Tales of Distortion: Sauvé's Crevasse
It's time to get back to my roots.
A privately maintained levee should not flood a city seventeen miles away from it.
A levee is a long embankment whose primary job is to keep water on one side of itself.
This is a good responsibility, but in Louisiana it is also an ambitious one.
On May 3, 1849, a guy named Pierre Sauvé owned a sugar plantation on the eastern bank of the Mississippi River, about seventeen miles upriver from New Orleans.
- Sauvé had a levee.
- New Orleans had houses, streets, businesses, canals, municipal governments, and around twelve thousand people who would soon become personally interested in Sauvé's levee.
The Mississippi River had been rising for weeks.
There was about to be a disagreement.
The Levee.
In 1849, a lot of the lower Mississippi's flood protection was maintained through individual landowners.
Plantations along the river were responsible for their sections of the embankment. Local inspectors and public regulations existed, but there was no integrated state or federal system maintaining a continuous engineered barrier to a uniform standard.
Sauvé's plantation was a sugar operation. The labor of building and maintaining its defenses depended upon enslaved workers, like much of the surrounding plantation economy.
- The owner held the property and its formal responsibilities.
- The workers supplied the labor.
- The levee held back the river.
This little arrangement worked for the ordinary purpose of keeping fields and structures dry, provided each section remained sufficiently strong.
There was one significant limitation.
- The levees belonged to different properties.
- The water behind them belonged to one big river.
The Mississippi was not physically divided into neat sections matching the responsibilities of the landowners whose slaves maintained its banks.
A weak section could therefore compromise a very large area whose residents had neither constructed nor controlled that section.
This became especially troublesome at Sauvé's plantation.
The property stood along a bend in the Mississippi, in a landscape through which the river had historically sent water away from its principal channel. During the spring of 1849, high water and erosion were working against the bank.
Other levees were failing too.
On April 17, the Fortier levee across the river gave way. Its water spread mainly across agricultural land and eventually found a route toward Lake Cataouatche.
Sauvé's levee had less agreeable geography behind it.
By May 3, the river had weakened the embankment enough to breach it.
Water began passing through the opening.
The flow eroded more earth.
More earth disappeared.
More.
More water entered.
More.
The opening eventually grew to approximately 150 to 164 feet across and more than six feet deep, according to later geographic reconstructions.
This is not the right amount of hole to have in a levee.
A levee is valuable because it has a continuous top.
You can stand on that top and observe that the water is on one side while the things you care about remain dry on the other.
A 150-foot section of missing levee changes this relation.
The river now has an entrance.
- Water enters the opening.
- The flow removes more soil.
Then more.
- The opening becomes larger.
Then larger.
- More river becomes available to the land behind it.
Sauvé's levee was no longer performing its assigned function.
This thing was beginning to perform an entirely different function.
The river had found a new distributary.
Unfortunately, New Orleans was in the way.
The Ground Had Been Here Before.
To understand where all this water went, we need to look at what New Orleans was actually built on.
The Mississippi River made much of this landscape.
Over thousands of years, it carried sediment downstream and deposited it when its water slowed or escaped its banks. Heavier sand and silt accumulated nearer the channel, producing elevated natural levees. Finer material settled farther into the surrounding wetlands, where swamps and low ground developed behind the higher banks.
The river also changed course.
An abandoned channel could leave behind an elevated ridge. The water might go, but the sediment stayed, preserving part of the geometry of an older river.
This is how the city acquired much of its ground.
New Orleans was built on the work of a river that had spent thousands of years refusing to stay in one place.
Then people began asking it to stay in one place.
The river did not sign anything.
Behind Sauvé's plantation lay low, swampy terrain. North of it stood the Metairie Ridge, formed along an old Mississippi distributary.
When the levee broke, water collected in the low ground between the river and the ridge.
There was a large lake to the north.
Lake Pontchartrain.
Things would have been very simple if the flood could simply have continued toward that lake and left the populated parts of New Orleans well enough alone.
But the Metairie Ridge made this difficult.
This ridge was high enough to obstruct the most direct northward escape.
So the water spread east.
Toward New Orleans.
The river had built this ridge during an earlier part of its life.
Now the ridge was redirecting the river's newest flood toward a city that had not existed when the Metairie formed.
A piece of the Mississippi's ancient architecture had acquired a new role in the disaster.
It was doing its job just fine.
The Canals.
