The steamboat Sultana exploded on the Mississippi River a few miles above Memphis at about 2 a.m. on April 27, 1865. It was carrying paroled Union prisoners in numbers several times its legal capacity, and it is still the deadliest maritime disaster in United States history. Nobody can say exactly how many died, because nobody made a reliable count of who was aboard. This piece traces where the information about that load existed and why it never reached anyone downstream with the authority to stop the boat. It also covers how thinly the news of the disaster traveled in the week after Lincoln’s assassination.

April 27, 1865: the boat and the parts of the record that are settled

The Sultana was a wooden side-wheel steamboat built in Cincinnati in 1863 and run in the lower Mississippi trade, carrying passengers and freight between St. Louis and New Orleans. In April 1865 its master and part owner was Captain J. Cass Mason. Its legal passenger capacity is usually given as 376. On April 24, 1865, at Vicksburg, Mississippi, it took aboard a very large body of Union soldiers who had recently been released from Confederate prison camps, mostly Cahaba in Alabama and Andersonville in Georgia. They had been waiting at a parole camp outside Vicksburg while exchange arrangements were worked out at the end of the war, and they were being sent north toward Cairo, Illinois, on their way home and out of the service.

The boat steamed north against a river in spring flood. It stopped at Helena, Arkansas, and then at Memphis on April 26, took on coal across the river from Memphis, and continued upstream after midnight. At about 2 a.m. on April 27, a few miles north of the city, its boilers exploded, and the fire that followed burned the wreck down to the waterline. The river at that point was high and cold and spread well beyond its normal banks. Many of the men aboard were weak from months of imprisonment, and survivor accounts describe men drowning in clusters as they grabbed at one another and at whatever floated.

Several points are not in dispute. These include the date and the general location, the fact that the boat carried far more people than it was licensed for, and the fact that most of the people aboard were paroled Union soldiers, along with civilian passengers and the crew. Captain Mason died in the disaster. Two questions remain disputed: how many people were aboard and died, and what exactly caused the boilers to fail. Retellings tend to settle both with a confident number and a confident villain, and the record supports neither.

For the narrative I rely mainly on Gene Eric Salecker’s book Disaster on the Mississippi, published in 1996. I also use Chester Berry’s 1892 collection of survivor reminiscences, Loss of the Sultana and Reminiscences of Survivors, and published descriptions of the military investigations held in 1865. Berry’s collection consists of survivors’ own recollections written down decades later, so I treat it as testimony rather than as a record of times and numbers.

The death toll: why the honest answer is a range

Published estimates of the number of dead run from somewhere around 1,200, or a little below, at the low end to more than 1,800 at the high end. Salecker’s book rebuilt the count from regimental records, survivor lists and hospital records, and it argues for a total toward the lower end of that range. Figures above 1,800 appear in older accounts and are still repeated in popular retellings. I have not been able to confirm which author first fixed the highest figure, so I am not attributing it to any one source. Estimates of the number aboard also vary, and most run above 2,000.

The reason for the spread is a communications failure that happened before the boat left Vicksburg. No complete roll or manifest was made of the men who boarded, so investigators in 1865 and historians since have had to work backward. Their sources are regimental rolls, the paroles issued at the camp, the lists of survivors taken in at Memphis hospitals, and the names that families reported as missing. Each source has gaps, and the gaps do not overlap neatly. A soldier who boarded without being entered on any list, and whose body went downstream and was never recovered, may appear in one reconstruction and not in another.

Some counting questions are matters of definition rather than of missing data, and different authors resolve them differently. One is whether to count survivors who were pulled from the river and then died of burns or exposure days later in Memphis. Another is how to count civilians and crew, whose records were kept separately from the soldiers’ records. These choices can move a total by a meaningful amount, which is why two careful researchers can work from much the same documents and still publish different figures.

For anyone using the Sultana in training or a presentation, the safe statement is a range, along with the reason the range exists. The disaster is often compared with the Titanic, and at the high end of the Sultana estimates the comparison favors the Sultana as the larger loss. At the low end it may not, which is one more reason to give the range rather than a single number.

