“The Most Original”

James Rumsey, a remarkable American inventor of the eighteenth century, is today almost forgotten. Most technological histories accord him little more than a footnote, as one of the less successful claimants to the invention of the steamboat. To some extent this is due to the loss of all his United States patents in the great Patent Office fire of 1836. But his four British patents, describing more than twenty inventions, have survived, and they and other sources make it apparent that Rumsey was one of America’s most creative inventors. He made several major contributions to the steamboat; his most important invention was his reaction wheel, which served as a starting point for the highly fruitful work by Americans on hydraulic turbines. Rumsey’s greatest achievement, however, lay not in any of his twenty to thirty inventions but in his role in changing the very process of invention. He forged new links between science and technology in an age when Newtonian mechanics was still new and invention was still largely a matter of craftsmanship and trial and error. He brought the two together by creating ways of thinking that have become integral to modern engineering design.

Born in 1743 in northeastern Maryland, Rumsey was an unusual combination of Southern gentleman and ingenious Yankee. He came from an impoverished branch of a prominent family; his father was a poor farmer, but his cousin William Rumsey was a prominent Maryland landowner and a member of the American Philosophical Society. Despite a rudimentary education, precarious finances, and almost ceaseless struggle, Rumsey clung to his gentility as perhaps only the downwardly mobile can. His polite manners did not, however, account for his ability to gain and keep the support of powerful patrons such as George Washington, Benjamin Franklin, and Thomas Jefferson. Rumsey impressed his contemporaries with a fertility of invention that has seldom been equaled. Jefferson called him the “most original and the greatest mechanical genius I have ever seen.”

Rumsey was the proverbial jack of all trades; a “born mechanic and skilled blacksmith,” he also had practical experience as a sawyer, building contractor, storekeeper, innkeeper, miller, and millwright. In 1782 and 1783 alone he built a sawmill, an iron mill, and several gristmills in Berkeley County, Virginia (now Morgan County, West Virginia), plus several buildings in Bath, the center of Berkeley County (now Berkeley Springs, West Virginia), for George Washington. It was during his years in and near Berkeley County, from 1782 to 1788, that he found his vocation as an inventor.

Rumsey’s first significant invention was a boat designed to mechanically propel itself upstream against the current. Power was supplied by a waterwheel, which moved a carriage back and forth inside the boat like a piston, parallel to the keel. Poles were loosely mounted to the carriage at about a forty-five-degree angle in the downstream direction so that they would drag along the bottom during the carriage’s forward motion but would catch and push against the river bottom on the carriage’s backward return.

In 1784 Rumsey, having made a model of his boat, sought a patent monopoly for it from the Virginia legislature. He was almost rejected, but Washington’s support saved the day. Congress promised a large land grant effective upon a successful demonstration of a full-size boat able to travel the Ohio and Mississippi rivers. Rumsey’s inability to make the pole boat work- and it must have seemed especially inadequate in the deep waters of those big rivers—and his increasing interest in steam were probably decisive in his subsequent shift to a scientific basis for his inventions. In July 1785 he was appointed, through the patronage of Washington, to a canal-engineering position with the Potomac Navigation Company. This undoubtedly furthered the scientific self-education he had undertaken.

It is possible to reconstruct, in broad outline, Rumsey’s shift from a craft to a scientific base for his inventions. Prior to 1785 his inventions, notably his pole boat, were craft-based. That is, they consisted of components familiar to craftsmen but arranged in unfamiliar configurations. But craft knowledge provided no method for perfecting such inventions; success depended upon luck and cut and try.

Science seemed to offer greater certainty. From 1785 on Rumsey’s inventions were scientific. They included a watertube boiler and a steam pump in 1785, a steamboat powered by the reaction of a backward jet of water in 1786, an improvement of Barker’s reaction water mill at about the same time, and a piston-driven hydraulic sawmill in 1787. (Barker’s mill was a hollow vertical cylinder with boxlike arms projecting from the bottom; water that was poured down through it would spin out from apertures on alternate sides of these arms, driving the mill by reaction.) Rumsey continued to work on these inventions and sketched about twenty more in a remarkable burst of creativity during his stay in England from 1788 until his death there, in December 1792.