New Orleans had canals.
This was very understandable. The city occupied a low, wet, deltaic landscape in which transportation, drainage, and the control of water were extremely important.
Canals could move boats, connect bodies of water, and carry water through routes that human beings had deliberately opened.
These things also had embankments.
Those were useful for maintaining the canal's own course and protecting surrounding land.
The flood arrived at the New Basin Canal on May 8.
Its guide levees were overtopped. Water crossed them and continued moving.
The canal had become part of the flood's route through the landscape.
The water eventually passed Broad, Claiborne, and Dryades streets. By May 15, it had reached Rampart Street. As the flood spread through the lower sections of the city, it reached as far as parts of Bourbon Street.
By May 30, it had attained its greatest urban extent.
Here is the interesting part.
Not all the embankments failed.
The raised track bed of the New Orleans and Carrollton Railroad, an ancestor of today's St. Charles streetcar line, helped obstruct the water in part of the city.
The guide levees of the Old Basin Canal, also called the Carondelet Canal, successfully blocked the flood from extending farther east.
The First Municipality had reinforced part of that protective embankment as the water approached.
These were local successes.
- The Metairie Ridge blocked the water north.
- The New Basin Canal's defenses were overtopped.
- The raised railway bed helped block the water elsewhere.
- The Old Basin Canal's embankments held.
None of these objects needed to know anything about municipal responsibility to affect the outcome.
The ridge followed its elevation. The railroad embankment followed its geometry. The canal levees followed their actual structural condition.
The water followed the resulting paths.
This is the difference between the mapped purpose of infrastructure and its actual participation in a changing field.
A railway was now a flood barrier.
A canal's protective embankment had become a failed obstacle. Another canal's embankment had become a successful one. The same kind of object could perform radically different work depending on its location, elevation, condition, and relation to the incoming water.
New Orleans had acquired a disaster large enough to reveal which parts of its engineering still worked.
The City.
The flooding reached approximately 220 inhabited city blocks.
Around 2,000 structures were affected.
Historical estimates place the number of residents flooded or displaced at roughly 10,000 to 12,000.
Some of the deepest inundations reached six feet, although most flooded homes experienced considerably less.
The physical footprint was enormous because so much of the surrounding land was still swamp, agricultural terrain, or undeveloped ground.
The flooded population was smaller than the image of an entirely submerged city might suggest.
A map can show water extending across much of a landscape without telling you exactly how many people lived in the inundated area, how deeply their homes were flooded, or what resources they possessed to respond.
The worst residential impacts fell disproportionately upon people living in the lower rear sections of the city, including poor immigrant communities and free Black residents.
For people, the flood was not an impressive line on a geographical diagram.
There was water inside the house.
Furniture could be damaged or lost.
Food and possessions needed moving.
An ordinary street could cease to function as a street.
Travel became a problem of finding a boat, making an elevated walkway, or identifying some other route that remained passable.
The flood had changed the actual transportation system of the city.
A person might still know where their employer, grocer, church, or neighbor lived.
They could no longer necessarily get there.
Boats began moving people and supplies through the streets.
These boats had previously belonged to a city whose streets were expected to remain streets.
They now belonged to a city in which certain streets had become waterways.
Their names stayed the same.
The transport requirements did not.
There is no strong record of deaths directly attributable to the crevasse itself, which is a remarkably fortunate feature of this disaster.
It is also no reason to confuse the flood with an entertaining interruption.
The city was filling with water.
Its people were doing the work of continuing inside it.
And the government was having a problem.
Three Governments.
New Orleans in 1849 had one mayor.
It also had three municipal governments.
This arrangement had been established by the charter of 1836, which divided the city into separate municipalities with their own councils and many separate administrative functions. A General Council retained certain citywide powers.
There were historical reasons for this unusual design, including the political and cultural divisions among the city's different communities.
- The First Municipality governed the older French and Spanish portion of the city.
- The Second governed the newer American section.
- The Third governed the area east of the old city.
- The neighboring city of Lafayette was separate.
All of this had been formally arranged.
The city had councils, offices, surveyors, records, powers, and boundaries.
These arrangements had a problem with Sauvé's hole.
Sauvé's plantation was outside the city, in Jefferson Parish.
The hole was therefore outside the municipal jurisdictions.
The water entering New Orleans was not.
