Confederate Secret Service Attack on St Louis Levee

The Confederate Secret Service Attack on the St. Louis Levee, September, 1864

By John B. Castleman

Excerpted and introduced by G. E. Rule

from “Active Service”, John B. Castleman, Major CSA, General USA, 1917

John B CastlemanMajor John B. Castleman had an active career. First as captain of one of John H. Morgan’s cavalry companies, and later as Thomas H. Hines’ junior partner in the Northwest Conspiracy. In late September of 1864, Castleman was captured in Indiana and spent the rest of the war in a Union prison wondering if he was to be executed. He was finally released on the condition that he leave the country and never return.

A few years later his exile was rescinded and he returned to his native Kentucky where he had a very successful career. Castleman was a U.S. General in the Spanish-American War, and many observers credit his cool actions for preventing another Civil War in Kentucky in 1900 when a disputed election for governor, and assassination of one of the candidates, threatened anarchy there.

Did The Confederate Secret Service Attack on the St. Louis Levee, September, 1864 actually occur? The documentation is sketchy. While Castleman’s account does not claim a specific number of boats burned, it certainly leaves the impression that at least some were destroyed. That he was in Missouri about the time in question seems proven by the Charges and Specifications brought against him by the Federals and by his own account. James D. Horan, working from Hines’ and Castleman’s papers in his account of Thomas H. Hines career (“Confederate Agent”, 1954), claims 5-10 boats destroyed at St. Louis by Castleman and his comrades in September of 1864.

The levee at St. Louis was a favorite target of the Confederate secret service and their copperhead allies, OAK (Order of American Knights). Not counting Castleman’s action, there are at least four documented instances of multiple-boat burnings at St. Louis during the course of the war. These accounted for the destruction of 18 boats. While one of them, in 1862, may (or may not) have been accidental rather than intentional, the other three were almost certainly the work of the Confederate secret service and/or OAK.

St. Louis was the top of the supply chain in the theatre. Men and supplies were funneled from the northern states down the river network to St. Louis and then distributed from there. In 1862-1864, Grant’s and then Sherman’s armies were both principally supplied from St. Louis. Each so-called “civilian” boat was a military asset to the Union, and recognized by the authorities as such. One government witness at Robert Louden’s trial testified “I consider every boat on the river to be in the government service, directly or indirectly.” The Confederates were quick to take notice of the state of affairs and act accordingly.

But did Castleman’s attack result in the destruction of any boats? Did it really take place in September of 1864 at all? We have not been able to find any documentary evidence outside of Castleman’s and Hines’ accounts that any attack took place in the timeframe claimed. Even the Charges and Specifications against Castleman only claim that his mission was to destroy public property, including in Missouri, not that he and his comrades had actually done so. Neither Castleman nor the charges name a specific boat either. Other sources that make no mention of this attack are the newspapers of the time, other memoirs, various official records, and “Way’s Packet Directory 1848-1994”. This last source is considered “the bible” for steamboats on the western waters (someday we’d like to share our research with the current editors to make it even more complete), and is a truly impressive achievement of 6,000 listed boats and their fates. Mr. Frederick Way, Jr. (now deceased) spent roughly 80 years of his life collecting this information. While we have identified approximately 100 “suspicious” steamboat burnings on the western waters during the war from Way’s, there are none at the levee at St. Louis later than July of 1864.

It is highly unlikely that if 5-10 boats had been destroyed at once at St. Louis it could have gone without notice from all these separate sources. Unless further evidence appears, it must be considered either Castleman had his dates wrong or his “Greek Fire” was even more ineffective than he complained –so much so the St. Louis papers disdained to even notice the attack. If several boats really were destroyed, then the other candidate dates for Castleman’s attack would be September of 1863 and July of 1864. The traditional timeline would have Castleman in Canada on the latter date, but September of 1863 is a mystery as to Castleman’s whereabouts. He was with Morgan on his Ohio raid that summer, but was not amongst those captured after the Battle of Buffington’s Island. Castleman’s memoirs leave this period entirely to Hines account of events –Castleman himself does not appear in the narrative from approximately April of 1863 until December of 1863. Another hint that September of 1863 might be the true date is Castleman’s description of “embarrassing the United States Army at Vicksburg”. While this would certainly be true in September of 1863, one year later that army was at Atlanta.

Our interpretation of “Military Direction” led to the contemplation of embarrassing the United States Army at Vicksburg by partially destroying its means of supply. It was known that the army at Vicksburg was with commissary stores, quartermaster stores, ordnance, forage, supplied chiefly from St. Louis by steamboats.

