English road system looked complete on the map, with adequate connections among villages and towns. But despite the early turnpikes and pioneering work by Tresaguet in France and John Metcalfe in England, much of the system was almost unusable by wheeled vehicles.4 In many areas, packhorses were the only means of transporting loads. Measurements of the carrying power of a horse varied greatly, but it was never suggested that horsepower was most efficiently employed in carrying instead of pulling loads. The Britannica reported that "the most disadvantageous way of employing the power of a horse is to make him carry the load up an inclined plane, for it was observed by de la Hire that three men, with 100 pounds each, will go faster up an inclined plane than a horse with 300 pounds. When the horse walks on a good road, and is loaded with about two hundred weight, he may easily travel 25 miles in the space of seven or eight hours.''S De la Hire's 2 hundredweight represented a cavalryman and his equipment, but packhorses could carry much more.6 Often they went in long lines, single file, along the narrow roads. In 1739, two gentlemen traveled from Edinburgh to Grantham, only I l0 miles from London, mostly on "a narrow causeway, with an unmade soft road on each side of it; . . . they met from time to time gangs of thirty to forty packhorses."7 Here, and in plenty of other parts of Britain, there was not room to pass, even with packhorses. A road system like this, with parts virtually untouched since the departure of the Romans, made the more efficient use of horsepower with wagons difficult.

Yet underlying these defects was a more fundamental one. The very complete parliamentary act of 1773,8 a major public policy statement on roads and their upkeep, shows the problem clearly by its omissions; it was simply not understood how to make a durable road over which a vehicle could travel easily; and the act was more concerned with protecting the roads against vehicles. It is so complete in other respects that the failure must be set down to ignorance. This lack of knowledge was not peculiar to legislators, since contemporary encyclopedias, otherwise strong on technical matters, betray the same ignorance of road building.9 They merely defined the term "road," describe how the Romans built them, and, in lengthier works, enumerate the major Roman routes.

The general ignorance of the principles of good road building persisted till the century's end, but it was moderated in a number of ways. Although there was no body of doctrine like that held by the engineers of the poets et chaussees in France, individuals like Metcalfe built soundly; and in other places sheer hard work improved surfaces, flattened slopes, or straightened curves. John Harriott's road harrow is a good example of the simple improvements taking place in the late 18th century.l°

The degree of confusion in all this pragmatic concern for roads shows up in the Rees's Cyclopaedia article, "Roads in Rural Economy."" The writer favored a moderately convex road of one-in-twelve camber, since excessive curvature defeated its own object; on highly convex roads, vehicles bore down too heavily on the outside wheel, overturned, or most usually, drove astride the crown of the road. In either case, deep furrows resulted and retained the water which the convexity was intended to dispel. Yet the article also conceded that Bakewell and Wilkes had made great improvements with their concave and sloping roads, which allowed water to flush surfaces clean and smooth. A further confusing factor was the variability of roadmaking materials and of the soils to be traversed. Parts of Lancashire, for instance, had limestone close at hand, which knit together for a good surface. But elsewhere in the county, stone and gravel were not hard enough to support the increasing weights and volume of transport, so costly paving stones were imported from Wales. Turnpike road near Manchester, for example, cost £2,000 a mile to pave. Perhaps the important point was not that a confusing picture emerged, but that the many detailed descriptions of roads in each county showed the scope of the problem and made possible a more fundamental approach. |

In any case, despite the disagreements apparent in Rees, the late 18th century brought a great increase in road travel. "Flying Machines"—coaches with springs—were in service on many routes, l and fast, regular mail coaches were making passenger travel and | postal services a commonplace rather than an adventure into the unknown.'2 Much of southwest England, especially Cornwall, continued to use packhorses, and travellers like Arthur Young had plenty of road horror stories to tell. But that did not stop Parson Woodforde confidently making for Bath or London from Oxford in a day, or even from London to Bath in the same timed G. L. Turnbull has given a quantitative idea of the frequency and volume of the road carrying trade during the Industrial Revolution in Britain. Around 1780, there were 126 carrier departures weekly from Birmingham, 156 from Manchester, and 126.5 from Bristol. More than half of the services from these and other major towns stopped within 30 miles, but each town had a frequent London service—18 departures weekly in the case of Birmingham. This clearly indicates a national network, but the defects of road transport were still there: perhaps 4 tons per wagon load and a three-week round trip between London and Liverpool. A great expenditure of horsepower and time was needed to move a small volume of goods.l4

The Rees's Cyclopaedia article on roads appeared in 1819, the same year as John Macadam's report on his system of road building and mending to the House of Commons. Five years later Encyclopaedia Britannica published a new article on road making by Thomas Young. As might be expected, Young's approach was rational, scientific, and coherent. Beginning from the premise that "the grand object of all l modern roads is the accommodation of wheel carriages," Young analyzed the natures of wheels and roads as a connected problem.'5 Previously, as we have seen, the tendency was to separate the problems of wheels and roads on the assumption that roads were inherently weak and that they needed protection against the traffic running | along them. Young considered the function of a wheel and deduced the need for smooth, hard roads. There followed a discussion of stone pavements and gravel roads and a brief account of the confused state of knowledge revealed in the Board of Agriculture papers, but Young concluded that Macadam's simple and economical system had superseded all these other real or imaginary improvements.'5

John Loudon Macadam himself was well aware of the shortcomings of his predecessors—indeed, the road-building profession had "become contemptible in the greatest degree" and "perfectly adapted to" .. . the most ignorant day labourer," while the system was worked by "surveyors selected from the lowest and most illiterate class of the community." The recent increases in commerce had merely led to great quantities of stone being thrown onto the subsoil, leaving the roads still barely passable. The core of his argument was that "Nothing has been written on the subject of the surface of roads, or the means of making them proper for the easy passage of carriages, though volumes have been published to recommend many useless and many vexatious restrictions on the carriages themselves." Macadam's methods for producing a"strong, smooth, solid surface" have often been described, but his reasoning is worth repeating. A surface of compacted 6-ounce stones would not cause jolting and shaking of the carriage and consequent repercussion upon the surface "which is the real cause of the present bad state of the roads of Great Britain." A firm hand was needed to establish the new standards of road engineering, but until that time, he concluded that "whatever carriages the law may compel men to draw through such roads (for at present they do not travel over them) must continue to act as ploughs."'8