The municipal authorities initially disputed responsibility for closing a breach beyond their limits. The property owner could attempt repairs, but the scale of the flood soon exceeded the practical capacity of his estate.
A city could authorize work within its own territory.
It could build a protective levee, arrange relief, and appoint engineers.
What it lacked was an immediately effective, coordinated response to the opening in the earth that was still supplying the flood from outside the city.
The First Municipality's defensive work along the Carondelet Canal demonstrates that the governments were capable of real local action.
The problem was getting the various authorities to act on the actual relation producing the disaster.
Eventually, officials from the municipal governments assembled a joint committee and organized engineering interventions.
This was progress.
By then, the river had been making progress for several weeks.
The Citizens Begin Building Levees.
Some residents decided to protect their properties themselves.
A perfectly understandable decision.
If the street outside your house is flooding, and an earthen barrier might keep water from entering your home, constructing an earthen barrier may be the best immediate option available.
So people began building makeshift levees.
These worked in some places.
Unfortunately, the water displaced by one person's barrier could enter another person's property.
The new levee protected one house by changing the geometry available to the flood.
The flood then entered another part of the street.
Their neighbor now had a problem.
That neighbor could build another barrier.
Or sabotage the first one.
Historical accounts describe people threatening one another, tearing down the defensive works of their neighbors, organizing posses, and taking up arms.
The city had begun reproducing its regional flood-control problem down at the scale of individual houses.
A wall could protect one property. But the water outside the wall still had to go somewhere.
A locally successful intervention could worsen the conditions of another resident without anyone having miscalculated the basic operation of a levee.
This is the much nastier problem.
The individual defensive works did not have to fail to produce a wider conflict.
Some of them worked.
That is why the neighbors were so angry.
E. B. White.
One New Orleans grocer deserves a little special attention.
His name was E. B. White.
The city's normal food-delivery arrangements had become unreliable, because, you see, many of its streets were flooded.
Because of that hole in that levee.
White therefore announced that he would continue delivering groceries to customers by boat.
This was an excellent adaptation.
- The streets were just full of water.
- Boats could travel on that stuff.
- Groceries could be put inside boats.
White had successfully translated the problem into a delivery service.
He also announced, according to historian Greg O'Brien's account of the episode, that he would refuse service to members of the various municipal councils.
White had apparently decided that the city's governing officials would not be included in his aquatic grocery program.
This is funny.
Whether this was a good commercial policy is another question.
It is difficult to imagine the municipal councils finding it satisfactory.
- The councils had legal authority.
- White had a boat.
During the flood, these were very different kinds of power.
Louisiana's Other Waterfall.
Meanwhile, the crevasse had become an attraction.
The river was pouring through a breach more than a hundred feet across, producing violent currents, spray, turbulence, and a substantial drop over the eroded bank.
Nineteenth-century observers compared the sight to Niagara Falls.
Of course they did.
Louisiana had managed to produce another impressive waterfall in a landscape not especially famous for its towering heights.
This was chronologically the first of our two Great Louisiana Accidental Waterfall cases.
The other would arrive in 1980, when Lake Peigneur learned about a salt mine, freaked out, and briefly rearranged the local hydrology to produce a waterfall of its own.
In 1849, the process was much simpler.
There was already a large river.
There was a wall.
Part of this wall was missing.
And so Louisiana had acquired a waterfall.
And somebody had discovered that people would pay to see that.
Steamboat excursions began operating to the crevasse.
An advertisement in May 1849 offered passage aboard the A. W. Vanleer on May 16 and 17. The boat would leave at four in the afternoon.
On the first day, the departure point was the foot of St. Louis Street.
On the second, it would depart from the foot of Canal Street.
The advertisement promised a cotillion band for the ladies.
It also specified that the excursions would not be postponed on account of weather.
The city was experiencing a major flood.
People had lost the use of their streets and homes.
The river was continuing to enter through a very large hole.
And the excursion operator wished to assure the public that adverse weather conditions would not interfere with their opportunity to visit this hole.
Tickets cost fifty cents each way.
Children traveled for half price.
A family attraction.
Passengers could board a steamboat, enjoy musical entertainment, approach the ruptured levee, and watch the Mississippi River enter the landscape with enough force to flood New Orleans.
Then at some point the steamboat would turn around and return them to the city.
That city would still be flooded.
The band had finished its performance.
Louisiana had monetized the crevasse.
The Engineers.
Back at the breach, the problem remained.