Camp Fisk to the landing: how the men were loaded without a count

The released prisoners waited at Camp Fisk, a parole camp a few miles from Vicksburg, and were brought to the river in groups for transport north. The government paid steamboat owners to carry them. The rate commonly reported is five dollars for each enlisted man and ten dollars for each officer, which gave boat owners a direct financial reason to want as many men as possible on their own boat. Testimony given afterward alleged that Captain Mason had an arrangement with Lieutenant Colonel Reuben Hatch, the chief quartermaster at Vicksburg, involving payment for a large share of the men. That is an allegation drawn from testimony, and Hatch was never tried on it, so I report it as alleged.

The officers involved in arranging the transport included Captain Frederic Speed, an adjutant on the department staff, and Captain George Williams, who was handling the paroled men. They gave conflicting testimony afterward on several points. These included who decided to send the remaining men on a single boat, what anyone believed the number to be, and whether the men were counted at the gangplank. Accounts describe a plan to complete the rolls once the men were aboard, and that was never carried out. Accounts also commonly describe at least one other steamboat at Vicksburg in those days, usually named as the Pauline Carroll, leaving with few or none of the prisoners while the Sultana was being filled.

From a communications standpoint, the important feature of the loading is that the critical number existed only as estimates held by a handful of officers on the landing. Those estimates disagreed with one another and were never written into a document that went anywhere. The officers loading the boat were the only people in a position to know the load, and they had a reason to understate it or not to look closely. The officers downstream who might have questioned it received nothing that would have prompted them to. Survivor accounts describe men packed onto every deck, and they describe the upper decks being shored up with extra supports because of the weight. I treat that last detail as reported in accounts rather than established by an engineering record.

I have not found, in the published descriptions of the investigations or in the books I rely on, any message sent from Vicksburg to Memphis or to Cairo reporting how many men were aboard the Sultana. That absence is consistent with everything else in the record, because a number nobody had written down could not be telegraphed to anyone.

The boiler repair at Vicksburg and what the record says about cause

The Sultana arrived at Vicksburg from New Orleans with a leaking boiler. A Vicksburg boilermaker, R. G. Taylor, later testified that he advised cutting out and replacing the damaged section, and that a smaller patch was put on instead so that the boat could leave on schedule. The repair and the loading took place over the same period. That sequence comes from testimony given after the disaster, and it is the most frequently cited fact in discussions of cause. I have not found evidence that the boiler problem was reported to the army officers deciding how many men to put aboard.

The cause of the explosion was examined in 1865 and has been argued over since. Explanations offered by investigators and later writers include the condition of the repaired boiler, the design of the boat’s tubular boilers, low water in the boilers, and the movement of a top-heavy, overloaded hull, which some writers suggest could have shifted water within the boilers. I am not going to rank those explanations or name one as the cause. The record does not settle the question, and cause is outside the scope of a communications piece in any case.

One explanation needs to be labeled at the point where it is mentioned. In 1888 a man named Robert Louden, who had been a Confederate agent, was reported to have claimed that he sabotaged the Sultana with a bomb disguised as coal. That is a later claim, not a finding. Historians I rely on, Salecker among them, do not accept it, and it is not supported by evidence from the wreck or by the 1865 investigations. It circulates widely online, and retellings sometimes present it as a solved mystery. It should be described as an unproven claim made more than twenty years after the event.

The boiler is relevant to the communications angle only because it shows that the information sat in a second silo. The captain, the engineer and the boilermaker knew about the repair, and the army officers held the loading estimates, and neither group’s knowledge reached the other in any documented form. Whatever weight anyone gives the boiler as a cause, no one in a position of authority ever had the boiler condition and the load in front of them at the same time.

Helena and Memphis: two stops where the load was visible

At Helena, Arkansas, on April 26, a photograph was taken of the Sultana with men crowded onto its decks. It is widely reported to be the last image of the boat, and it shows the overloading more clearly than any written description. The photograph matters for the record because it is direct, contemporary evidence of the crowding, independent of testimony that was given later by people with reasons to shade it.