Benjamin Franklin helped begin all this activity. In December 1785, after his return from France, Franklin gave a paper before the American Philosophical Society in which he reported the conclusions of Daniel Bernoulli, one of Europe’s leading scientists, that a boat might be advantageously driven by the force of reaction of a jet of water shot backward from it. Franklin added his own suggestion that the jet of water might be produced by a steam engine. Rumsey evidently heard about Franklin’s proposal, and he designed a steamboat based on this plan.

The greatest scientific influence upon Rumsey, however, was not Franklin but John T. Desaguliers, a British lecturer of French ancestry who had popularized Newtonian science. Desaguliers believed that modern science should provide new foundations for technology, and he addressed the second volume of his two-volume treatise, A Course of Experimental Philosophy , to craftsmen without advanced training in mathematics. The first volume had been published in 1734, and the second in 1744, the year following Rumsey’s birth. Despite its age, this work provided Rumsey with an excellent guide to Newtonian physics cast in experimental form and employing only simple arithmetic. The technology was somewhat outdated, but Rumsey found in it much inspiration for invention, including the first public description of Barker’s mill.

Rumsey learned from Desaguliers to think about technology in the abstract manner developed by scientists, and he learned to break down machines into their component parts and analyze the workings of these parts as independent entities. He also benefited from Desaguliers’s emphasis on building scientific models of machines. Nonetheless, Desaguliers’s legacy was a mixed blessing; to him as to Rumsey, machines were simply expressions of scientific laws, and both he and Rum- sey tended to neglect the development process that translates theoretical principles into a workable full-scale invention.

Rumsey was also misled by an erroneous scientific principle that Desaguliers presented. Desaguliers reported on the “discovery” by Antoine Parent of the inherent inefficiency of waterwheels powered by the direct action of water. Parent had deduced mathematically that according to Newton’s first law of motion, all such wheels are limited to no more than about 15 percent efficiency (the ratio between the available power and that utilized by the machine). Since the paddle wheel of a boat is essentially a waterwheel run backward, the limitations on the efficiency of water mills should apply equally to paddle wheels. It appeared that a better mode of propulsion for steamboats would have to be found. In fact, Parent’s theory had already been proved wrong in experiments by the great British civil engineer John Smeaton, which had” been published in the Royal Society’s Philosophical Transactions for 1759-60. The actual limits of efficiency of vertical water mills varied from 50 to 100 percent. Rumsey was apparently unaware of Smeaton’s work.

Many leading scientists and engineers were convinced that the use of reaction in place of action offered a scientific high road to novel and efficient mechanical inventions. (Reaction means recoil, as with a cannon; this force is always equal to the action with which it is paired.) This line of reasoning was the direct inspiration for a number of reaction motors, including Barker’s mill and a continental equivalent, Segner’s mill. Leonhard Euler, the great mathematician, and his son Johann designed reaction mills inspired by Segner’s. The lure of Parent’s theory provided indirect inspiration for the reaction boats suggested by Bernoulli and Franklin as well.

On December 3,1787, Rumsey held a public demonstration of his steampowered jet-of-water reaction boat before crowds gathered on the cliffs of the Potomac River in Berkeley County. The spectators cheered the successful test, but mechanical difficulties suggested the need to leave the Virginia frontier for a place where more advanced technologies were available. It was time for Rumsey to move to a larger stage. In search of support he traveled in March 1788 to Philadelphia. There he presented his ideas for inventions to the American Philosophical Society. Led by Franklin as first subscriber, a number of members formed the Rumseian Society to raise money to send Rumsey to England to exploit his inventions.

After his arrival in England in June 1788, Rumsey entered into negotialions with the firm of Matthew Boulton and James Watt. Boulton offered him a partnership in the manufacture and sale of steamboats, but the firm broke off negotiations when Rumsey tried to get more favorable terms.

