The Ocean Railway: Isambard Kingdom Brunel, Samuel Cunard and the Revolutionary World of the Great Atlantic Steamships

As Symington later told the story, in March 1802 the Charlotte Dundas took on board Lord Dundas, his son Captain George H. L. Dundas of the Royal Navy, and others, and towed two loaded vessels of seventy tons each a distance of nineteen and a half miles along the canal in six hours, against a strong head wind. ‘This experiment not only satisfied me, but every person who witnessed it, of the utility of steam navigation, ’ Symington later wrote. But the canal proprietors worried that the steamboat’s agitation and wake would harm the banks of the canal, and so rejected the plan. Lord Dundas then arranged for Symington to meet the Duke of Bridgewater, the leading canal entrepreneur in England. The duke at once ordered eight of Symington’s vessels – but he soon died, cancelling the deal. This double rejection after apparent successes left Symington too disheartened to persist. ‘This so affected me,’ he recalled, ‘that probably I did not use the energy I otherwise might have done to introduce my invention to public notice.’

This version of events has become the standard historical account, but it is wrong in certain particulars. Drawing from memory some twenty-five years later, Symington compressed two separate trials into a single event. On 4 January 1803, the Charlotte Dundas, with the two Dundases and others on board, towed a 100-ton boat from Stockingfield to Port Dundas at three miles an hour ‘amidst a very large concourse of people’, according to a newspaper report, ‘who were exceedingly well pleased with the performance.’ On 28 March 1803, the steamboat also towed two loaded vessels, a combined 130 tons, from Lock 20 on the canal to Port Dundas, eighteen and a half miles in nine hours and fifteen minutes – a speed about 40 per cent slower than Symington later remembered. For this trial he had incorporated suggestions by Captain George Dundas for how to manage the tow lines around sharp bends in the canal. The Glasgow Herald and Advertiser praised ‘the very appropriate mode in which the machinery is constructed, and the simple yet effectual manner its power is applied in giving motion to the vessel’. The newspaper also credited Lord Dundas for his generous financial support and perseverance in the ‘costly experiments’.

A few days later, the Herald and Advertiser published a testy letter from a Forth and Clyde Canal proprietor which fleshes out Symington’s later explanation of why his steamboat was banned from the canal. The letter writer pointed out that a vessel passing through one of the canal’s thirty-nine locks used a lockful of water, so a towboat plus barge consumed twice as much water (and the canal had recently been closed by low water); that the Charlotte Dundas, contrary to another report, would save no money over tow horses given her initial expense, the cost of coal, her crew, and general wear and tear; and that Symington’s earlier steamboat of 1801 could not run with any ice in the canal, and this problem had perhaps not yet been solved. After all these objections, the proprietor added, ‘It will be observed too, that the motion of the boat raises such an agitation in the water, as to injure the banks.’ In conclusion – and this probably clinched the matter – the writer regretted that Lord Dundas had been given all the public credit for funding Symington’s efforts. ‘It should have been added, that the Proprietors of the Forth and Clyde Navigation have already paid about £1700. for these experiments of this ingenious mechanic, without reaping any benefit from them, and without even getting any credit for their liberality.’

Given this bristling mixture of unmet criticisms and wounded, unappreciated generosity, and (one may assume) competitive resistance by the local owners of horses and stables, it is not surprising that Symington got no farther with the canal proprietors. Hoping for other wisps of interest in the Charlotte Dundas from somebody else, he laid her up near the canal at Bainsford. There she lingered on for almost sixty years, rotting and rusting away, a waning curiosity of the early steam age. Like James Watt, Symington was a gifted inventor saddled with a fainthearted personality, too easily deflected from his purposes. His singular misfortune was that – unlike Watt – he never found his Matthew Boulton.

Robert Fulton, the American painter and inventor, knew all the precedents in steam navigation. During twenty years spent abroad, in England and France, he studied the efforts of other steamboat pioneers and tried out his own improvements. In contrast with most of the other innovators, he was blessed with an overpowering confidence and persistence which, along with good looks and a gift for friendship, brought him the continuing support of rich, powerful patrons. Ultimately he returned to America to build and run the first commercially successful steamboat. Today most Americans consider him the principal originator of steam power on water. The process by which he achieved this reputation – and thus the reputation itself – demands a renewed examination.