Water was still coming through.
The municipal committee had organized engineers to attempt repairs. They tried filling the opening with sandbags and other material.
These efforts failed.
According to a later historical reconstruction, they even attempted to sink a riverboat into the crevasse, hoping the sacrificed vessel would obstruct the rushing water.
This is a real method, in principle.
A large object can be sunk deliberately to obstruct a channel or support a barrier. Engineers have used ships, timber, stone, and other substantial objects in comparable work for centuries.
But this vessel has to remain in place.
And it has to obstruct enough of the opening.
And the water has to cooperate with the resulting geometry.
The dead riverboat did not close Sauvé's Crevasse.
This is an important distinction between the two great Louisiana hole disasters.
- At Lake Peigneur, the water would swallow boats without being asked.
- At Sauvé's Crevasse, people apparently tried to put one in the hole deliberately, to appease it.
So the waters of Louisiana may have discovered their dark hunger then. Neither was an especially good day to be a vessel regardless.
The engineers continued working.
They were also investigating other routes through which the accumulated floodwater might escape.
This led to some unusually informative administrative proceedings.
On May 24, engineer George T. Dunbar submitted a report concerning the D'Aquin and Labarre canals.
The relevant committee had apparently asked him to examine a proposed widening of the canals.
Dunbar surveyed the proposed work.
He concluded that excavating the channels to the required depth would cost more than $100,000 and require a large workforce for about a month.
He also doubted that the proposed enlargement would accomplish very much.
The elevation and connection to the surrounding basins were troubling. Digging a wider ditch was of little use if the ditch did not reach the water it was supposed to remove at a sufficient depth.
This is one of those very important features of canals.
A canal is a cut through land.
It is not an administrative promise that water will go where the cutting is most convenient to the council.
Dunbar recommended a much smaller intervention involving lateral cuts and removal of obstructions.
And so then the committee began disputing whether the original resolution had even instructed this guy to widen the canals.
The revised work was authorized under a limit of $1,000.
So the city had a proposed excavation costing more than $100,000, a much smaller plan costing no more than $1,000, a dispute about what had actually been ordered, and a flood continuing to rise. Things were under control.
This was all recorded in the Daily Crescent on May 25.
The newspaper also reported a visitor.
A man named H. Y. Nichols had arrived at the editorial office with a model of an invention he believed could close Sauvé's Crevasse.
This model used rows of supported piles arranged to resist the incoming water.
The editor thought it looked quite nice.
The model was apparently persuasive on its own terms. It contained supports, braces, and the principles by which the structure might remain standing.
The editor had, however, become reluctant to trust another engineering demonstration forged without contact with the actual conditions at the crevasse.
He ended with an observation:
"while the models are exhibited, the water is rising."
The Hole Finally Closes.
The repair work continued into June.
The river eventually fell enough to make the physical conditions at the breach more manageable.
Engineers and laborers succeeded in constructing a closure, described in later accounts as an effective floodgate or barrier.
On June 20, 1849, the crevasse was plugged.
Forty-eight days had passed since the levee opened.
The Mississippi finally stopped pouring through Sauvé's plantation.
The people of New Orleans could now begin the less spectacular business of removing a flood from a city.
Some of the water fled through canals and Bayou St. John.
Some soaked into the soil.
Some meekly evaporated.
Streets became passable again. Residents returned to damaged homes. Public infrastructure required repair, including pavements, gutters, bridges, wharves, and other works.
The Second Municipality would later impose a substantial special tax to meet flood-related expenditures.
The city had stopped the incoming water.
It still had the cost of everything the water had already done.
There was also something else left behind.
Mud.
Here is the part of the history that makes the final accounting more difficult.
The Mississippi River carried sediment into the city.
That sediment remained after the water receded.
This was the same basic process by which the river had built much of the surrounding delta in the first place.
Water escaped the channel.
Its velocity fell.
Sediment settled onto the low ground.
The deposited material became part of the landscape.
Richard Campanella identifies the 1849 flood as one of the last major episodes of natural Mississippi River sediment deposition within New Orleans.
The city had been damaged by a process that also supplied a material resource important to the long-term survival of delta land.
This is an awkward distinction.
Nobody needed to have their home flooded to receive a beneficial quantity of sediment.
The uncontrolled inundation was dangerous and destructive.
But simply forbidding all river water from reaching the surrounding wetlands also interfered with a process that had helped build and sustain the landscape.