Rumsey felt he was entitled to more partly because of the high value he placed on certain inventions that existed only as sketches and as theories. Boulton rejected Rumsey’s claims; in one instance—Rumsey’s calculation of the theoretical efficiency of his tubular steam boiler—Boulton responded that “you may easily determine its ratio [i.e., its efficiency] by experiments which are worth a thousand fancys or theories.”

The results were tragic for Rumsey. Boulton and Watt had capital; Rumsey’s lack of funds was to plague him until his death. They also had knowhow. Moreover, Boulton was right: Rumsey placed too much faith in deductions from scientific theories, and he failed to appreciate the need for a long period of expensive development even for a science-based technology. In this he had been led astray by Desaguliers, who, though he advocated experiment, thought inventors could contrive new machines by deduction “ a priori , that they may always be sure of what the effect of the performance will be.”

When Rumsey had presented his ideas in Philadelphia in 1788, one of them had been for a reaction mill, an improvement on Barker’s mill with the water admitted from below, and Rumsey made the claim that “there is no other method of applying the same quantity of water that can give so much motion and power in an equal time.” But when he measured the power and efficiency of his mill, by means of an ingenious experimental apparatus, Rumsey found that high efficiencies could be obtained only at unattainably high speeds. He measured an efficiency of just 29 percent.

Reaction mills based on Barker’s mill are inherently inefficient. In modern terms this is because they lose so much kinetic energy through the high speed of the water at discharge. That is why mills similar to Rumsey’s improvement of Barker’s mill are widely used as spinning lawn sprinklers. There the excess kinetic energy serves the useful task of spreading the water over a wide area.

While he never explicitly acknowledged it, Rumsey evidently knew that his reaction steamboat would have the same low efficiency characteristics as his mill. Both in reaction mills and in boats of Rumsey’s sort—as in modern jet aircraft as well—the theoretical limit of efficiency is 100 percent, but it can be approached only as the speed nears infinity. A serviceable reaction boat was quite feasible, but screw propulsion or paddle wheels were inherently more efficient at realistic speeds.

Though Rumsey sometimes failed disastrously when he tried to use science deductively, he was remarkably successful in using methods suggested by science to produce new inventions. In essence he used scientific abstraction to decompose machines into their component parts. These components could then be analyzed and recombined in novel configurations to produce new inventions. Rumsey had a powerful visual imagination; few inventors in history have rivaled his ability to produce highly original ideas by picturing them in his mind.

Rumsey’s methods have been incorporated into modern engineeringdesign technique, usually as aids to creativity. First he would gain an abstract understanding of the elements of a machine. Then he would manipulate these elements by mental operations similar to those employed by scientists in considering the spatial relations of geometric figures or the interactions of material bodies. That is, he would combine mechanical elements, separate them, reverse their roles, change their spatial relationships (e.g., move an object from vertical to horizontal, turn it upside down, etc.), reverse the direction of motion or flow, or reverse the motion of an entire machine (that is, imagine it running backward).

Reversibility has proved an especially powerful analytical tool for technologists, and Rumsey was one of the first to use it systematically. He apparently realized that just about any machine, if reversed, will result in another machine. When he invented his reaction wheel, he saw that upon reversal it became a centrifugal pump. This insight was not shared by other engineers until many years after Rumsey’s death. He likewise perceived that if the direction of the water’s flow in an Archimedean screw pump is reversed, it will become a sort of turbine (that is, if water is poured in at the top, it will produce mechanical power through its rotation). If a large Archimedean screw pump had its casing removed and were laid on its side, it could be employed as a stream or tidal mill.

He also found new possibilities by combining components or systems. He linked steam engines not only with boats but with wagons. He invented his reaction mill by combining the method of delivering water employed in the column-of-water piston engine (in which water at high pressure was delivered through an iron pipe) with the rotor of Barker’s mill. This enabled him to introduce the water from the bottom of his rotor, lessening the weight on the bearing on which it spun—a major disadvantage of Barker’s design—while utilizing high falls of water.

