Engineering The Erie Canal
Thomas Jefferson had a good eye for real estate on a grand scale. But when the notion of a canal linking the Great Lakes with the Hudson River near Albany, New York, was put before him in 1809 by two New York State legislators, he dismissed it out of hand. “Why, sir,” he said, ”… you talk of making a canal three hundred and fifty miles long through a wilderness! It is a little short of madness to think of it at this day!”
The idea of the Erie Canal tended to arouse that kind of skepticism wherever it was broached, and it has been difficult for later historians writing about the canal not to disparage its many early critics. The Erie Canal turned out, after all, to be a resounding success, perhaps the single most important public work ever built in the United States. But at the time, its opponents were merely being prudent. There were many more reasons to believe the Erie would fail than that it would succeed.
Canal building was hardly an advanced craft, much less a science, in early nineteenth-century America. The two most ambitious artificial waterways attempted prior to the Erie were the twenty-two-mile-long Santee Canal near Charleston, South Carolina, and the twenty-seven-mile-long Middlesex Canal linking the Merrimack River with the Charles River and Boston Harbor. Neither had much to commend it. Completed in 1800, the Santee took eight years to build and was a fiscal nightmare. It wasn’t much of a canal either. Political cronyism had pushed its path away from natural water sources, so that eventually two-thirds of its channel lay bone dry. The Middlesex, finished in 1803 after nine years of work, didn’t pay a dividend to its stockholders until 1819.
The Santee and the Middlesex were built at the approximate rate of two miles per year. If they were any yardstick, the Erie, with its 363-mile length and its significantly larger channel (forty feet wide and four feet deep) might well be open by the millennium.
The most serious impediment to progress on both the Santee and Middlesex works was the lack of trained engineers. In the early nineteenth century there was not a single native-born engineer in America. The Santee Canal had been engineered by a peevish and none-too-honest Swede named Christian Senf. The Middlesex began in American hands, but the local magistrate chosen to lay out the canal line quickly proved incompetent, and an Englishman named William Weston was called in as consulting engineer.
Of the handful of European engineers working in this country around 1800, Weston was the most active, but he was not a man of prodigious talent; many of the structures he designed for the Middlesex Canal simply fell apart. The historian Elting E. Morison writes of Weston: “Knowing not much, he knew a great deal more than anyone else and was in frequent demand.” He was in demand once again in 1816, when—after years of discussion, debate, and political maneuvering—New York State seemed poised to proceed with its canal undertaking. The state’s board of canal commissioners offered him seven thousand dollars to come from England and oversee construction of the Erie, but Weston declined on the grounds of his advancing age and his desire to stay with his family.
At that point the canal commissioners discontinued their search for foreign expertise and instead appointed four residents of upstate New York—Benjamin Wright, James Geddes, Charles Broadhead, and Nathan S. Roberts—to be principal engineers on the canal. None of these men had ever even seen a canal before. Wright, Geddes, and Broadhead were judges. They knew surveying because such knowledge was useful to magistrates when they heard property cases. Roberts was a schoolteacher who had taught himself surveying at Wright’s urging.
But simple surveying, the kind that goes into making boundaries, is of limited use when it comes to building a canal. Canal engineers must be able to measure elevations with a precision that allows for vertical errors measured in inches over horizontal stretches measured in dozens of miles. None of the four principal engineers appointed to building the Erie had ever taken a level before. But within a year they had taught themselves well enough so that when Geddes and Wright ran levels by different routes from Rome to Syracuse in the spring of 1818, enclosing a loop of one hundred miles, their final readings differed by less than two inches.
In all the many other details of canal building, they learned as they went, becoming engineers long after the title had been conferred on them. This was as true for the younger engineers on the canal as it was for men like Geddes and Wright. Virtually every American engineer of consequence during the first half of the nineteenth century learned his profession either on the Erie Canal or from an engineer who had been there. The Erie was truly, as a number of historians have said, America’s first school of engineering. Men learned things there because they had to, and they learned them in whatever way they could: by mistakes, by watching and asking questions, and by accepting expert authority without regard to rank.
In 1818, for example, a minister and amateur mathematician named David Bates was serving as the resident engineer along a stretch of the canal east of Syracuse, and a young farmer named John Jervis was working as a target man in Bates’s surveying party. Jervis had gained a rudimentary knowledge of leveling the previous year while felling trees for the canal and had buttressed this knowledge by studying two books on the subject. He ended up teaching Bates how to measure elevations.