Macadam may have underrated Metcalfe and overrated the durability of his own roads, but he was right in emphasizing the hard, smooth surface. Thomas Young had calculated the resistance offered by a soft road to the passage of a wheel. Normally a horse exerted a pull of about 200 pounds. If a 3-ton wagon with wheels 4 feet in diameter sank just 1 inch into the surface, it required a force equal to at least one-seventeenth, and more likely one-ninth, the weight of the wagon to pull it along. This meant an additional pull of between 400 and 700 pounds—in other words, two or three extra horses. A 2-inch sinkage of the wheel would add half as much again, say three or five more horses.l9 These calculations were borne out by experiment. McNeill found that a wagon which needed a pull of 33 pounds on well-made pavement, or 46 pounds on 6 inches of broken stone laid over large stone pavement, required four times the pull—147 pounds—on thick gravel laid on earth. These and other contemporary calculations offer a rough but useful agreement on how much horsepower was needed on the bad, old roads.20 On a poor 18thcentury road, a horse could scarcely pull 1 ton, while a better road increased this to 2 tons.


The horse which moved 1 ton on land could pull 30 tons or more in a floating barge. This was the great advantage of canals. Britain's long indented coastline made her fortunate in her natural waterways, and from 1750 she began to emulate and then surpass the canal systems of France and Holland. By the 1820s some 2,200 miles of canal had been built and "there was no place in England south of Durham that was distant 15 miles from water communication."22 Canals offered more than a great savings in horsepower. Fragile loads like pottery suffered far less risk as waterborne freight than if they were subjected to the jolting of the roads, and the dense but yielding medium of water sustained weighty loads which could not be carried on the roads. For these reasons the artificial waterways spread, growing straighter, wider, and deeper as capital and profits furthered their expansion.

When we consider the level of hydraulic skill possessed by the canal builders, we appreciate better the magnitude of the technical switch as railways replaced canals. Even in 1800, most engineers were oriented toward hydraulic rather than steam power, and the common steam engine was a massive, slow pump. Watt's rotating engines were only fifteen years old, and compact high-pressure engines were almost unknown. Meanwhile, engineers were employing a developed hydraulic technology.23 At Chemnitz, hydraulically created air pressure forced drainage water up 96 feet from the mines, and Montgolfier's ingenious hydraulic ram also raised water with scarcely any moving parts. Smeaton's experiments in 1753 on waterwheels and French achievements, like Poncelet's efficient undershot wheels, are paralleled by Joseph Bramah's beer pump, flushing lavatory, and hydraulic presser Their sureness of touch and self-confidence can be seen in many of Brindley's works, like Wet Earth Colliery, where he supplied an underground waterwheel from an inverted siphon which took water from the IrwellRiver and passed underneath the river to reach the mine.25

Britain's canal system presented massive opportunities for applying hydraulic techniques. Great canal terminals like Bugsworth had many wharves and feeder tramways, resembling in function the later railway marshaling yards. Constructional feats accentuated the water engineering. Thomas Telford's cast-iron aqueduct at Pontcysyllte stood 120 feet high on its masonry pillars and carried the canal 1,007 feet in nineteen spans.2fi Of the 42 miles of tunnels, Standedge was the longest, at 5,415 yards, and took sixteen years to build. Canals ran miles underground in the Duke of Bridgewater's collieries, where they had inclined planes to pass barges from one level of canal to another. Thus the nuisance of water to be drained from the workings was ingeniously turned into the benefit of a useful transport system. Few canal engineers had such unwanted water near at hand, and their successful completion of a canal depended on a reliable supply of water. They sought springs and streams, built reservoirs, and, as the last resort, sometimes set up pumping machinery to fill their artificial navigations.

Subordinate Railways

The early iron railways were an ancillary to this impressive maturing of British canals. Although, as we shall show, the canals were to reach a near-crisis state through inherent technical problems which proved intractable, the railways which served mineral traffic and fed the truck network of canals were, at first, not viewed as any challenge. The cost of constructing and maintaining canals limited their successful application to the main arteries of trade, and short runs of railway were a useful supplement. Wooden rails had long been used to facilitale mine transport, and when cast-iron rails were introduced at Coalbrookdale in 1767 their use spread quickly—a horse could pull five or six times as much on these as on a common road. However, they were far inferior to canals in reducing traction.

The early iron railways were short, local, and private, so that there was no call for standardization of gauge or pattern.27 South Wales and Shropshire generally preferred tramways, sometimes called plate ways, whose raised edge kept the plain wheel on the track. They were popular after 1787, largely on the grounds that suitable road wagons could also use the rails. At first they were cast iron, and their structural weakness led to fishbellied flanges underneath and a variety of other modifications. The other kind—the edge rail—was earlier and at first was simply a cast-iron bar set on edge. This called for trucks with flanged wheels, although these were unsuitable for ordinary roads. The advantages of the edge rail were that it did not hold stones and dirt or offer as much friction as the flat template with its raised flange. It was soon realized, too, that edge rails formed a stronger load-carrying beam, since their material could be disposed to give greater depth, but even so the plateways persisted in South Wales, the Surrey railway, and some other places. Wrought-iron edge rails began to replace cast iron at SirJohn Hope's colliery near Edinburgh, and in 1820 John Birkinshaw of Bedlington patented his new wedgesectioned rails which he rolled from puddled iron28 (see fig. 1). At first the edge rails only predominated in northern England and most of Scotland, but on a plateway or tramroad, a horse could only pull 5 or 6 tons, whereas on SirJohn Hope's edge railway 10.5 tons could be taken at 4 miles an hour. This was a great improvement, though it still left canals with a comfortable margin of superiority, since on a canal a horse pulled three times what it could manage on the best of the railways.

The new railways were usually no more than 10 miles long, in keeping with their roles as private lines or feeders to the canals. The Hay Railway with only 24 miles was the longest. Their independent public life began with the Surrey Iron Railway of 1803 from the Thames at Wandsworth to Croydon. This was followed in 1806 by the Oystermouth Railway in South Wales. Rails suited the mineral traffic in hilly regions, where canal locks would be too frequent, and careful surveying could save animal power by downward inclines on the loaded run.2'' South Wales soon acquired over 300 miles of tramway.