For most of his two decades abroad, Fulton was preoccupied with other inventions than a steamboat. Living in France from 1797 to 1804, he devoted himself to an elaborate, quixotic, finally unworkable scheme for submarines and explosive mines, intended to revolutionize naval warfare. His intermittent interest in steamboats was revived when Robert R. Livingston arrived in Paris late in 1801 as the US minister to France. A man of enormous wealth and political influence in New York, Livingston hoped to develop a steamboat service for the Hudson River back home. Fulton had found his final, most significant patron.

During the summer of 1802, Fulton conducted a series of trials with a model boat powered by a clock spring. After considering all the propelling devices used by his predecessors, he settled on an endless chain with paddles or buckets attached to it. Resembling the tread of a modern tank or bulldozer, the chain was draped over two wheels across the side of the model, dipping into and seizing the water at the bottom of its cycle. Livingston, drawing from his own previous sallies at steamboat invention, preferred paddle wheels; but after Fulton reported on his trials with the model, arguing his case quite vehemently, Livingston was converted to the endless chain. In October 1802 the two men signed an agreement to build a large steamboat in New York, designed for the Hudson River traffic to Albany.

Now came a surprising, puzzling twist in the story. At some point that autumn, after insisting so aggressively on the superiority of his endless chain, Fulton decided to adopt paddle wheels as his propelling device. His biographers have guessed that Fulton switched to avoid infringing a French patent, granted earlier that year to an inventor named Desblancs, for a similar steamboat with an endless chain. But Fulton had learned of this patent in June, and as late as September he was nonetheless still urging his own version of an endless chain. Something else must have persuaded him to change this crucial aspect of his design.

A possible explanation was later provided by William Symington. As he told the story in the 1820s, Fulton had come to Scotland to see one of Symington’s vessels, explaining that he intended to return to America to build a steamboat, and that his project could lead to a rewarding business for Symington as the inventor. Flattered and intrigued, Symington ordered steam up in his paddle wheeler and took Fulton and others for a ride. From Lock 18 on the Forth and Clyde Canal, they went four miles west and back in one hour and twenty minutes, at an average speed of six miles an hour – ‘to the great astonishment of Mr. Fulton and the other gentlemen present’, according to Symington. Fulton asked questions, took notes, and made sketches of the steamboat. After this single encounter, Symington recalled, he never saw or heard from Fulton again.

The dating of Fulton’s visit presents problems. Symington placed it in July 1801 or July 1802. In 1801, however, France and England were at war, severely limiting travel between the two countries. Fulton would have had great difficulty in making his way from France to Scotland; at the time he was also still quite focused on his submarine and mines, to the exclusion of other interests. The Peace of Amiens in March 1802 allowed a brief lull in hostilities, easing travel restrictions. By then Fulton, with Livingston’s beckoning patronage, had turned his attention nearly full-time to inventing a steamboat. He spent the summer of 1802 at a resort in the Vosges Mountains of northeast France, too far from the English Channel for a convenient trip to Scotland. That autumn he was back in Paris, intent on his steamboat. The most probable date of Fulton’s encounter with Symington is thus the autumn of 1802, when the Charlotte Dundas was almost ready for her first major trial of January 1803. The journey from Paris took three days to London, then about sixty hours by mail coach to Glasgow. He could have made the round-trip in two weeks.

With travel again flowing between France and England, Paris was full of British tourists from whom Fulton or the widely acquainted Livingston might have heard about Symington’s boat. A trip to England was clearly on Fulton’s mind that autumn. His friend Joel Barlow, also interested in promoting a joint steamboat scheme, had recently urged Fulton to go to England ‘silent and steady… quiet and quick’ to obtain a steam engine. His formal agreement with Livingston in October also bound Fulton to go ‘immediately’ to England for the same purpose. Fulton left no surviving record of such a trip at that time. But he could have gone secretly – silent and steady, quiet and quick – on steamboat business, especially to examine the Charlotte Dundas, the most promising such experiment in the world at that time. In late September, he was conspicuously absent from a dinner party given in Paris by the painter Benjamin West. Fulton was a close friend to West, his main mentor in painting. Joel Barlow and his wife, with whom Fulton lived in a ménage à trois, did attend the dinner. If Fulton had been in Paris, he surely would have joined the party. Perhaps he was then quietly off to Scotland.