Rumsey also could rearrange the spatial relations of the parts of a design. In adapting his old pole boat for steam, he made a patent drawing that pictured the steam cylinder in the conventional upright position, but he soon realized that a steam cylinder could work as well on its side, so he suggested that arrangement to allow the engine to connect directly to the propulsive mechanism (the wheeled cart carrying the poles). In the case of his steam wagon (never built), he proposed likewise to connect the piston of a steam engine directly to the wheels. His insight was critical in the later evolution of the Western steamboat. The direct connection of a horizontal cylinder to the paddle wheels had a number of advantages; among them, it allowed the engine and boiler to be located on the same deck, eliminating the need for a deep hold. Thus the shallow-draft steamboats so familiar to us were based on an arrangement of boiler, cylinder, and drive evidently invented by Rumsey.

Rumsey arrived at still other inventions by reversing the roles of material elements. His watertube boiler is a case in point. Other innovators, including James Watt, had proposed to heat water for a boiler by means of fire tubes immersed in water. Rumsey put the water inside the tube and the fire outside. This simple operation increased the relative surface area of water exposed to heat, producing faster vaporization and a more efficient boiler better suited to high pressure. Rumsey elaborated on his watertube boiler while in England and anticipated several features widely adopted later.

Rumsey made frequent use of analogy. He exploited analogies between steam and water. To cite one of many examples, he designed a steam version of Barker’s mill; a century later a variant of this device fascinated Carl Gustaf Patrik de Lavai, the Swedish pioneer in steam turbine development. When Rumsey arrived in England in 1788, he was impressed with the dry docks he observed and saw that’a floating dry dock was analogous to a boat; it might be used not only for repairs but also for transport of smaller boats or ships along shallow waterways. He suggested a flat-bottomed hull form similar to the landing craft of World War II, with doors in front to enable other vessels to enter and leave. In basic terms of buoyancy, his dock was quite similar to the shallow-draft Western steamboats that appeared later. He proposed to power his dock with a steam engine but did not illustrate this in his patent drawing. The centrifugal pump he invented by reversing his reaction wheel would presumably have been used to pump out water after a vessel had been taken on or discharged.

James Rumsey did not invent the steamboat. No one person invented it. It evolved from idea to working reality through the efforts of many people over more than a generation. Rumsey’s contributions were neither few nor minor, however, and no one worked harder to realize the dream of steam navigation. The horizontally mounted steam cylinder and the watertube boiler both came to be widely adopted. Rumsey quickly grasped (and published) the basic buoyancy conditions needed by a flatbottomed vessel for use in shallow waters. And although reaction propulsion was a blind alley, it was a potential that had to be explored. Rumsey was a pioneer who, at the price of great hardship, blazed a trail that others were to follow to success.

It was Rumsey’s reaction-wheel designs that had the greatest impact on American technology. By 1850 reaction water wheels—essentially Barker’s mills within wheels—had become the workhorses of American industry. In 1851 Scientific American commented that “the subject of reaction wheels is more interesting than any other, because there are twenty of such wheels used in America for one overshot or breast wheel, we suppose.”

Rumsey’s reaction-wheel work brought together two threads in his thinking about waterpower, his use of reaction and his use of water pressure. Among the inventions that he had presented to the American Philosophical Society in 1788 was a sawmill driven by the pressure exerted by a column of water. He had realized then that higher efficiencies could be obtained if a mill was operated by pressure rather than simply by the impulse of moving water. (This is because when a stream of water hits something, its smooth motion becomes turbulent, often involving a significant loss of energy.) But the 1788 mill, in which the water pressure drove a piston, was efficient only at a slow speed. It would have been uneconomical, and the Rumseian Society reported a lack of public interest.