In the early years of railways, before 1825, engineers were remarkably free to devise alternative systems. H. R. Palmer's overhead monorail proposals claimed the advantages of freedom from difficulty during snow and ease of construction over uneven ground.30 A stone tramway was built into London's commercial road for 2 miles to the West India docks. There were 22 feet of macadamized surface for horses and light carriages; 9 feet of stone paving carried stagecoaches; and a tramway of heavy stone blocks, their outer edges 7 feet apart, was to take the heavy wagon traffic. On this tramway a horse could pull 8-10 tons. The early 19th century thought radically about systems and did not separate ideas sharply into compartments. Sir George Cayley, for example, dubbed his proposals for caterpillar tracks a "new universal railway"—a logical use of the word which would not occur to many modern minds (see fig. 2).

Opinion continued to favor iron rails, and a common argument was that the deeply worn stone road blocks recently uncovered at Pompei showed that even ancient Rome had found difficulty in maintaining such a surface. Apart from all these localized experiments, railways had their broad dreamers too, like Thomas Gray and William James,3' who both called for a national system. But it was still normal in 1824 to see canals as the greatest economizers of horsepower, and Britannica concluded that the railway "is principally applicable where trade is considerable and the length of conveyance short, and chiefly useful therefore in transferring the mineral wealth of the Kingdom from the mines to the nearest land or water communication, whether sea, river or canal. Rees's Cyclopaedia, 1819, put the railways among the canals in its thorough gazeteer of British canals. This, after all, was t e sensible place for a minor system which common opinion saw as a supplement, not competitor, to canals. Twelve years later, Priestley's Navigable Rivers and Canals also put railways in with canals but, as we shall see, had a quite different view of the railways' potential.

Canals' Internal Problems

Although water transport remained the natural choice for a main system until the middle 1820s, serious drawbacks slowly emerged arising from the problems of water supply, the nature of locks, and behavior of a barge in water. As soon as canals needed to change level, some considerable discontinuity in the journey resulted. Brilliant feats of surveying reduced these discontinuities to a minimum: near Coventry there were 73 level miles, Manchester had 70 miles, and there were at least twenty sections of canal with more than 10 miles of level pound. Even the unlikely terrain between Abergavenny and Brecon, where the Welsh mountains begin, had 14 miles free from locks. A lock could easily take 50,000 gallons—over 220 tons—of water. Only one lockful was needed for a pair of boats if they were travelling in opposite directions, but a series of barges following each other up or down needed a lockful for each vessel. The canals found themselves in serious competition with industry which still relied mostly on waterpower, for whose high torque and low speeds many industrial plants were designed.33 To make matters worse, the depth of locks was limited by the great pressure of water, which at 30 feet exerts a force of more than 2,000 pounds per square foot, "but as the lower gate is strained in proportion to the depth of water which it supports, when the perpendicular height of water exceeds 12 or 13 feet, more locks than one become necessary."34

In broken country staircases of as many as thirty locks were built, and the water expenditure—especially in time of drought or when | crossing a watershed like the Pennines—became formidable. Ander- | son, arguing against locks, pointed out that even with good locks only | one boat could pass each ten minutes. "In short, in a canal constructed | with water locks, not more than six boats on average can be passed is an hour, so that beyond that point all commerce must be stopped." -The constrictions imposed by water shortages and the hindrance of locks set a sharp limit on this main transport system of a fast growing economy 36 l

Engineers made a major inventive effort in the thirty years after 1790 to alleviate the water shortage or to do away with canal locks. Their suggestions fall into three broad categories, which take up no fewer than eleven quarto pages of Rees's Cyclopaedia: devices for savings water, methods of lifting boats or loads, and the use of inclintd planes.37 Ordinary locks were vulnerable enough to rough usage, and some of the suggested machines asked too much of a hasty or clumsy canal employee. But, as we shall see, some of the schemes were worthless, even assuming the technical means of achieving them in 1800, and few gave promising results.

A number of methods of saving water were attempted. Where barges varied in size, narrow and wide locks were set side by side to avoid using more water than necessary. Another approach was to build side pounds, which could store the first water drained from a lock; only the bottom portion was passed into the lower level of the canal, and the first part of refilling the lock was undertaken from the l side pounds. Playfair's design used no fewer than ten side pounds, each taking 1-foot depth from the lock, and only one-sixth of the water was lost. There were simpler arrangements, but all added substantially to the time spent passing through locks. Lawson Huddleston and Robert Salmon independently came up with methods of raising and lowering the lock water by means of huge plungers forced mechanically into the lock, so that no water needed to pass into the lower pound; these we may safely list as improbable.

Lifting devices offered a promising line of development, for it was cell within the power of 18th-century technology to lift 20 tons or so (see fig. 3). It was another matter, however, to lift, move, and lower boats the 12 feet between pounds often, quickly, and safely.39 Lift designs were put forward by Anderson (1794), Fussell (1798), Rowland and Pickering (1794), and Robert Fulton (1794). Most lifts used caissons of water which counterbalanced each other even if only one held a barge, and lifting was achieved by letting water out of the ascending caisson. This had the advantage that a barge constantly supported by water did not need to be so strong as one made to be suspended by a few points. A crane inside the Duke of Bridgewater's canal tunnel at the Worsley colliery worked by the weight of water tubs which were filled by an underground stream. The crane had a brake wheel "of sufficient size that a man who stands before the lever thereof has his two hands at liberty to pull the lines which connect ``ith the valves, and give signals to those below, while by lunging or stepping forward with his breast against the lever, he can in an instant stop the machinery."40 The crane moved boxes through a height of 180 feet. Brindley also made the water tubs pump water from the lower to the upper canal when they were not needed for the crane. Apart from Brindley's crane, and an early handworked one on the Churprinz Canal in Saxony, there is no evidence that any of these lifts worked satisfactorily.4' Later, in the 1870s, a good lift was built at Anderton, but the canals needed successful solutions in 1830, and lifts were not among them. Robert Weldon made the most remarkable attempt to supersede pound locks with his "hydrostatick lock," which was tried in 1794 in Shropshire and built in 1797 at Combe Hay on the Somerset coal canal. The barge was floated into a great tubular wooden caisson. When this was shut, water was pumped in and the caisson sank to the lower level where it butted against the exit. The doors were opened and the barge floated out. It is noteworthy that when the brickwork bulged, the sinking caisson experiment was dropped. An inclined plane, and, later, pound locks were substituted. The new locks for the 138-foot rise forced the company to add £45,000 to the £80,000 capital they had originally sought for the canal.42