At Memphis that evening the boat unloaded cargo, and some of the soldiers went ashore for a time. Memphis was a major Union military post with a large waterfront, and a boat in that condition would have been plainly visible to anyone on the levee. The published accounts I rely on do not describe any officer at Memphis trying to hold the boat or reduce its load, and I am not going to guess at what anyone there thought or why they did nothing. After midnight the Sultana crossed to the Arkansas side for coal and then turned upriver.

A structural explanation fits the record without requiring speculation about individuals. The transport was moving under orders issued at Vicksburg in a different military command, and nothing in the record suggests that Memphis had received any report that the load was a problem. Under those conditions an officer seeing a crowded troop boat at the end of a war, when crowded troop boats were common, had no document in hand that marked this one as different and no clear instruction about whose problem it was. The boat also held a federal inspection certificate that set its legal capacity, and enforcing that limit at an intermediate landing would have meant someone deciding to countermand another command’s movement on his own judgment.

This is the downstream half of the information failure. Vicksburg held the load and did not send it, and Memphis saw the load and had no report to compare it against. Neither stop had a person whose job it was to ask how many men were aboard and whether that number was legal.

How word traveled: the flooded river, Memphis, and a country reading about Lincoln

After the explosion, the first rescue came by chance and not in response to any alarm. Accounts describe the steamer Bostona, which was coming downriver, arriving at the burning wreck and pulling men from the water. Survivors were also carried downstream on the current, clinging to debris, and some drifted past the Memphis waterfront in the dark, where people heard them or pulled them out. Word reached the military and civil authorities in the city from those survivors and from the rescue boats. Vessels then went out from Memphis, and people living along the flooded banks took men out of trees and off driftwood. I have not found published timings precise enough to say how long it took for the alarm to reach Memphis, so I describe it only as hours, measured in the drift of men on a cold river.

The disaster had no way to raise its own alarm. A boat on the river at night had no means of signaling anyone beyond the reach of its whistle and its fire, and the fire was the only signal anyone ashore was likely to see. Detection depended on survivors reaching people, and in a flood that had spread the river far beyond its banks, many survivors were nowhere near anyone.

From Memphis, the news went out over the telegraph and appeared in newspapers. It arrived in the worst possible news week. President Lincoln had been shot on April 14 and died the next morning, his funeral train was traveling across the North toward Springfield, and John Wilkes Booth had been killed on April 26, the day before the Sultana exploded. Accounts of the disaster, including Salecker’s, describe it as receiving far less attention than its scale warranted, and newspapers of the period gave far more space to the assassination story. Families in Ohio, Indiana, Michigan, Tennessee and Kentucky, the states most of the soldiers came from, learned of their losses unevenly and often late, through survivor lists and regimental word rather than through a single official notification.

The national neglect of the story had a lasting consequence for the record. Fewer contemporary accounts were written, less pressure fell on the investigations, and much of what survives about individual experience comes from the reminiscences Berry collected in 1892, which is one reason the questions of numbers and timing are still open.

The investigations, the one court-martial, and the missing manifest

Several inquiries were held in 1865. Commanders in the region convened boards to examine the loading, including one at Memphis under Major General Cadwallader Washburn and one connected to Major General Napoleon Dana’s command at Vicksburg. Secretary of War Edwin Stanton ordered a further investigation by Brigadier General William Hoffman, the Commissary General of Prisoners. Hoffman’s report placed responsibility on officers at Vicksburg for the overloading. Published summaries differ in detail on how responsibility was divided among Hatch, Speed and Williams, and readers who need the exact language should go to the records themselves, which are held by the National Archives.

Only one officer was tried. Captain Frederic Speed was court-martialed and convicted, and the conviction was later set aside by Judge Advocate General Joseph Holt. Reuben Hatch left the service and was never tried. Accounts commonly report that he did not respond to summonses to testify, and I treat that as reported in the secondary literature. No one was ultimately punished for loading the Sultana.

The investigations ran into the same gap that keeps the death toll a range. Without a manifest, investigators could not establish with certainty how many men had been put aboard. The officers’ testimony about the count conflicted, and the investigators had no document against which to check it. The absence of a count therefore damaged the inquiry as well as the rescue, because it left the central fact of the overloading to be argued from estimates. The written record that a gangplank count would have produced would also have served as the best evidence afterward.

