Bonkers
“Bonkers” message. EC 19. Ciphers Sent. 1864, Jan. 27 — 1864, Dec. 2. P. 293.
I’ve often felt that that the vocabulary of archival description could benefit from some content designators. When it comes to ciphered telegraphic messages, the term “bonkers,” as applied to the telegram above, seems very apt.
The USMT cipher men indeed appear to have been on a mission not dissimilar to the staff of the “Confuse-a-Cat” outfit from the Monty Python sketch.
The ciphers were constantly changed, modified, updated, and otherwise fiddled with. The first version of the cipher, developed by Anson Stager for Governor of Ohio in April 1861, was so short that it fit on the back of a business card. The last iteration to be used during the war, cipher No. 4, contained some 1500 arbitraries. The ciphers were numbered out of conventional sequence: the ciphers developed early in the war were numbered 6 and 7, followed by the series of 12, 10, and 9, then 1 and 2, and, towards the end of the war, 3, 4, and 5. These series did not replace each other but rather were often used simultaneously. Occasionally they had to discarded. No. 6 and 7 were discontinued in August 1862, after Nathan Brooks was captured by John H. Morgan’s men, and No. 12 followed suit after Stephen L. Robinson fell to guerrillas in July 1864. The cipher No. 1 which, according to William R. Plum was used to send “more important telegrams,” had to go after James E. Pettit and John F. Ludwig were captured by Forrest’s men in September 1864.
The arbitraries were confusing enough. But there were also the blind or check words, used solely to confuse the interceptor, that could be thrown in either at the end of each column or after every sixth word; commencement words that could indicate the number of either words or lines, and key words concealing different routing techniques for the entire telegram or its parts, coded by columns, numbers assigned to each square made by the column lines, or both.
All these were in constant flux. After he met a “reformed gambler” on a train, Stager updated the routing instructions in cipher No. 12, to include a mnemonic devise revealed by that venerable gentleman. All you had to do was to memorize the formula “K 842 W 795 M 361 B,” or “The King had 842 women, 795 men and 361 boys” and use it to shuffle the deck to know the exact location of every card. The numbers stood for the spotted cards, with 1 designating the ace. “Boys” indicated a ten spot; “Women,” the Queen, and “Men,” for Jack. Stager appropriated the formula, replacing the king, queen, jack, and ten-spot with numbers 13,12,11, and 10, to scramble a message following the sequence 13- 8-4-2-12-7-9-5-11-3-6-1-10. Ciphers No. 9 and 10, adopted for the use of the Western Departments, relied on a further elaborated version of this technique, with each word numbered and counted from the top or of the bottom of a column, with X standing for a check word.
What we see in the “bonkers” message is the process of finalizing a new cipher. The message itself was of no particular importance. As seen from this publication in OR, it merely informed Assistant Quartermaster Captain H.C. Hodges to take charge of Sherman’s supplies.
Official Records of the War of the Rebellion. Series 1, vol. 47 (Part 22) P. 475. The “bonkers” message appears at the top of the page.
Stager, Stafford G. Lynch, and other cipher men used this message as a guinea pig for testing various routing instructions being developed for a new cipher. What they were working on was Cipher No. 4, the last cipher to be used during the war. Its predecessor, known as No. 3, was devised by Samuel H. Beckwith in winter of 1864 and included a new set of arbitraries which, among other things, addressed the problem of a “stuck” sounder by avoiding certain letters. (A stuck sounder often resulted in making dashes instead of dots and which, on one memorable occasion turned “pacific” into “fairfye.”) Beckwith’s cipher was unveiled on December 25, 1864 but, for some reason, saw little use.
The top message, first addressed to Stafford G. Lynch, was instead sent to Stager, with the instructions to return it “by train of or in Some confidential manner.” Two different ciphers of the same message were also sent to John Horner in New York.
EC 19. Ciphers Sent. 1864, Jan. 27 — 1864, Dec. 2. P. 294. Note a different version of the “bonkers” message, the second on the page.
EC 19. Ciphers Sent. 1864, Jan. 27 — 1864, Dec. 2. P. 295. Another version of the “bonkers” message, on the top of this page.
The result of this work was sent to the operators at Grant’s, Sherman’s, and Thomas’s commands on March 23, 1865. The final product, known as Cipher No. 4, numbered some 1608 arbitraries; the key proper comprised 12 pages, each differing in the words used and the route employed. Significantly, the book contained no directions for the use of the cipher; if captured, it was of little use to the anyone who was not a USMT cipher man, including, I am afraid, archivists.