Inclined planes turned out better than either the water savers or the lifts. William Reynolds made the first successful attempt at Ketley in Shropshire, where boats twenty feet long entered a lock and settled on wheeled cradles, on which they descended 73 feet on rails. This plane could pass 400 tons a day, with counterbalanced loads requiring no outside power. Five more planes were built near Coalbrookdale, of which the largest—the Hay incline—still remains. The Hay plane rises 207 feet in a horizontal distance of 1,000 feet. Instead of a lock, the boat and its cradle are wound out of a basin in the top pound before traveling down. An ingenious arrangement of wheels and two rail gauges allowed the boat to remain horizontal on both legs of the journey. The Hay incline saved building twenty-seven locks and passed a pair of boats in three and a half minutes.43 The group of planes at Hay worked well for as long as they were needed, from 1793 till 1907. Even more remarkable was the underground plane, rising 107 feet, which joined different levels of the Duke of Bridgewater's Worsley mine.44 A few dozen inclined planes were built, and many worked successfully. It seems fair to comment, however, that this sole successful alternative to locks on canals before 1840 was essentially an application of railway principles rather than a true hydraulic solution.

It is easy in retrospect to see that hydraulic engineering was reaching a crisis which did not touch the coal-powered technologies. Water vas a limited resource which gave both power and transport. To the canal engineer it seemed that transport should take precedence over industrial needs, since this would facilitate the transition of industry to steam power by means of easier coal supplies. Rees's Cyclopaedia argued in 1819 that canals and their extension "remove one of the principal objections to steam engines, by enabling new mines of coals to be daily opened, and the products thereof, as well as of the old mines, to be regularly and cheaply conveyed to every situation where engines can be wanted."45 This seems almost a wistful treason against the hydraulic tradition. If steam transport and steam-powered industry had not triumphed, no doubt the hydraulic tradition would halve produced an integrated system to optimize the limited water resources. But the Cyclopaedia further maintained that:

"we would not, however, be supposed to recommend the annihilation of water mills; on the contrary it bath long appeared to us that their number and their power might . . . be increased, and yet all the purposes of canals be fully answered, and those most capital improvements of irrigation and drainage at the same time be extended, to very large tracts of land; for this purpose it would be necessary, that an entire valley of considerable extent, that has a good stream of water through it, as the Colne or the Lea near London for instance, should be put under a system of improvements."46

Improvement meant drainage and the exploitation of springs; the conversion of mills from undershot to overshot, to double efficiency; and better exploitation of waterhead, by transmission in pipes under pressure and siphoning. Had the demon of unlimited fossil fuel not beckoned, perhaps the paradise of living on renewable resources would have come into the future then, instead of now! But all this was said in 1819, and in 1824 George Buchanan could still write that "On some of the railways near Newcastle, the waggons are drawn by means of a steam engine working in a waggon by itself, the wheels of which are driven by the engine. But this application of steam has not yet arrived at such perfection as to have brought it into general use."47

The New View of Railways

If, despite their problems, canals retained their position as the most favored form of bulk transport until well into the 1820s, we must ask how and when they lost this position to locomotive-powered railways. If we seek practical demonstration of the locomotive's effectiveness, then it is almost indisputable that Locomotion on the Stockton and Darlington, 1825, or the performance of Rocket at the Rainhill trials in 1829 were the two decisive events. It is also important to ask, however, at what point the theoretical advantages of railways became manifest as a part of popular technical opinion. When was the ideological victory? It is not necessary in this context to do more than mention the locomotive builders: Trevithick in 1804 at Pen-y-darren; Blenkinsop and Murray in 1812 with toothed wheels and rack rails, but also two cylinders to smooth the power output; Hedley's Puffing Billy, 1813; and finally the series of George Stephenson designs beginning with Blucher in 1814. However, these men did not create public opinion, and our own familiarity with locomotives should not make us forget how alien the steam locomotive was in an age when traction was synonymous with horses. Public opinion was perhaps first alerted by I William James, whom L. T. C. Rolt called "the John the Baptist of railways."48

In 1822 William James wrote: "In comparison with navigable canals, generally speaking, articles may be moved by this improved engine system three times as fast, at one third the expense, and with the advance of only one-seventh the capital in the construction."49 James is not expressing common knowledge in this report on the possibilities of a railway line in East Anglia. Lord Hardwicke replied: "Though I am very sensible of the advantage of iron railways, yet I should certainly prefer the execution of the canal, for which the Act was obtained with so much difficulty...."50 Even dealing with a railway company, James could be equally unconvincing. At the first shareholders' meeting of the Stratford and Moreton Railway, July 13, 1821, James (a member of the committee of management) recommended Birkinshaw's malleable iron rails and Stephenson's locomotives; the new railway never used any power but horses and nearly chose castiron rails.5' Simmons has shown that the directors probably chose correctly, but the incident certainly indicates that locomotives on rails were not yet the conventional wisdom.

In 1824 Mechanics Magazine reprinted from the Scotsman a serialized unsigned article on railways which constitutes an ideological breakthrough, stating in a detailed and sound argument the case for locomotive railways as the rational choice for a transport system.52 It begins with the familiar argument that a horse on a canal will pull as much as thirty horses with wagons or 120 packhorses can carry, and it admits that, though cheaper to build, the railway hitherto had lacked some of the canal's advantages. It then further argues that "We are quite satisfied, however, that the introduction of the locomotive steam power has given a decided advantage to railways [which will lead to] ahllost boundless improvement and is destined, perhaps, to work a greater change in the state of civil society, than even the grand discovery of navigation."53

Two main reasons were put forward for this swing of the technical balance in favor of railways. In the first place, a horse exerted its best power at low speeds. At a dead pull a horse exerts a traction equal to 150 pounds which is reduced to less than one-half when he travels 8 miles per hour. At 12 miles per hour his whole strenght is expended in c arrying forward his own body, and his power of traction ceases. At 2 miles per hour, this still meant that a traction of 100 pounds would pull 90,000 pounds on a canal and only 30,000 on a railway. In the second place, however, the resistance of the canal rises with the square of the barge's speed, so that, at 4 miles per hour the 100-pound traction would only move 22,500 pounds. At the same time, the friction of a railway (excluding air resistance) did not rise with velocity— "the friction is the same for all velocities."54 The conclusion was clear: at horse speeds, a canal could compete with railways; at steam speeds it could not, and there was little power loss on railways at increasing velocities. This new comparison, unfavorable to canals, became the accepted view.55