This mystery turns on hard questions about Fulton’s character. Could he have made a clandestine trip to Scotland, borrowed from Symington’s work, and later hidden the entire episode? His subsequent history of lies and deceit suggests that he might have. In 1806, for example, he claimed in writing that he had held an American steamboat patent for fourteen years, and that some $280,000 had been subscribed to build twenty of his vessels for service on the Mississippi River – none of which was even remotely true. Later, when embroiled in patent controversies, he forged a ‘copy’ of a drawing he had supposedly made in June 1802 of a Hudson River steamboat with paddle wheels, at a time when he was actually still committed to an endless chain for propulsion. He also forged a letter, which he dated to 1793, about his supposed interest in paddle wheels at that time. In 1815, shortly before his death, he was caught committing perjury with this letter. All these manipulations were intentional, self-serving lies on Fulton’s part.

Symington’s later recollections, by contrast, erred in some details, but the essence of his account of the Charlotte Dundas is verifiably true. His version of the Fulton story was also corroborated by Symington’s engine man, Robert Weir. In 1824, after the matter had become controversial, Weir signed a sworn affidavit that he had fired up the boiler of the Charlotte Dundas on the occasion of Fulton’s visit and had heard Fulton identify himself by name and nationality. After their brisk eight-mile demonstration, according to Weir, Symington had lamented the difficulty of running his steamboat through the narrow Forth and Clyde Canal, and Fulton had replied that the broad rivers of America would present no such problem. The details and certainty of Weir’s affidavit seem authentic.

Fulton’s own explanation of how he converted to paddle wheels, later given under duress, must be weighed carefully. In 1811 he asked Joel Barlow to endorse his version of certain events for a potential patent lawsuit. ‘I want your deposition as follows,’ he instructed: that in the autumn of 1802, while living at Barlow’s home in Paris, he had conducted experiments with various propelling devices, which by Christmastime had convinced him to adopt paddle wheels. ‘You will have this copied on foolscap,’ Fulton told Barlow, ‘and sware to it.’ Barlow apparently complied. It was at about this time that Fulton also forged other documents to bolster his claims of steamboat originality.

The smoking gun in this mystery is the vessels that Symington and Fulton actually produced. In January 1803 Fulton drew up the plans for his first steamboat. Overtly she did not much resemble the Charlotte Dundas. long and lean instead of short and stubby, with a different arrangement of the machinery and a distinct means of converting the engine’s reciprocating action to rotary motion. But in four crucial respects the boats may be linked. In both cases the engine’s cylinder was put in the exact centre of the hull, with the boiler behind it. Like the Charlotte Dundas, and unlike the vessel recently proposed in his agreement with Livingston, Fulton’s first steamboat was a towboat, with room on board just for the machinery, fuel and crew. Both vessels were propelled by paddle wheels: Symington’s by a single wheel at the stern, Fulton’s by two wheels attached to the sides. And – the most telling detail – Fulton’s paddle wheels were placed quite high in the boat, as in the Charlotte Dundas, so that only three paddles were under water at once, avoiding the wasted up-and-down motions of a more deeply immersed wheel.

It seems more than probable that Fulton did see the Charlotte Dundas and borrow from her design without ever acknowledging the debt. His first steamboat, built to the plans of January 1803, underwent a successful trial on the Seine later that year. Fulton eventually returned to the United States and, with Livingston’s support and a Boulton and Watt engine imported from England, made the paddle wheel steamboat later known to history as the Clermont. Her machinery and paddles closely resembled those of Fulton’s first steamboat of 1803 – and therefore may also be linked to the Charlotte Dundas. With the Clermont and her successors, Fulton ran a profitable steamboat service between Albany and New York City, marking the first sustained commercial use of steam navigation. The unfortunate Symington faded into obscurity and died penniless in 1831.