There is one additional takeaway from the story of the “bonkers” message: the fact a message appears in OR certainly does not diminish its historical value.
Olga Tsapina, Norris Foundation Curator of American History. The Huntington Library.
Laughing Matter
To McCaine
4:30 pm Washn. Aug. 26th 1864
Katy Nabob The provisional battalion of Pacific belonging
to Greggs quitman which is tower upper
Attica saddling the Windsor while the Platina Andover
Panama is absent pekin has been ordered to
Blubber stop This will leave me without means
of saddling the windpipe while the paddle Andover
is absent stop Taunton Waite whites wedge that
he cannot get his whip ready to Talbot
before Monday zebra The forges coal &c had
toby sent from here I think he will
get ready Persia possible unity If you think
it advisable I will send out the Gas
Amos sligo direction of Aldie They cannot raise
more perfume prolong forth field & they cannot
go to Laughter Sheffield zodiac they may wolf
about Aldie & pick up rumors Cork = screw
Draw off your water out of town
Decoded, reads:
To McCaine
4:30 pm Washn. Aug. 26th 1864
4:30 P.M. P. H. Sheridan The provisional battalion of Cavalry belonging
to Greggs Division which is over the upper
Potomac guarding the River while the 8 Illinois
Calvary is absent comma has been ordered to
City Point stop This will leave me without means
of guarding the River while the 8 Illinois
is absent stop Major Waite reports today that
he cannot get his Regiment ready to movement
before Monday period The forges coal &c had
to be sent from here I think he will
get ready as soon as possible period If you think
it advisable I will send out the 16
New York in the direction of Aldie They cannot raise
more 3 hundred for the field & they cannot
go to Snicker Gap period they may scout
about Aldie & pick up rumors Augur
Draw off your water out of town
On August 26, General Christopher Columbus Augur, the commander of the the defenses of Washington, D.C., sent a ciphered telegram to Phillip Sheridan who was in the process of planning an expedition against John S. Mosby’s men. Major John M. Waite of the 8th Illinois Infantry was charged with leading the force. He, however, needed some time to assemble the men: the regiment, formally part of Augur’s XXII Corps, had been scattered on all over Virginia and Maryland: six companies were assigned to Lew Wallace at Baltimore, four were guarding the Potomac between Great Falls and the Monocacy, another was at Port Tobacco, and one was with the army of the Potomac. Augur, still reeling from Jubal Early’s raid on Washington on July 11, was not overly enthusiastic about the arrangement. The best he could to reinforce the expedition was the 16th New York which had already been cut up by Mosby’s men.
As with many telegraphic communications between generals, this telegram was published in the Official Records of the War of the Rebellion, after the deciphered recipient’s copy. 
Our telegram, which appears in the USMT ledger of sent messages (EC 18), is not just the ciphered version of the OR publication. For one thing, it was sent from the USMT office at the War Department an hour and a half after the original message went out from the headquarters of the XXII Corps, located on the corner of Fifteenth and a Half St. and Pennsylvania Avenue. The message is addressed to R.R.R. McCaine, Sheridan’s operator, rather than Little Phil himself.
Robert Rezin Redman McCaine, 1833-1873
And then there is the strange appendix involving a cork screw and instructions to “draw your water out of town.” In fact, the message offers a glimpse into how codes were modified and adjusted, which often happened on the fly.
The message had been ciphered a variation of Cipher No 1. This version, found in EC42, EC44, and EC46, features significant and apparently recent changes. The term “commencement word” was replaced with “blind word” which indicated the number of columns rather than lines. 
As seen from the handwritten corrections made in EC 46, “the sum of the numbers set the opposite the next two words” indicated the number of lines. “Town” equaled to 6 lines, and “water,” to 10, which translates into sixteen lines of the eight-column message.
Instead of simply instructing McCaine to add the values, however, his Washington counterpart opted by a coded message involving the “cork-screw” and drawing off, i.e. “decanting,” his “water out of town,” and added a few words at the top and bottom of the message. When the operator followed the respective routing instructions, he saw a tip to “use tower for on the Not Received letter.” (The value of the arbitrary “tower” as listed in the code book is “over the.”) Most likely, this message was not even intended for Sheridan’s eyes, but rather served as a test case for changes and modifications in ciphers.