The Penny Cyclopaediagave similar traction figures, but included roads as well. At 13.5 miles per hour, even a turnpike road was better than a canal.Sfi Joseph Priestley's preface to his compendium of canals and railways recognized the same situation.s7 Thus, from about 1825 railways were potentially superior to canals. This is not to say that railways had indisputably acquired machines that fulfilled expectations. Brandreth's Cycloped,an engine propelled by a horse on a treadmill, would yet be entered for the Rainhill trails, and discussion of the merits of pneumatic railways, stationary engines, and steam locomotives still lay in the future.

Technical Reaction by Canals

The reactions of canal companies support the view that an ideological change came in the mid 1 820s. Until this time, the canals' technical aspirations were-largely directed toward solving internal problems having to do with locks and water supplies. After this, their problems were set externally and they tried to emulate the railways. They did this by concentrating on increasing speeds and applying steam power—before there was serious widespread competition from railways. On the Paisley and Ardrossan canal, and near Birmingham, light barges carried passengers at more than 9 miles an hour. The Paisley passage boats were 70 feet long and weighed only 32 hundredweight, drawn by two half-blood horses which could comfortably work 12 miles a day in stages of 4 miles. Up to seven miles an hour, a high wave built up ahead and the going was hard, but at nine miles an hour the tractive force diminished and the boat rose on a wave behind the bows.58 Sir John Rennie commented that these swift boats had come too late, but earlier use might have retarded the adoption of railways.S9 This was not so. In 1832, the Mechanics Magazine was already convinced that the 10 miles an hour of horses was not sufficient and that steam promised little for canals.60 Nevertheless am was tried. The Victoria locomotive pulled passenger barges on he Forth and Clyde at 20 miles an hour, and one American enthusiast even wanted to put the rails on the canal bed and float the load.

John Lake designed another rail-canal hybrid consisting of piles supporting rails along each side of the canal. A steam engine in the I tug drove wheels which were pressed down onto the rails by levers. Thus the tug was propelled without reaction against the water. The Lake system also did away with locks. Instead, there were inclined planes with rollers, and the rails gave way to a rack up which the tug could wind itself on toothed wheels. The rack continued along the upper pound to allow the tug to pull up its train of barges.fi2 A successful trial was made at Leighton Buzzard in 1852, and there seemed promise that 10 horsepower could work 1,000 tons at 3 miles an hour. Yet again, there was hope that "the canal interest of this country may be restored to its former magnitude and importance, and raise its head from beneath the oppressive load of railway ascendancy."fi3

The more conventional response to railway challenge was using paddle or screw tugs. The Duke of Bridgewater had tried this before 1800, but the boat "went slowly and the paddles made sad work with the bottom of the canal, and also threw the water on the bank."64 Steam towing was certainly preferable to tacking a frigate through the Caledonian Canal, but the difficulty came in applying the principle to the narrower waterways. William Fairbairn wrote a tract about steamboats for canals and constructed stern-wheelers on the American plan for the Forth and Clyde company. He thought that stern wheels would not damage the banks like side paddles and named forty-five canals as suitable for his new boats.65 The Birmingham and Liverpool had two daily trains each way pulled by towboats with disc engines. and the Union Canal tried mounting screw propellers each side of the bow to prevent wash.66 Puddled clay banks were certainly vulnerable' to damage, but it cannot have been easy to combine a worthwhile' payload with a bulky steam engine or, alternatively, to tug the barges through locks. Thus, throughout the 1830s and 1840s there came claims that some new system or other would allow successful steam working on canals, but in practice the canals never competed with railway speeds. Nevertheless, they survived and even increased their carrying for a while, helped by the fact that manufacturers and carriers had integrated their operations with the canal network, while the smaller railway companies were slow to develop transfer arrangements. Soon after 1850, however, the railways began to carry a greater volume of goods than the canals.67

Road Transport Improves

The evidence of technical comparisons among transport systems | cannot be stretched to suggest that complete victory was possible for one or another, for they were not in simple competition. Certainly, canals could never have replaced road transport as effectively as rail could supplant canal, and there would always have to be some residual road system to supplement either of the more inflexible and costly channels. In the late 18th century, road transport improved in parallel with canals, especially providing better passenger services. lanes, reliability, and the completeness of services all improved steadily until the mid-1830s,68 and these organizational changes were accompanied by technical improvements in roads and vehicles. The railway locomotive after 1840 destroyed long-distance stagecoach serv~ces more completely than rail replaced canal goods transport, but road and vehicle improvements proved more than enough for survival in local work. It might even be suggested that the improvements in road transport before 1840, though not sufficient to challenge longd~stance rail transport for the next half century, nonetheless laid a ba|t for later victories.

improvements. It was never likely that roads would be completely replaced by rail or canal, though wooden rails were laid on some farms. Previously a vehicle had done well to survive journeys over the appalling surfaces, but as the latter improved, signers could look for speed, safety, economy of effort, and even comfort for passengers. The stagecoach and many forms of lighter vehicles—phaetons, chariots, landaus, britzskas, chaises, curricles, so on—all came to be With Telford and Macadam improving the roads, vehicle designers could begin to make real better designed. Weight distribution, fric_ and the function of wheels were analyzed in the "Mechanics" articles of early 19th-century Britannicas, and the Penny Cyclopaedia articles on "Springs," "Carriages," and "Steam Road Carriages" give more evidence of the strong contemporary interest, as well as proplding a popular, balanced account of the state of knowledge. The plentiful patents connected with improvements in vehicles are less instructive, however, and usually throw little more than an amusing sidelight on the innovation taking place.