Bates learned well from his subordinate. Later in life he was the principal engineer of the Ohio canal system. Jervis went on to become one of the greatest American engineers of the nineteenth century. Years later he wrote that on the Erie “the mechanical department of engineering was practically in its infancy.… The plan for a timber trunk for the aqueducts was prepared and submitted by a carpenter, Mr. Cady of Chittenango. This plan was adopted in nearly all the wood trunk aqueducts on the canal. At this day it stands as a well designed plan.”
By 1819 local contractors and mechanics like Cady working on the first section of canal under construction—the ninety-four miles between the Seneca River and Utica—had invented three immensely valuable laborsaving devices. The land they were clearing was thick forest, and without their new machinery, the entire enterprise might have ground to a halt early on.
The first of the three inventions made it possible for one man to fell a tree of any size without using a saw or an ax. The worker would secure one end of a cable to the trunk of a tree some sixty feet above the ground and the other end to a roller turned by a gear with a crank. After anchoring the apparatus to the ground one hundred feet from the base of the tree, the worker would turn the crank. The tremendous leverage obtained by fastening the cable so high up made it only a matter of time and exertion before the tree crashed to the ground.
The stumps left behind could be extracted by another local invention. It rested on two huge wheels sixteen feet in diameter joined by an axle almost two feet thick and thirty feet long—in other words, a fair-sized tree. Midway along this axle-tree was a smaller wheel, fourteen feet in diameter, with its spokes firmly united to the axle barrel. A rope was fastened to the rim of this middle wheel, wound around it several times, and its loose end attached to a yoke of draft animals.
The middle wheel would be positioned almost directly over a stump, the two larger wheels braced, and a chain made fast to both the thick axle and the stump. When the team of horses or oxen pulled on the rope, the rotation of the wheel made the entire axle turn, winding the chain around it and gradually uprooting even fresh, green stumps. With this huge machine, seven men and a pair of horses could pull thirty or forty stumps in a day.
The third invention was a plow with a heavy piece of sharpened iron attached to it; when draft animals pulled the plow, the plate traveled below the ground, cutting through roots as thick as two inches so that they could be easily scraped away.
From almost the beginning, plowing and scraping were the preferred method of excavation on the Erie, since the continual traffic of men and animals packed and strengthened the banks in a way that shoveling and carting could not. Spades and wheelbarrows did have to be used when the ground was wet, but on the Erie new spades were , designed to cut through roots more easily, and new wheelbarrows provided greater ease in carting dirt away.
All these means for clearing and shaping the land were in use by 1819. So was a well-organized system—based on the accountability of contractors—for letting many small excavation and construction contracts to private citizens along the canal’s route (the state put up major structures, such as aqueducts and dams). There was an ample supply of local labor, supplemented with Irish immigrants shipped north by New York City’s Tammany Hall. There was a corps of engineers whose diligence more than compensated for their inexperience.
And yet despite all that, the Erie Canal might have been a failure—even at that early stage—for lack of one vital commodity: material for building durable locks. If the locks were built of wood, they would rot in a few years. Good stone locks needed hydraulic cement for waterproof mortaring, but the only known sources of hydraulic cement were in Europe, so the cost of that would be prohibitive. The only solution was to build the locks by uniting stone to stone with ordinary mortar and applying a thin coating of imported hydraulic cement at the joints between them. It was a concession to the apparently inevitable: The Erie’s locks were destined to fall apart fast. The only question was how fast.
That question became academic almost as quickly as it became critical. In 1818, quite by accident, contractors along the canal line discovered natural cement rock. It was also discovered by a Herkimer County physician named Andrew Bartow, who demonstrated its potential for the benefit of Benjamin Wright and Wright’s chief assistant engineer, Canvass White. In a tavern in the village of Chittenango, Bartow mixed the pulverized rock powder with sand and placed a ball of it into a bucket of water. By morning the mixture had hardened to the point where it could be rolled across the floor like a stone.
Canvass White, easily the most gifted engineer on the canal, had spent the previous winter in England—at his own expense—studying existing canals and learning about hydraulic cement. By the start of the 1819 construction season, he had perfected the process for refining this local rock into true cement powder. By the time the canal was completed, more than four hundred thousand bushels of it had been used. It firmly held together every bit of masonry on the canal, from mundane little culverts to gigantic aqueducts—and, of course, good stone locks. (The gates were of wood.)