Possibly borrowing from Desaguliers, Rumsey then invented a family of more conventional waterwheels that would operate by pressure rather than impulse. To reduce wasted energy, the vanes were constrained to move slowly in tightly fitting channels. These slow mills were also uneconomical. Before he left for England, he had the idea of trying a wheel-like version of his spinning reaction mill, but he at first laid it aside, failing to see that reaction wheels could operate by pressure without the sacrifice in speed that other water-pressure motors required. He also left behind in America some sort of outline or sketch for a “water fly wheel,” and he mentioned a “wheel” version of his reaction mill in his British patent of 1790, which included some drawings that suggested wheelshaped motors. But he apparently did not get around to working out a completed design until he was forestalled by a rival American inventor, James McComb, of Princeton, New Jersey.

McComb received an American patent in 1791 for a two-outlet reaction wheel—a sort of Barker’s mill in the form of a wheel, whose outward shape was approximately that of a cartwheel, with two outlets on opposite sides of the rim; McComb called it a “hollow wheel.” Rumsey designed a flume to make comparative tests of models of his original mill and McComb’s hollow wheel. He found that McComb’s wheel was far superior, particularly in its ability to operate underwater. Conventional waterwheels were often brought to a halt by a rise in the downstream water level (termed backwater ). The hollow wheel’s ability to operate in backwater would allow the continuous operation of mills and factories at times of high water, thus giving it an important competitive advantage. In 1792 Rumsey received a British patent for a similar two-outlet wheel of his own.

Not long before his death, Rumsey began making improvements in the hollow wheel. His comparative tests probably revealed another advantage it enjoyed over his mill: its larger outlets gave it a higher “capacity”- that is, it processed more water in a given time. High capacity meant smaller machinery and lower cost for a given power output. Rumsey increased the capacity further by increasing the number of outlets from two to six, arranged symmetrically around the wheel’s periphery.

Rumsey next devised a version of his six-outlet reaction wheel with six curved pieces of wood inside it forming six vanes to channel the water. He didn’t live long enough to prepare an analysis of the wheel, but he must have had some inkling that the internal vanes or channels would serve another purpose, one that would unite the two major focuses of his thinking about mills: reaction and pressure.

By introducing internal vanes, he raised the possibility that the reaction might take place inside the wheel. In such a wheel the water could exert its force of reaction by pressure against the vane. This would, in principle, eliminate the need to expel water at high velocity. Rumsey’s vaned reaction wheel carried the seeds of the ideal water motor, capable of approaching 100 percent efficiency—the kind of motor to which the French were to give the name turbine .

Rumsey died in London in 1792, at the age of forty-nine, shortly before he was to try out a new steamboat, and was buried by friends in an unmarked grave near Westminster. His reaction wheel did not die with him; it was the starting point for very productive American work on reaction wheels. Between 1799 and 1837 fifteen horizontal waterwheels and fifteen “reacting” waterwheels were patented; the patents for almost all of them were lost in the Patent Office fire of 1836. But even after that date a number of reaction wheels very similar to Rumsey’s six-outlet vaneless wheel were patented.

In a few cases Rumsey’s influence can be clearly documented. There seems little doubt that even McComb’s wheel was inspired by Rumsey’s work, directly or indirectly. In anticipation of broad patent coverage, the Rumseian Society had displayed models of his inventions and publicized his ideas in every possible way, with the result that a number of imitators sprang up. The influence of Rumsey and McComb may be seen in the names given to three subsequent wheels: in 1808 Joel Farnham, of Oswego, New York, patented an improvement “in McComb’s horizontal reacting water wheel” in 1814 David Cooper, of Jericho, New York, patented an improvement of “Farnham’s reaction wheel”; in 1817 Theodore Burr, a distinguished inventor of truss bridges, patented a reaction wheel labeled an “improvement of Rumsey and McComb’s.” It is not common or prudent for inventors to advertise their indebtedness to prior patents in the title of their own. It is therefore only fair to conclude that Rumsey’s direct and indirect influence extended to many other patents that did not explicitly acknowledge their antecedents.

Rumsey was one of the first of a long line of American inventors who developed and used vernacular science as an aid to invention. It is time to end the long neglect of Rumsey. He deserves better from his countrymen.