The wheel was argued about most and changed least. The practice of "dishing" wheels,—that is, of arranging the spokes in a flat cone pointing in toward the vehicle—was condemned by theoreticians as an unnecessary source of weakness.' Lewis Gompertz went so far as to design an alternative to the wheel (see fig. 4), but his elaborate arrangement of cranks and cams working rotating legs, like the early attempts at caterpillar tracks or Boydell's plate-laying wheels, had nothing to offer to the new light, fast traffic appearing on Britain's roads. Such designs belonged on the old roads or on cross-country vehicles.70 Even Jones's and Sir George Cayley's elegant tension wheels, forerunners of the modern wire wheels, were not generally adopted. The wooden-spoked wheel went on until the days of motorcars, sometimes in the form of Hancock's improved artillery wheel, but most often as the traditional, maligned, dished wheel.

Around 1800, legislation and road trustees mistakenly encouraged vehicle owners to use wide-rimmed wheels which, tilting outward from the side of the wagon, needed conical rims to make them lie flat on the road. The grinding and rubbing action of such a wheel did far more harm than good to road surfaces. Improvements in wheels' however, grew from a clearer understanding of their function

Wheels reduce wasted effort by rolling rather than rubbing over the road; instead, friction is transferred to the small, smooth area of contact between the wheel and its axle. The higher speeds possible on the new roads made this friction more noticeable; but it is worth remembering that after 1800 there was a general need in engineering to reduce friction in faster, lighter machinery and in the new high pressure engines. Even if Gamett's "friction rollers," a forerunner of the modern roller bearing, were unnecessarily refined for their day, Collinge's patent axle of 1792 was certainly not. In this device, the axles and the wheelbox which turned on them were accurately ground together and their rotation pumped a supply of oil to the rubbing surfaces, so that a well-made Collinge axle could run for 5000 miles without attentions Such axles went on to mount the wheels of early motorcars a century later.

\\'hccls also smoothed the effect of bumps in the road by acting as a lever which allowed the carriage to ride less sharply over the obstruction. The larger the wheel, the less a given stone deflected the path of the carriage. If a wheel was 48 inches in diameter and had to be pulled by a horizontal force over a 7-inch bump then the horse had to exert a equal to the whole weight of the carriage; a smaller wheel made an even larger pull necessary. Hence, it was concluded, wheels should be as large as possible—but here the trouble began. A four-wheeled vehicle needed steering at the front wheels, and this was done by pivotting the two-wheeled axle about its middle. Swinging the axle like this brings the wheels rubbing up against the sides of the wagon or carriage—unless the wheels are small. The vehicle designer's dilemma was how to have big wheels and still steer the carriage. W. Bridges Adams proposed his curious "equivotol" rotary carriages, arguing that if large wheels could not be mounted on a pivotting beam, the n the beam should be abolished. Adams pivoted the vehicle itself, 50 that the front end with big wheels was hinged to the back end which had equally big wheels; to go round a corner, the carriage bent in the middled Rudolph Ackermann's improved steering was no more successful w hen first introduced, and this elegant system for linking the front wheels so that each followed the true curve of the carriage's turn (see fig. 5) did not replace the swinging-beam axle for nearly a century. By some freak of tradition, when early gasoline-powered cars with Ackermann steering were built, they still had front wheels smaller than the rear ones.

The simplest way to fit large wheels was on two-wheeled carria and carts which needed no steering gear at all; they merely follov the horse by the wheel on the ouside of the curve traveling faster tl the inside wheel. Such carts and carriages dealt with obstacles well c cornered smoothly, but needed to be loaded carefully since tt weight tended to bear down on, or lift, the horse through the leverage of the shafts when they were on a hill. The superiority of the cart over the four-wheeled wagon seemed to be unquestionable by 1850 when Pusey reported: "It is proved beyond question, that the Scotch and Nlorthumbrian farmers, by using one-horse carts, save one half of the horses which south-country farmers still string on to their three-horse waggons and three-horse dung carts, or dung pots as they are called . . . last year at Grantham, in a public trial, Eve horses with five carts here malt bed against five waggons with ten horses, and the five horses beat the ten by two loads."74 Two-wheelers were best for concentrated loads, like dung, but four-wheelers gave the room for lighter, bulky materials. A large farmer would usually have both. The analytical approach was not always welcome, however, and George Sturt and his workmen clung to their heavy designs until 1920. Sturt's classic of technical reaction, The Wheelwright's Shop, needs to be read and contrasted with W. Bridges Adams's English Pleasure Carnages in order to see how slowly changes make their way against settled tradition.'

Improved springs provided the main step forward in carriage huilrling. Earlier vehicles had used leather, wood, and occasionally steel springs to protect passengers against shocks, but around 1800 springs were used to improve performance. The difference was that springs were designed to mount as much as possible of the total weight on the springs, so that when the vehicle hit a bump, the springs absorbed the vertical movement and limited the shock. Without them, anti with small wheels, the horse was virtually lifting the weight over the hump, as we have seen. Lovell Edgeworth investigated springs, conthtding that, "Whatever permits the load to rise gradually over an obstatle without obstructing the velocity of the carriage, will tend to facilitate its draught, and the application of springs has this effect to a very considerable degree.76 A load of 8 hundredweight on springs was easier to pull than 5 hundredweight without. Unsprung, four~eeled carriages were vulnerable to "discordant motions of fore and Hill wheels" and likely to strain their frames by the weight being For\ sharply upon three wheels. Both two- and four-wheeled carriages without springs were subject to shocks from the road causing violent vibratory motion [and] the wheels to leap from one prominanceto another . . . tending to the rapid destruction of the vehicle and extremely unpleasant to the riders.""