From Marshall [Illinois], George B. Eastin was sent to inspect and report in particular detail. Where such inspections were made, immediate report was required. Within twenty-four hours verbal information was brought of the approximate number of boats lying at the St. Louis wharf, between what streets, the names and approximate size of the steamboats, character of cargoes and probably sailings.

The same date ten of us went back to St. Louis to attempt partial destruction of this government service and embarrass the supply to the Vicksburg army.

We stopped separately at the Olive Street Hotel, where we arrived early in the morning. Directly after breakfast, without seeming concert of action, each one went aboard of the boats, previously assigned, lying between the foot of the streets allotted and each quickly knew his boats by name and location, and where was to be found the most combustible or most easily ignited feature. Citizens were not denied permission to go at will about the boats.

By eleven-thirty this was done. We took our luncheons separately and proceeded to make the best use of the information obtained by the personal inspection, advices of which had been considered by our little “conference of war”, held in my room.

We had had the misfortune to have had made a quantity of small bottles of liquid designated “greek fire” [typically made of turpentine and phosphorous]. “Greek fire” was a combination of chemicals which, when exposed to the air, ignited and had, or was designed to have, the advantage of ignition after a minute had elapsed in which time the user of the liquid could move from the scene.

It is probable that had the little band of fearless Confederate soldiers used a few boxes of matches, there would have been none of the seventy-three steamboats left on that day or landed at the St. Louis wharf.

But “greek fire” was not reliable and in most instances the self-ignition did not occur. We dared not go back to complete the work and, as previously arranged, we quietly left–taking passage separately—on the train that afternoon.

One cannot, in the fifty years that have passed, forget the deliberate courage of that little body of men. It is a picture still vivid in memory that made lasting the quiet demeanor of those boys, each taking life in hand, and going with nonchalance in performance of service.

Those boys are now all dead, but one –God bless those fearless boys.

On the first of October I had an engagement at Sullivan, Indiana, to meet some men who were trusted by Hines and me. The chief of those men was Mr. Humphrey.

I drove across from Marshall, Illinois, and instructed my comrades to come via railroad to Sullivan. To do this they went to Terre Haute, and thence over the Evansville Road. I was arrested at Sullivan, and afterwards, after detention, was taken to Terre Haute on a train that arrived there in the early morning.


Charges and Specifications Preferred Against John Breckenridge Castleman,


Major of the Second Regiment, Kentucky Cavalry of the Rebel Army.

Charge 1st: Lurking and Acting as a Spy.

Specification 1st: In this, that the said John Breckinridge Castleman, Major of the Second Regiment, Kentucky Rebel Cavalry, did, on or about the 26th day of August, 1864, secretly, in disguise and under false pretenses, enter and come with the lines of the regularly authorized and organized military forces of the United States, and within the states of Michigan, Illinois, Missouri and Indiana, and did secretly and covertly lurk and travel about as a spy in the dress of a citizen, and under and assumed name, and did seek information with the intention of communicating it to the enemy, and remained within said military lines until arrested as a spy at Sullivan, Indiana, on or about the 30th day of September, 1864. All this within the states of Michigan, Illinois, Missouri and Indiana, during the months of August and September, 1864, and within the military lines and the theater of military operations of the Army of the United States, at a period of war and armed rebellion against the authority of the United States.

Specification 5th: In this, that the said John Breckinridge Castleman, Major of the Second Regiment, Kentucky Rebel Cavalry, was found lurking and acting as a spy in the state of Missouri, at or near the city of St. Louis on or about the 14th day of September, 1864, within the military lines and the theater of military operations of the Army of the United States, at a period of war and armed rebellion against the authority of the United States.

Charge 6th. Conspiring to Destroy Government Property in Violation of the Laws of War.

Specification: In this, that the said John Breckinridge Castleman, Major of the Second Regiment, Kentucky Rebel Cavalry, did, on or about the 26th of August, 1864, enter the states of Illinois, Missouri, and Indiana, in pursuance of an agreement with Jacob Thompson, Clement C. Clay, Jr. [Confederate commissioners in Canada], James A. Barrett [Missouri OAK leader], Captain Hines, and others unknown, to burn and destroy government arsenals, depots, and storehouses, and steamboats in government employ, and incite others thereto, with the purpose and intent of hindering and impeding the efforts of the lawfully constituted authorities of the United States in suppressing an armed rebellion against its authority.

[All emphasis added by editor]