It should be noted that the term “Cork=screw” is a triple play on words. The message was sent by Augur, a homophone of which is “auger”. An auger is a helical screw, often used for boring holes in things. A short step sideways and you have “corkscrew,” split in two to fit the columns. (Thank you MEinaudi for this find. I must confess, my knowledge of drilling equipment and terminology a bit limited; I was thinking more along the lines of an augur as in the ancient Roman priest. This, of course, would be very hard to code).
And another thing. Because there was no arbitrary for Snickers Gap, Va., the operator had to improvise: Snickers was replaced, of course, with “Laughter.” I would probably go with “Chortle” or “Giggles.”
Now, Jesse …
No 47 P.
J. H. Bunnell (1) Washn. Mar 31st 1864 5 P.M
Grapes, Laugh, Plug, March, Lucy, For, Knapsack Embrace
or Bridle Unity I would Quince the Welch
as far to the Ripley of Chart as
possible Zebra It is a good plan to
Pine all the Saints you can when Rapturing
is expected and make all other Territory Tartar
to hold defensible for the smallest possible number
of Whiskey yoke John Lieut Shelter Now Jesse
As with many ciphered telegrams, this message, found in ledger EC 19, looks mostly like gibberish. This was the end product of a complex and sophisticated system of word substitution encoding originally developed by Anson Stager. Words that described sensitive data — names, time indicators, numbers, military terms, places, etc. were replaced by replacement words or arbitraries. The text was broken into squares formed by columns and lines and then scrambled during the transmission. The keys listed not only the arbitraries but also the commencing words and line indicators specifying the number of columns and lines and routing instructions listing the order of the transmission that scrambled the sentences. Thus encoded, the text assumes the appearance of an assemblage of random words, impossible to make sense of, let alone decipher.
This telegram, luckily, can be relatively easily deciphered, as we happened to have a copy of the key to this particular code. This would be the ledger EC 44, titled Cipher for Telegraphic Correspondence; arranged expressly for Military Operations, and for important Government despatches and known as Cipher No. 1. It contains codes for commencing words, line indicators, routing instructions, and arbitraries as developed in 1862 but implemented in February 1864.
For example, as seen on this page, “John” was one of the codes used for Ulysses S. Grant, and “Knapsack” was reserved for William T. Sherman. The entire telegram reads:
J. H. Bunnell (1) Washn. Mar 31st 1864 5 P.M
Washington, 30, 1, March, 5 p.m, For, W. T. Sherman Nashville
or Chattanooga Period I would destroy the railroads
as far to the east of Knoxville as
possible Period It is a good plan to
concentrate all the forces you can when fighting
is expected and make all other Territory necessary
to hold defensible for the smallest possible number
of troops signed U.S. Grant Lieut General Now Jesse
The message is part of a rather anxious back-and-forth between Grant, just two weeks into his tenure as the commander of the Armies of the United States, and Sherman, just appointed the commander of the Division of the Mississippi. The exchange was prompted by Forrest’s raids in the Union occupied Western Tennessee. Grant, who was passing through Washington on his way to Fortress Monroe scribbled this telegram and handed it to a USMT operator. Grant’s original note, now held by the United States Military Academy and published in the Papers of Ulysses S. Grant, reads:
“I would destroy the rail road as far to the East of Knoxville as possible. It is a good plan to make the country to be held to concentrate all the forces you can wher fighting is expected and make all other territory necessary to hold defensible for the smallest possible number of troops. ”
The telegram appears in OR (Ser. 1, vol. 32, part 3, p. 213), but only as part of another telegram from Sherman’s aide-de-camp Lewis Mulford Dayton sent to John Schofield; in this case Grant’s “territory” has been replaced with “preparations.”
But what are the words and numbers that swarm around the message? The telegram was directed to Jesse H. Bunnell (1843-1899), the operator attached to the George H. Thomas of the Army of the Cumberland. The note “Sent from Book 5:20 P.M. Tinker” tells us that the telegram was re-transmitted twenty minutes later by Charles Almerin Tinker (1838-1917). The scribble on the bottom of the telegram appears to indicate that 75 words were charged in total. The main body of the telegram contains 63 words, including “Now Jesse” that the operator managed to squeeze in at the end of the line. Apparently he had more to say. Using the entry in the ledger, Mr. Tinker sent the same message with some additional words and an extra layer of coding intended for Bunnell’s eyes only.