It British roads adapted to spring carriages were like this, it is hard , imagine the roads in Canada, France, and the United States whose "imperfect condition precludes the possibility of using steel spring. with a due regard for economy."'8 Springs brought a great change for stagecoach passengers, as previously "the danger of sitting on the coach was then never hazarded by outside passengers; they were stuffed in straw in a huge clumsy basket that was fastened precisely over the hind axletree of the coach."79 Now, with springs, passengers moved to the roof of the coach and at the same time the horses managed to draw a greater load. Obadiah Elliot's elliptical springs were a particularly neat arrangement, cutting down the unsprung weight and lowering the coach's center of gravity. Bagwell considers their importance to coaching equal to the introduction of multitubular boilers for railway engines.80

Road Steamers: A Technical Failure

It is not hard to find reasons why a great deal of time and ability were devoted between 1825 and 1840 to establishing steam travel on common roads. Steam was coming to its maturity, a victorious period in which the basic advances such as condensing, high-pressure work- l ing, and compounding were giving engineers a flexible technology. Marine engineering produced many different engine configurations; steam hammers, agricultural applications, textile mills, waterworks, and other industrial applications produced more. John Herapath wrote to the Duke of Wellington in 1829 that "Adam Smith, I believe it is, informs us that one horse consumes for food as much land as would maintain eight persons ... every steam coach constantly in work would save that from horses, which would maintain one thousand four hundred and forty more of the human species." There is a long list of steam-carriage builders including Goldsworthy Gurney, Francis Macerone, Richard Roberts, the Heaton brothers, WilliamJames, Dr. Church, F. H. Hills, William Scott Russell, Burstall, and Walter Hancock. Hancock deserves special mention, for he came nearest to succeeding; but their efforts all failed in one way or | another, and the steam road carriage was to lie dormant for twenty I years after 1840. |

Conflicting reasons are given for these failures. Outside influences I such as the hostility of magistrates, heavy tolls, and the opposition of horse-owning interests are usually blamed.32 Was it really the case that external reactionary forces stopped the steam coaches? John Copeland recounts the problems encountered with tolls and coach owners and notes the difficulties inventors met in raising finance, but he also points out that "mechanical and operational difficulties were still being experienced in the late 1840's."83 Anthony Bird goes further in discounting the external causes for steam-coach failure and emphasizes the mechanical and metallurgical ills the steamers suffered.84 W. WorbyBeaumont earlier condemned steam carriages on technical grounds. Although "they were creations which were much more than the roughing out of a good working arrangement", they were too heavy; wore out quickly; used too much water; vibrated from the slow, large cylinder engines; and were smelly and hot.35 Fletcher, in his technical history, Steam on Common Roads, was more generous to the steam pioneers, but we may note of Beaumont and Fletcher that the first was an internal-combustion engine man and the other a steam engineer.36

These disagreements point to the interesting historical problems presented by the failures of the first generation of road steamers. There is a great deal of contemporary material in the newspapers, issues of Mechanics Magazine, and pamphlets by inventors; but much of the information is so prejudiced or incomplete that it needs to be treated very cautiously. Exaggerated claims began with Julius Griffith's pioneering carriage of 1821, which never left Bramah's works and they reached a climax with Gurney and Macerone vaunting thctoselves and vilifying others. These enthusiasts could be very persuasive. The steam-carriage group convinced a House of Commons committee in 1831 that all problems were solved and the future rosy f'or steam locomotion: hills could be climbed, boilers were safe and adequate, and no danger or annoyance awaited the public and their animals on the roads. Historians have relied heavily on this 1831 committee as evidence for the viability of steam locomotion, but there `` as little practical achievement before or after to justify its assertions A continuous story of disillusioned inventors, breakdowns, and explosions is a more reliable testimony to the real practicability of steam arriages in the 1830s. Yet some inventors stood aside with dignity from such mendacity. The Heaton brothers did their best, but "have been compelled to doubt the possibility of steam locomotion on common roads at an average speed of ten miles an hour, the wear and tear of the machinery, with other expenses, being so great as to exceed any probable receipts."87 Richard Roberts of Manchester likewise retired with good grace. Walter Hancock—who worked longer, built nine vehicles, and came nearer to success than the others—never descended to abuse; his letters and book are restrained and truthful. There is no doubt that Hancock's machines, after his first threewheeler experiment, ran long distances with fair reliability. He made trips to Brighton, to Cambridge, and once took the Stratford Cricket Club to a match; he carried more than 12,000 passengers on a regular service which ran for twenty weeks in 1836 (see fig. 6). Yet in July 1839, we still find Hancock calling out for a proper trial and backers: "I have sufficient confidence in my own (carriages) not to shrink from a fair trial with any rival whatsoever—only I am now ready, and therefore no delay whatever need take place as to any sufficient trial to satisfy parties really in earnest to begin." It is sad that Hancock, who financed his own experiments and came so near success, did not attract capital as easily as, for example, Gurney. But it is difficult to judge whether Hancock's carriages were reliable and economical enough, unless perhaps by building an exact replica—for which it would be very hard to find complete plans, identical materials, and working tolerances.

Certainly there was no contemporary agreement that high tolls or opposition from horse owners were decisive obstacles to successful steam road locomotion. The editor of Mechanics Magazine remarked acidly, "We have been informed that Sir Charles Dance, the proprietor of that carriage, [Gurney's Cheltenham-Gloucester coaches] does *ot hesitate to say in private that he was infinitely obliged to the road trustees for furnishing him with so plausible a pretext as they did for abandonning a losing concern."90 Hancock wrote in 1839 that "the whole evidence of Clerks of Trusts, and others connected therewith, goes to prove . . . that tolls are and will be diminished in consequence of the rivalry of railways." The Penny Cyclopaedia bears this out further.92 A similar doubt exists about the hostility which, it was claimed, the coaching interest showed to steam. A leading coach propnetor, Benjamin Home, who was interviewed by a parliamentary committee on tolls in 1836 argued that "You think you should be able to compete with steam carriages: Do you think you should beat theme—I fancy so. I only hope that steam carriages will be on the high road instead of being on railroads; there is every probability of our coaches doing very well if they draw about half the weight. With a tramroad we could maintain about twelve miles an hour very easily."93 If there is reason to doubt that turnpike or coach owners, whose livelihoods were threatened by the railways, were principally responsible for suppressing steam carriages, it is worth turning to the technical aspects of their history.

It was naturally tempting to repeat the railways' success on common Dads, but a steam carriage could notjust be a steam engine on wheels. The ease with which railway locomotives developed contrasts strongly with the difficulties of steam carriages. A good example of this is the simplicity and effectiveness of Stephenson's Rocket. Railways were, it turned out, ideally suited to steam locomotion, and the Penny Cyclopaedia explained the difference.94 Railways, unlike roads, bore an almost unlimited weight, and the locomotive ran on hard smooth surfaces (needing little springing), with few changes of level to surmount. Flanged wheels spared railway engines the need for steering, and the bends were too slight to need differential gears between th' inside and outside wheels. A road carriage faced not only these prob lems but more. It had to be light to spare the road surface and carry a good proportion of payload; yet all the propulsive machinery, fuel water, passengers or freight had to travel on the one carriage with perhaps a single trailer. By contrast, a railway locomotive with massive machinery could have its tender and then a great string of wagons. The design of horse-drawn carriages was equally little help to road locomotive designers. Unlike carriage wheels, the road locomotive's wheels had to transmit the thrust of an engine to the road. Its body had to carry a hot, heavy, vibrating steam engine and boiler as well as the driver, passengers, and useful load. All this had to be sprung against the jolting of the road, and it was not easy to carry the engine's power to the wheels while bouncing on top of springs. Thus, the designer had to achieve simultaneous major improvements in steering, suspension, transmission, boiler, and engine, and one can see to what extent Hancock, James, Hills, and the others were working at the limit of contemporary technology.