From this point on, geography largely determined the separate development of steamboats in America and Great Britain. In the United States, with its vast internal networks of inland lakes and long, broad, navigable rivers, steam navigation typically took the form of riverboats: large, fragile craft of shallow draught, driven at top speed by high-pressure boilers prone to explosion and disaster. In Britain, the characteristic steamboats were smaller and slower but safer, with low-pressure boilers, and sturdy hulls and high bulwarks designed to survive the heavier seas of coastal and ocean traffic. The future of Atlantic Ocean steamships would unfold mostly in the British Isles.

William Symington’s many frustrations had an apparent chilling effect on steamboat building in Great Britain. After he finally laid up his unwanted creation at Bainsford, nine years passed before another British steamboat was launched. The Comet, completed in the summer of 1812, became the first passenger steamer in Europe. Her planner and owner, Henry Bell, had been interested in steam navigation for over two decades. But his mercurial nature – his ‘restless volatile genius’, as a friendly biographer put it, ‘flying from one daring scheme to another’ – kept Bell pushing on to the next experiment before finishing his last one. It took him a long time to settle down and produce his first actual steamboat.

Like Watt and Symington, Bell was a Scotsman, born in 1767 near Linlithgow, west of Edinburgh. He came from a family of millwrights and was trained as a mason, millwright and shipbuilder, with early stints in Glasgow and London. (‘I was not a self-taught engineer, as some of my friends have supposed,’ he later insisted.) Settled in Glasgow, he built houses and public works and started to focus intermittently on steamboats around 1800, after Watt’s patent expired. Bell tried to interest various patrons and governments but got no favourable responses. He hung around the Carron Works when the engine and machinery of the Charlotte Dundas were being constructed, to the point even of making himself a nuisance to the workmen. Later he repeatedly inspected Symington’s boat at Bainsford.

When Bell became the owner of the Baths Hotel in the resort town of Helensburgh, on the Clyde some twenty miles west of Glasgow, he acquired the necessary practical goad that pushed him finally to build a steamboat – for bringing Glaswegian customers out to his hotel. The Clyde, as yet undredged, was then a winding, shallow stream, often filled with sandbanks. Sailing boats drawing only five feet still might be grounded for an hour or two; passengers would be obliged to run on deck from side to side, rocking the hull and loosening the keel from the sand. To reach Glasgow, at the river’s eastern and narrowest point, Bell’s steamboat for the Clyde had to be small.

In the autumn of 1811 he contracted with John Wood, a shipbuilder in Port Glasgow, for a hull forty-two and a half feet long, eleven and a half feet wide, and five and a half feet deep, and a total capacity of only twenty-five tons. John Robertson of Glasgow, a builder of textile-mill machinery, made the engine: a cylinder eleven inches in diameter, stroke of sixteen inches, and four horsepower. Four small paddle wheels hung on the boat’s sides. Her smokestack at the bow doubled as the mast for a single square sail (as on an old Viking ship). The Comet was named not to suggest her speed but in tribute to Halley’s Comet, recently visible in the night sky. Launched in July 1812, she began her Glasgow to Helensburgh to Greenock service a month later. As she puffed along the river, local boys would run down to the water’s edge, expecting or hoping to see her blow up. She made the trip three times a week in each direction, covering the twenty-six miles to Greenock reliably in four hours, sometimes under three and a half – as fast as horse-drawn travel by land, and cheaper and much more comfortable than heavy, unsprung vehicles on bad roads. Within a year, four road coaches that had been taking passengers to Greenock stopped running for lack of business.

This quick success provoked a productive steamboat competition. For some years before the Comet, Bell had worked on steam navigation designs with John Thomson, a Glasgow engineer. Thomson had made sketches of a boiler and machinery, and he expected to help Bell produce his steamboat. But Bell instead went ahead on his own, leaving Thomson angry and disappointed. He took his revenge by building a bigger, faster boat, the Elizabeth. Also constructed by John Wood, she was fifty-nine feet long by twelve feet wide, and forty tons, with a nine-horsepower engine. Her cabin included such touches of luxury as carpets and a sofa, windows with tasselled curtains and velvet cornices, and even a small shelf of books. The Elizabeth ran from Glasgow to Greenock and back every day, instead of only thrice weekly, carrying as many as one hundred passengers at speeds up to nine miles an hour, cutting steadily into Henry Bell’s business.