White’s discovery exploded on the scene so quickly that his patent on the process was conveniently ignored by all the manufacturers. Eventually the state legislature considered awarding him ten thousand dollars in compensation. He was entitled to at least six times that in royalties, but even the attempt at partial reparation fell through. It was a typical outcome for White. Perhaps an engineering genius, he was luckless in his financial affairs. He died young, in 1834, leaving his widow little more than the furniture she was compelled to sell.
The discovery of native hydraulic cement came as the building of the Erie Canal was about to enter a more technically difficult phase. The middle section had been chosen as the starting point for construction in 1817 because it offered advantages to both the engineers and the pro-canal politicians. Politically the advantage of starting in the middle was twofold. Results, measured in navigable canal miles, could be effected quickly there and then used as leverage to obtain more state funds for further work. At the same time, as the middle section was completed, popular support for the rest of the canal would grow in the areas to the east and west.
Engineering a canal forty feet wide and four feet deep through the ninety-four miles of wilderness between Utica and the Seneca River was no mean technical feat, but the problems it presented were minor compared with the ones that lay in the later segments. If the canal was a school of engineering, the middle section offered the appropriate introductory course.
That course was completed in the fall of 1819, and the middle section was opened for navigation the following spring. Along the 270-odd miles of unfinished channel remaining, the most striking engineering problems (and solutions) were to be found in the 158-mile western portion, between the Seneca River and Buffalo, but the more difficult, if less spectacular, engineering had to be done in the east. There the canal dropped 419 feet in the 109 miles between Utica and the Hudson River; between Lake Erie and Utica the drop was only 146 feet over 252 miles.
The magnitude of the descent in the eastern section would have offered enough of a challenge by itself, but the difficulty was compounded by the inhospitable topography of the Mohawk River valley, through which eighty-six miles of the canal’s line had to be laid out. The banks of the Mohawk were cramped by steep hillsides, which in places ended at the water’s edge. This required construction of the canal channel in the river itself, supported on a masonry base and protected by high embankments. The canal’s eastern section was a potential nightmare, and it fell to White, by now a principal engineer of the canal in everything but title, to make the best of it.
The key to his solution lay in his placement of locks, which automatically determined the location of the pound levels—the stretches of channel between the locks. In order to take advantage of the better line available on the north bank, he ran the canal across the Mohawk four miles below Schenectady on a 748-foot-long aqueduct. Twelve miles farther east, at the Cohoes Falls, he recrossed to the south bank via a 1,188-foot-long aqueduct. White saw the work through in three years. By the end of 1823, the Erie was open from Brockport, some twenty miles west of Rochester, all the way to the Hudson River at Albany.
At that point, about eighty miles of channel, between Brockport and Buffalo, awaited completion. Already standing were two of the western Erie’s engineering triumphs: the Irondequoit Embankment and the Genesee River Aqueduct.
Not far east of Rochester, an unexceptional stream called Irondequoit Creek had carved out a valley and an engineering challenge. Taking the canal across it without adding about one hundred and fifty feet of up-and-down lockage was imperative. The only thing that made the task even remotely possible was the presence of several natural ridges that could carry the canal at least partway over the valley it would have to span. James Geddes had long advocated linking these ridges together with great earthwork embankments and running the canal across the top; the canal commissioners were hesitant to approve so bold a plan, but finally realizing that they had few real options, they authorized work to proceed as Geddes had proposed.
The Irondequoit Embankment, built entirely during the season of 1822, consisted of three natural ridges joined together by two man-made ridges, one 1,320 feet long and the other 231 feet. The canal ran along the narrow summit for 4,950 feet, passing 76 feet above Irondequoit Creek, which flowed through a 245-foot-long culvert. Since the valley’s soil was unsuitable for such enormous earthworks, small mountains of earth had to be hauled in from elsewhere. Even so, there was no great confidence that the embankment would hold up; from its completion in October until the close of the 1822 season, the work was drained nightly.
A few miles farther west, in Rochester, a stone aqueduct carrying the canal over the Genesee River was completed in 1823; its combined span of 802 feet made it the second longest aqueduct on the canal. But impressive as the aqueduct and the embankment were, the engineering work that captured the most attention lay about sixty-five miles to the west, in Lockport, where the canal had to be lifted sixty feet up onto the Niagara Escarpment and where, for two miles, its channel would be blasted out of solid rock.
The work there began in 1822 and took three years. What stood at Lockport upon completion of the job were five double locks, one set of five for going up, a parallel set of five for going down. West of the locks, which quickly won popular renown as the Lockport Fives, the channel ran for seven miles through the Niagara ridge. To get through the most difficult portion—the two miles known as the Deep Cut—workers had to blast free and haul away nearly 1.5 million cubic yards of rock.