As flood-control systems improved, the river was progressively confined more tightly to its channel.
The city became much better protected from Mississippi River flooding.
The delta also lost important routes through which freshwater and sediment had historically reached its wetlands.
That loss would interact with drainage, subsidence, development, canals, coastal erosion, and many other processes across the following centuries.
None of this means Sauvé's Crevasse should have been left open.
It means that closing one harmful route can remove a useful function carried by the same physical process.
That function may need another route.
A river can be prevented from entering houses without pretending that the land surrounding those houses no longer needs sediment.
Both problems belong to the same river.
And underneath parts of New Orleans, the flood of 1849 remains physically present.
The water left.
The sediment stayed.
The city built over it again.
A Better Levee.
Sauvé's Crevasse became a reference point for subsequent flood protection.
The height reached by the 1849 water was used to establish new protective requirements. Later Louisiana law instructed that levees along the New Basin Canal be raised to the level of the high water recorded during the disaster.
The divided municipal government also came to an end.
In 1852, New Orleans adopted a new charter unifying its three municipalities and incorporating the previously separate city of Lafayette.
This consolidation had causes and political pressures extending beyond Sauvé's Crevasse. But the flood had provided an extremely practical demonstration of the difficulty of coordinating physically connected public works through separate governing authorities.
The old arrangement could perform useful work within its parts.
It struggled when a dangerous transition began outside those parts and moved through all of them.
The new arrangements did not make the city immune to flooding, but they gave it different ways to organize a response.
Over the following decades, levees, drainage works, and larger engineering institutions would continue changing the physical relationship between New Orleans and the river.
Some of those changes saved lives and property.
Others introduced their own dependencies, tradeoffs, and long-term problems.
The river and the city remained.
The work of making them coexist continued.
The Ruling.
Sauvé's Crevasse was a hole in an earthen levee on private property.
The flood traveled seventeen miles.
The Metairie Ridge kept the water from following an easier path toward Lake Pontchartrain.
The city flooded unevenly.
Three municipal governments struggled to coordinate a response to a breach outside their jurisdictions.
Residents built their own levees and began fighting over where the water should go.
E. B. White delivered groceries by boat and refused to feed the councillors.
Steamboats took paying passengers to see the waterfall.
Somebody tried to put a boat in the hole.
Eventually, workers closed the breach.
The water receded.
The sediment stayed.
The governments changed.
A flood-control system that had assigned responsibility to individual sections of a continuous river gradually gave way to larger arrangements designed around the actual extent of the danger.
We have now examined two remarkable Louisiana holes.
- At Lake Peigneur, a drilling operation connected a lake to a mine that the administrative arrangements had failed to make sufficiently present in the transition.
- At Sauvé's Crevasse, a river passed through a private levee and entered a physical landscape whose connected consequences exceeded the governing arrangements responsible for stopping them.
- One map was missing below.
- The other was missing across.
- Neither gap was imaginary.
- The water found both.
No postponements on account of weather.
Sources
Richard Campanella, Disaster and Response in an Experiment Called New Orleans, 1700s–2000s, Oxford Research Encyclopedia of Natural Hazard Science, 2016. Principal geographic reconstruction, flood progression, inundation figures, and sediment consequences.
Richard Campanella, The Katrina of the Nineteenth Century Was Called Sauvé's Crevasse, Times-Picayune, June 13, 2014. Historical route and contemporary eyewitness reconstruction.
The Daily Crescent, May 25, 1849, page 2, Library of Congress. Contemporary reporting on Dunbar's proposed canal works, municipal committee proceedings, Nichols's model, and the rising flood.
Travis Lux, NOLA vs. Nature: The Other Biggest Flood in New Orleans History, WWNO, November 29, 2017. Historical reporting with Greg O'Brien, Craig Colten, and Richard Campanella; source for the grocery boycott, improvised defenses, steamboat excursions, and attempted riverboat obstruction.
New Orleans City Archives, Charter and Ordinance History. Documentation of the municipality period and 1852 consolidation.
John Kendall, History of New Orleans, Chapter X, 1922. Retrospective historical account of municipal defenses, flood extent, and engineering repairs.
1849 excursion advertisement, reproduced in the historical Sauvé's Crevasse collection and in WWNO's historical reporting. Source for the A. W. Vanleer, St. Louis Street departure, fares, weather policy, and cotillion band.