Over the next few years, steamboats appeared on most of the major rivers of Great Britain. Just before the first railroads, they started to speed and discipline the pace of life, ratcheting up to the predictable, rationalized clock time of the Industrial Revolution. Steamboats ran at man’s pleasure, ploughing along through adverse winds and waves, coming and going as ordered. A clock soon became a necessary instrument for doing business. ‘The merchant, knowing the time of the tide, can count to an hour, in ordinary weather, when his goods will arrive; and will not be disappointed in one case out of thirty’ Henry Bell asserted. ‘I expect in a short time to see all our ferries, and our coasting trade carried on by the aid of steam-vessels.’

In May 1815, the first long ocean passage by a steamboat in Europe tested steam’s potential for that coasting trade. The Glasgow (later renamed the Thames) had been built by John Wood a year earlier. She showed steady progress in size and power: seventy-two feet long by fifteen feet wide, sixteen horsepower, and seventy-four tons. Sold to London interests for service on the Thames, she put to sea just for delivery to her new owners, not to start a regular ocean service between Scotland and England. Under the command of George Dodd, a young architect and civil engineer, she set forth from Glasgow with an eight-man crew of a master, four sailors, and a cabin boy – and a smith and fireman for the engine.

The Glasgow ran easily down the Firth of Clyde into the narrow channel between Scotland and Ireland. Here she encountered more difficult sailing than anything normally seen on the Clyde, as the ebb tide collided with strong swells sweeping in from the North Atlantic. Unable to make progress, Captain Dodd had to seek shelter in Loch Ryan. The Glasgow ventured out again, was tossed around, and nearly wrecked on the rocky Irish coast. She stopped at Dublin for several days of rest and repairs. Naval officers came to see her, agreeing that she would probably not survive a true stormy sea and had better hug the shore. Watched by thousands of spectators ranged along her way, she left Ireland with just two brave passengers for London.

Away from the coast in the Irish Sea, she again met heavy swells. ‘The movement of the vessel differed entirely from that of one pushed by sails or oars,’ noted Isaac Weld, one of the passengers. ‘The action of the wheels upon the water on both sides, prevented rolling; the vessel floated on the summit of the waves, like a sea-bird. The most disagreeable movement took place when the waves struck the ship crossways; but here too its particular construction gave it a great advantage; for the cages which contained the wheels acted like so many buoys.’ As water flooded into the paddle box on the windward side, the compressed air exploded in an alarming report whose percussive force made the whole boat tremble. This noise exploded again, by reaction, on the other side of the Glasgow, then again, much diminished, on the first side. At this point she at least stopped rolling for a while. ‘During the rest of the voyage,’ according to Weld, ‘the vessel made what the sailors call, a dry way, that is, it danced so lightly over the waves, that it never took in one; and in all the passage we were not once wet…which could not be expected in any common ship.’

As they neared Wexford, at the southeastern corner of Ireland, the Glasgow’s thick coal smoke convinced local pilots that the approaching boat was on fire. They scrambled out to sea, expecting to save lives and perhaps seize some profitable salvage – and were surprised and disappointed that the Glasgow was just steaming along in safety. She crossed St George’s Channel to England, near Cape St David, and was again greeted by a flotilla of would-be rescuers not anticipating a smoking steamboat in those waters. Heavy seas tossed up waves so high that at times the crew of the Glasgow could not see the coast. Captain Dodd picked his way through, leaving far behind a fleet of sailing vessels trying to keep pace. They stopped for two days at Milford Haven for inspections and to scrape the saltwater scale from the boiler, a problem not encountered when sailing freshwater rivers.

Rounding Cape Cornwall into the English Channel, they encountered their highest swells yet. ‘It seemed impossible to pass,’ Weld recalled. ‘The vessel appeared to suffer… Night approached, and no harbour presented itself, except that which we had quitted, and which was already too distant.’ Captain Dodd hoisted sail, which helped steady her, and struggled against the waves for hours until reaching calmer waters. The rest of the trip was smooth and easy. At Portsmouth, tens of thousands of people came out to stand back and be amazed. The Glasgow reached the mouth of the Thames on 11 June, intact and in good order. She had covered 760 miles in a bit more than 121 hours of actual sailing time, spread over almost three weeks.