While progress was being measured in feet and inches at Lockport, the final section of the canal, between Lockport and Buffalo, was being built more quickly—but nevertheless to an extraordinary standard of care. To propel water down from Lake Erie, the fifty miles of channel from Buffalo to Lockport were sloped at exactly one inch per mile. By the summer of 1825, the job was done. A few final details remained; then, on October 26, 1825, the Erie Canal formally opened amid statewide ceremony and celebration that lasted for weeks.
The ceremony and celebration ended with the year, but the effect of the canal on America had just begun. The Erie had cost the state about $7.9 million to build, but it attracted such a huge volume of commercial traffic that it paid for itself through toll revenues in less than ten years. Its awesome vitality as an avenue of commerce catapulted New York City into the position of preeminence that Philadelphia had always assumed would be its own. It turned western New York State from a wilderness into a prosperous country of farms, towns, and busy manufacturing cities. And as emigrants passed over the canal heading west, it had the same civilizing and nurturing influence on Ohio, Indiana, and the other states of the Old Northwest. The Erie appreciably advanced the timetable of American development.
But what if the canal had not been built? What if the task had proved too great and the work had been abandoned?
In his definitive History of the Canal System of the State of New York , written in 1905, Noble E. Whitford indulged in some fascinating speculative history. Without the Erie, he wrote, Canada would have been “enriched … commercially and strategically almost in proportion as it would have tended to impoverish us.” With the completion of the Welland Canal across the isthmus between Lake Erie and Lake Ontario in 1831, the already established tendency of Northwestern trade to gravitate up the St. Lawrence would have been greatly accelerated. Moreover, Canada would have gained strategic control of the outlet to the Great Lakes, making it possible for the British government to translate that advantage into naval control of the lakes. All this at a time when their vital importance in the War of 1812 was still fresh in the memory.
Within the United States, the natural outlets for the northwestern trade were the Ohio and Mississippi rivers, and prior to the building of the Erie Canal, trade was drifting south along those routes just as it was drifting north to Montreal. Without the Erie Canal and the impetus it gave for the building of other canals and, later, east-west railroads, geographic expediency would have routed that trade north and south. “Chicago could hardly have become so great an emporium,” wrote Whitford, ” … and not a little of the commercial prestige of Boston, New York and Baltimore … would then, perchance, have descended upon New Orleans and Mobile and Galveston. More portentous still than this commercial alliance between the Northwest and the South is the consequent probability that out of it there would have grown racial sympathy and political kinship, with what effect upon the great issues which culminated in the Civil War or upon the present constituency of the American land and people, we can only conjecture.”
But the canal had been built, and there was no conjecture as to its benefits. The men who built it found themselves in high demand as “canal fever” swept the country in the wake of the Erie’s success. Benjamin Wright would still be referred to as the father of American engineering had he retired after the Erie, but he served as either the chief engineer or the consulting engineer for practically every major canal built in the United States for the next sixteen years, and for the Harlem and Erie railroads as well.
James Geddes was sixty-two when the Erie was done, but he continued canal work in Ohio and Maine before retiring. Canvass White served as the chief engineer of several canals, including the Delaware and Raritan. Nathan Roberts, who designed the famous Lockport Fives, was chief engineer on the Pennsylvania State Canal and at Muscle Shoals on the Tennessee River. After serving as chief engineer of the Ohio Canal System, David Bates moved on to success in railroad engineering. Before he too moved on to railroads, John Jervis’s triumphs included the Delaware and Hudson Canal and New York City’s Croton Aqueduct.
They did not build a perfect canal— no one does. The old Erie required a great deal of maintenance and repair, and it was alternately bedeviled by floods along streams that fed it or crossed its path and by low water due to leakage through its bed and banks. In 1836 the state began a twenty-six-year program of enlarging and improving the canal, guided by an authoritative, canal-long survey led by such men as John Jervis and Nathan Roberts. The survey found problems along every section of the canal, and some required substantial changes.
None of this detracts from the original accomplishment. That there was a profitable and maintainable canal in operation at all in 1837 speaks volumes about the kind of men who put it there. The few of them who had any technical knowledge at all had been nothing more than plain old-fashioned country surveyors before the Erie began. The rest had been farmers, craftsmen, merchants: the ordinary settlers of a wilderness. To do what they were asked to do, they had to reinvent themselves, and in reinventing themselves, they accomplished something unimagined and extraordinary. They invented the American engineer.