The marks on the lower left side contain the commencing word (Mobile) followed by the line indicator (Horse – Deal). In the cipher key, “Mobile” indicated a nine-column transmission. However, the line indicators called for an eight-column set up, and the telegram is indeed broken into eight rather than nine columns. According to the key, the telegram was to be routed in the following order: up the 8th column, down the 5th, up the 7th, down the 1st, up the 6th, down the 3rd, up the 2nd and down the 4th.
The little numbers indicated the place of the word in the column, with additional words supplied on top and bottom of the columns. For example, the number 4 on top of the 1st column pointed towards the 4th word from the top, i.e. “possible.” The last two words which began the private message (“Now Jesse”) were not included in the count, so the 1st from the bottom in the column 8 is “number” rather than “Jesse.”
If we follow the coded instructions, the embedded message would read: “Now Jesse your number you would full when possible bad good draw and you about Bridle (Chattanooga) Grant here.” This sort of makes sense to us. It certainly made sense for Jesse.
At this point we can only speculate as to the nature of this exercise. Most likely it offered the operators, who, after all, were not supposed to use government communications for their own needs, a way to bypass the rules.
This little puzzle, (which took yours truly some two hours and elicited some highly descriptive epithets), must have been a piece cake for Bunnell. By the ripe age of nineteen, he boasted a six year career with the telegraph, (yes, he became a full-fledged operator at 13), and a speed record which he set in 1860, transmitting President Buchanan’s last message to Congress (14,040 words) in two hours.
Studies of the USMT personnel is one of the new and exciting directions in the Civil War studies offered by the Decoding the Civil War project. I’m certain that historians will find out what Mr. Bunnell and Mr. Tinker were up to in March 1864. Until then, we can simply stand in awe of the ingenuity and the skill involved in this multiple-layered coding and decoding.
Olga Tsapina, Norris Foundation Curator of American History. The Huntington Library.
When in Doubt, Eat the Cipher!
By Daniel Stowell, Decoding the Civil War Independent Researcher.
Courtesy of Seth Kaller, Inc.
When working with numerous collections, sometimes one stumbles across a little nugget that just calls out to be told. Working with the Eckert telegrams has made me familiar with the codes and ciphers used by the United States Army and Government. So, imagine how my curiosity was piqued when I saw the following story from August 1862:
“During all the spring months I alone in all the Army Corps was entrusted with the Government Cyphers. During General Pope’s retreat, I was one day sent for by Generals Pope & Banks, to put into cypher a very important dispatch to General McDowell, with whom direct communication had been cut off by the enemy.
I was obliged to reply that during the severest part of the Battle of Cedar Mountain when I was in the greatest danger of being killed or captured at any moment, I had felt it my duty to destroy the cypher which I tore up into a hundred or more very small pieces & swallowed some of them. My action was approved. I then offered to carry the orders, unwritten, myself to General McDowell, if I could find him, and take my chances.
My offer was accepted, but while the instructions were being prepared, the advance of General McDowells Corps came in sight, & I was relieved from a duty which would have put me in the greatest danger of capture or otherwise.”
This is excerpted from a twenty-five-page handwritten memoir written by Frederick d’Hauteville on his service in the American Civil War. d’Hauteville was writing about the the Battle of Cedar Mountain, fought on August 9, 1862. Brigadier General Crawford in his report on the battle specifically commended Captain d’Hauteville, “who from the first rendered me especial and important service, attended with great personal exposure.” Crawford probably wasn’t thinking of d’Hauteville’s willingness to eat the cipher when commending him, and it is impossible to tell whether the cipher which he ate was one of Stager’s telegraph ciphers, but d’Hauteville’s memories do illustrate the extent to which those entrusted with the ciphers went to protect them.
After the battle d’Hauteville discovered that a ball had pierced his blankets, strapped behind his saddle, with more than a dozen holes. Among the regiments in d’Hauteville’s division at the battle was the 2nd Massachusetts Volunteer Infantry, with officers from “the leading families of the City of Boston,” many of them, like d’Hauteville, graduates of Harvard University. The regiment with fewer than 500 men suffered 173 casualties, and 16 of its 23 officers were killed, wounded, or taken prisoner. “All of them were my friends, & very dear friends,” d’Hauteville recalled, “Their loss was enormous, but they went to their deaths with sublime courage.”