I am not going to say who deserved the blame. The investigations reached their conclusions, the one conviction was overturned, and later writers have argued over the outcome ever since. The communications point holds whichever officer is held most responsible: the boat left with no record of its load, so no one downstream could act on that load and no one afterward could prove it.

The record applied to an evacuation, a dispatch floor, and a casualty count

The closest modern equivalent to the Sultana is a mass movement of people who are not able to look after themselves. Examples include a nursing home evacuated by bus, a shelter population moved to another county, or a group of patients transferred during a hospital evacuation. In each case the number and identity of the people moving exist at the sending point. The value of that information depends entirely on whether it is written down and sent ahead before the vehicles leave. A roster that is going to be completed on the bus has the same flaw as a roll that was going to be completed on the Sultana, because the moment of loading is when everyone is busiest and least able to stop and count.

For the communications center, the lesson concerns who receives the expected count and what they do when the number is wrong. A receiving facility or a neighboring dispatch center that knows to expect forty people on two buses at a given hour can raise an alarm if one bus is late or the count does not match. A receiving point with no expected count can only notice that something seems off, which is what the Memphis levee could do and did not act on. The expected count should travel by a channel that leaves a record, such as a CAD note, a resource order or a message to the receiving facility, and not only by radio voice traffic that nobody logs.

For mass casualty operations, the death toll dispute is a warning about patient tracking. Triage tags, transport logs and hospital receiving lists exist so that a later count can be reconciled. Counting rules matter as well: whether to count people who die after reaching the hospital, and how to count people who left the scene on their own. If those rules are not settled before the incident, two honest officials can give two different numbers for the same event, as historians do for the Sultana.

The news angle applies to public information. A large incident that coincides with a bigger story will get less public attention, and families may learn less and later than they should. That makes the agency’s own direct notification to families more important, because the press cannot be counted on to carry it.

What to do at your agency

• The emergency management coordinator pulls the evacuation and transportation annex of the existing emergency operations plan this month and confirms it requires a written roster for every vehicle before departure, sent to the receiving site, and adds that requirement in one paragraph if it is missing.
• The communications center manager writes into the existing mass evacuation or patient movement procedure the CAD field or message where an expected count and expected arrival time are recorded, and names the dispatcher position responsible for confirming arrival against that count.
• The EMS coordinator checks the mass casualty plan for a written rule on how deaths after hospital arrival and self-evacuated patients are counted in the official casualty figure, and adds the rule to the plan before the next regional MCI meeting.
• The public information officer revises the prepared casualty statement template so that every figure released carries the source and the time it was confirmed, and so that it states plainly when the count is incomplete.
• The training officer adds a mismatch to the next scheduled evacuation tabletop, in which one bus arrives with fewer people than its roster lists, and records in the exercise notes who noticed the difference and how long it took to reach the incident commander.
• The fire chief or emergency manager confirms in writing, in the existing family assistance section of the plan, which named person notifies families directly during a mass casualty incident, so that families do not depend on news coverage to learn what happened.

Takeaways

• The Sultana exploded on the Mississippi a few miles above Memphis at about 2 a.m. on April 27, 1865, carrying paroled Union prisoners far beyond its legal capacity, usually given as 376.
• Published estimates of the dead run from around 1,200 or slightly below to more than 1,800, with Salecker’s 1996 reconstruction toward the lower end and older and popular accounts at the higher end.
• The death toll is disputed mainly because nobody made a complete roll or manifest of the men as they boarded at Vicksburg.
• Knowledge of the load and knowledge of the boiler repair sat with different people at Vicksburg, and I have found no record of either being reported to anyone downstream.
• The overload was visible at Helena and Memphis, where no one with a report of the load was in a position to compare it with the boat’s legal limit or to stop the movement.
• The cause of the explosion is not settled by the record, and the 1888 sabotage claim is an unproven later claim rather than a finding.
• News of the disaster spread during the week of Lincoln’s funeral and Booth’s death, and it received far less attention than its scale warranted.
• Captain Frederic Speed was the only officer court-martialed, his conviction was set aside, and no one was ultimately punished for loading the boat.