The Penny Cyclopaedia suggested that the time had come to combine the best elements of the steam carriages so far designed.95 Admittedly, the achievements were impressive. Hancock's multiple-plate boilers were noted for their lightness and steaming capacity; they were fed air by a fan and consumed exhaust steam, making for a very silent vehicle. Hancock, Russell and others overcame the problems of springing by a radius arm and chain drive from a separate crankshaft. Many other carriages suffered because their road axle was also the crankshaft, making it impossible to have a gear change or effective springing, and the machines were overstrained by road shocks, resistance from loose gravel, or steep inclines. Redmond used the Ackermann steering system, though it was perhaps anachronistic even in 1832 to operate it by means of reins. The Penny Cyclopaedia's list of individual technical advances can be extended. The Heatons, (see fig. 7) and W. H. James gave their vehicles sets of speed-changing gears driven by chains. Roberts in Manchester fitted his driving axle with a "jack-in-the-box," or differential gear, which he had used for cotton--' spinning machinery. Hills proposed a crankshaft running in oil and al condenser. The condensing idea was especially important. Heaved water consumption, usually lOO pounds per mile, was a perpetual headache on cross-country trips and even in London; and it is very doubtful whether muddy village duck ponds were a lesser evil than obstructive London water companies. When Hancock hoped to go 60 miles to Reading without a stop he must have had 2 tons of water in his trailer, and on trips like this a condenser would have saved man' times its own weight. Hancock's strong wheels were another successful innovation. If the best of these ideas had been combined in one design, it might have been financially successful. But the Penny Cyclopaedia's advice was not taken before the steam carriages began their sleep of twenty years about 1840. If a technology does not include a method of making correct choices, there is a defect in the entire system, whatever its piecemeal merits.

We are also left with the impression, however, that the steam carriages were lacking, despite all efforts, in a number of technical respects. Constructors spoke more readily of propelling 4 tons at 20 miles an hour than of the difficulty of retarding such inertial motion. Crankshafts, hand welded and forged from bundles of wrought-iron bars, were sometimes flawed and may partly explain the axle breakages which Gurney and Russell blamed on loose stones laid by malic~ous opponents. Boilers exploded (even Hancock's), hills overstrained machinery, and carriages were ignominiously towed home by horses. Hancock got to Brighton in 1833, but there "unluckily an inferior part of the mechanism, technically called a clutch, gave way and led to the fracture of a cogged wheel, which gave motion to the centrifugal fire funnel, and the carriage was brought to a dead stand."''6 Similar foreign technical failures reinforce the conclusion that it was not just British turnpike tolls or local opposition which stultified progress. The difficulties of Dietz in France, inventors in Prussia and Denmark, and J. R. Fisher of New York all support the conclusion that it was technology rather than social or economic institutions which stymied steam coaches.97 It is also not fair to blame those who put their money into two-horse omnibuses and thereby starved steam coaches of capitals When steam succeeded on roads after 1860 the situation was quite different, though even then it was preeminently the large slow traction engines, built for massive strength rather than speed, which succeeded. And when Serpollet built his successful steam cars at the end of the century, even greater changes had taken place in materials and manufacturing methods. The successful White or Stanley steam cars came from an advanced technology using liquid fuels, machine tools, steels, and boilers unknown in Hancock's day. They had more in common with their gasoline-engined contemporaries than with the slow-rewing juggernauts of the 1830s.


Three general conclusions follow from this study of transport in the industrial revolution. There were fairly clear technical criteria for comparing the various forms of transport, and comparisons severe available in popular accounts like encyclopacdia articles. Furthermore, despite many local exceptions, the general outcome of development was broadly congruent with the technical merits of the systems as they appeared to be emerging at the time. Railways began as feeders to canals and only came to the forefront with new ideas as well as new locomotives after 1825. Canals predominated, despite their difficulties, until their general inferiority was shown; and steam road vehicles, which could reasonably have been expected to succeed, turned out to have unexpected drawbacks.

Thus, there seems to be a strong argument for laying greater emphasis than is normally done on technical factors, even though many cases exist of personal or local rivalries dictating the choice of system. ( nntemporaries were plainly interested in technical comparisons, and further work is needed to evaluate their influence. The great floods of investment in railways after 1840 were doubtless due to a "bandwagon" effect and the expectation of high profits based on past experience, without any regard to careful technical knowledge. But before the bandwagon investment, presumably, technical expectations could have been decisive.

Second, it is tempting to argue that something like a paradigm shift occurred around 1825 in the changing ideas about horses and canals compared with railways and steam. The canals' change in conduct, before and after 1825, and the way the new ideas preceded the great mania for building railways seem to support this. However, there must he reservations about using the paradigm idea, since many amhiguities existed. Horse-powered railways remained; steam-powered canal boats were built, and for decades there was an extraordinary flexibility in assembling combinations of power sources and transport channels—railed, flat, or tubular, water, iron, or stone. A strong theoretical case for steam railways was created, but, equally, engineers continued to choose the combination of technical means which suited the particular situation. Concentration on locomotives, like the stancPard gauge, only came later with a rational national system. On balance, I would argue that engineering pragmatism, fundamental to the success of the profession, renders concepts like paradigm shifts of dubious usefulness in the history of technology.

Finally, there is much to respect in the way the pre-Victorians held public debate on the best methods of transport. It is interesting to reflect that a similar debate is being held again today, partly because the fossil fuels and materials which carried the early 1 9th century out of the hydraulic tradition are now scarcer. Thus we find ourselves reconsidering those decisions of the 1 820s which excluded renewable resources.