Frederick Sears Grand d’Hauteville (1838-1918) was born in Boston, Massachusetts, to a Swiss nobleman and American mother, who returned to the United States while pregnant. Their marital conflict led to a contentious custody battle over Frederick in 1840 in a Philadelphia court, which his mother eventually won. D’Hauteville graduated from Harvard University in 1859. He was appointed volunteer aide-de-camp to General Nathaniel Banks in December 1861, and served at the Battle of Winchester in March 1862. Commissioned captain on June 30, 1862, he served on Brigadier General Samuel W. Crawford’s staff, including action at the Battle of Cedar Mountain in August and at Antietam in September 1862.
By December, 1862, d’Hauteville had returned to General Banks’ staff and traveled to New Orleans, where Banks commanded the Department of the Gulf. D’Hauteville resigned his commission on March 1, 1863. Later in 1863, he married the daughter of former New York Governor Hamilton Fish, but she died the following year. In 1872, he married Susan Watts Macomb, whose grand-father Major General Alexander Macomb was a hero of the War of 1812 and general-in-chief of the United States army from 1828 to 1841. The d’Hautevilles kept a home in Newport, Rhode Island, but they spent much of their time in his paternal family’s thirty-room chateau overlooking Lake Geneva in Switzerland.
Add Hooker
Wash’n Jan’y 27th /63
Sheldon
Enter the name of Joseph
Hooker cipher the bottom word
on page ten of no
nine to Borgia and berry
are the arbitrary words given
A Stager
Sometimes changes to codebooks required entirely new printings, but occasionally the cipher clerks would only need to add a single entry. In the case of General Joseph Hooker, this late addition may have occurred because he was not commissioned immediately at the start of the war, and only rose to prominence during the Peninsula Campaign of 1862.
Coded Nonsense
The Bethune desires that you
will prepare to tablet Whack Tappan and
at all hazards make a Pigeon raid
to break the Saginaw at or near
Kiss and at such other places as
may be practicable period
Project member Daniel Stowell recently pointed out to me the amusing and nonsensical statements that often result from text being encoded. It makes sense, of course, since a word substitution code is supposed to confound readers who don’t have the codebook, but sometimes the messages end up sounding like 19th century Mad Libs. We’ll be posting some of these over on Twitter, so be sure to follow along. And if you come across any #19thcentmadlibs while you’re transcribing telegrams, please share them!
Daniel Stowell is Director and Editor of the Papers of Abraham Lincoln at the Abraham Lincoln Presidential Library and Museum.
Telegram Tails
As Daniel Stowell discussed in his post on routing, it was sometimes necessary to add null words to the end of messages to ensure that every line was the same length. The person who composed the message didn’t care how long the telegram was, so it was up to the telegraph operators to fill in the blank space. When confronted with a message like this:
| for | Adam | the | ninth | Walpole |
| corps | are | under | orders | to |
| join | you | the | first | Raleigh |
| are | embarked | ready | to | start |
| signed | Borgia | Lucy |
the operator had to add two words to round out the line. In this situation the operator chose the mundane option:
| for | Adam | the | ninth | Walpole |
| corps | are | under | orders | to |
| join | you | the | first | Raleigh |
| are | embarked | ready | to | start |
| signed | Borgia | Lucy | good | evening |
There are plenty of examples of niceties being passed over the wires, but there are also some delightful insights into the personalities of the operators. I call these little discoveries Telegram Tails, and will be sharing them regularly on Instagram and Twitter using the hashtag #telegramtails. If you find any good examples, be sure to post them on a talk board or put them on social media!
Ch-Ch-Ch-Changes!
mssEC_24_076
Chicago 21
Sheldon
I thank you for
the suggestion the printer
is now at work
printing arbitraries for months and
hours and which will be
pasted inside cover of book
I designate arbitraries as follows
hamlet for James Hebrew for Pamunky
valley for Brigadier vernon for General
venus Colonel Vienna Major virtue Captain
vulture lieutenant Vincent quartermaster violet paymaster
vista surgeon how you like it
The codebooks used in the Civil War evolved as time went on, as is evident in this telegraph. The updates here correspond to copies of Cipher, or Code, Book 9, which has gaps in the arbitraries even after these additions. Half of the “P” section, most of “T”, and all of “U” are still unassigned, so there was still room for expansion of the code if needed. We will be recording these unassigned arbitraries as we will the assigned ones (e.g. Hamlet = James, Vernon=general, Pacific=[none]) so that we will have a complete data set when it comes time to decode the messages.
Keeping Secrets, Part 2: Route Ciphers
By Daniel Stowell, Director and Editor of the Papers of Abraham Lincoln at the Abraham Lincoln Presidential Library and Museum.
Maintain
ing the secrecy of high-level political and military communications during the American Civil War was a necessity for both the Union and the Confederacy. In the previous post, we learned that the Union had the advantage of a simple but ingenious system developed by Anson Stager, the civilian superintendent of the Western Union Company in Ohio. Early in the war, Stager developed a cipher system for use by the governor of Ohio and Union generals in the area. When he became the head of the United States Military Telegraph in October 1861, Stager brought his cipher system with him and used it throughout the Union high command.
Stager’s system depended on absolute secrecy, and not even top civilian or military officials were allowed to see the cipher codes. The cipher Stager developed relied on two key principles. First, a cipher telegrapher replaced key nouns and verbs for persons, places, and military terms with arbitraries, or code words, known only to his counterpart at the other end of the telegraph line. Although this technique was effective, so long as the cipher remained absolutely secret, over time, dedicated Confederate agents who intercepted such messages might be able to guess at the true meaning of many or most arbitraries.
However, Stager’s system had another key component. It was also a sophisticated “route cipher.” At its simplest, a route cipher simply reorders the words in a message from their ordinary reading order. For example, a backwards route would simply reverse the order of all words in the message, and the recipient could read the message from the end to the beginning to understand it. The system employed by the Union during the American Civil War was substantially more complicated.
The key to decoding a route cipher is to place the words in the proper number of rows and columns, similar to a modern-day spreadsheet. The key unit in Stager’s system was the word, not the letter. So, a cipher telegrapher would take a message and write it out in a series of rows and columns. If the message did not end on the last column of the last row, he would add additional “null words” to fill out the grid. These null words could be nonsense or they could even send a short message to his counterpart on the other end of the line, such as “Nab those Rebs” and “Rained nicely tonight.”
After the message was arranged in a grid, the cipher telegrapher chose a route through the grid to reorder the words. For example, if the message was 10 columns wide by 11 rows long, then the cipher telegrapher could use the commencement word “Morton” (among others) to indicate that the message had 10 columns and the commencement words “Next” or “News” to indicate that it had 11 rows (2 rows + 9 rows). The receiving cipher telegrapher consulted his cipher 1 code book and knew that the incoming message of 110 words should be arranged in 10 columns and 11 rows. Then, he placed the words in their proper order by reading the route indicated in cipher 1: down the sixth column, down the tenth column, up the first column, down the eighth column, up the second column, down the fourth column, up the seventh column, down the third column, up the fifth column, and down the ninth column.
Reading the words in this manner placed them in their proper order, but the receiving cipher telegrapher still needed to replace the arbitraries in the message with their clear counterparts, either words or punctuation. The telegrapher also stripped out the commencement words and the null words to leave only the words of the message.
The scrambling of words through the route cipher method made Stager’s system far more difficult to decipher than using arbitraries alone. In fact, the Confederates never broke any of the Union ciphers, and the ability for political and military leaders to communicate important information securely was important to the overall success of the Union war effort.
Fortunately for researchers, none of the telegrams in the Thomas T. Eckert Papers are scrambled in the route in which they were originally sent. Even those with arbitraries for words and punctuation are in their proper order, making decoding them a much simpler task than Confederate agents faced in the 1860s.
Keeping Secrets, Part 1: Arbitraries
By Daniel Stowell, Director and Editor of the Papers of Abraham Lincoln at the Abraham Lincoln Presidential Library and Museum.
Keeping secrets from the enemy is vital to combatants in any war, and the American Civil War is no exception. Union and Confederate leaders sent directions to and received reports from their subordinates over wide areas. The telegraph made “lightning” communication between distant points possible. Leaders who could use it effectively had a decided advantage, both on the battlefield and in mobilizing their society’s resources for war.
The simple substitution of words or letters for other words or letters is an ancient technique for keeping messages private between the sender and the recipient. Lovers sent messages to each other in their own private codes, and diplomatic and military leaders also sent messages that messengers or interceptors could not understand. Critical to transmitting such messages successfully was that both parties knew and understood the code and that the encoded message made it safely between them, usually by a personal courier or perhaps even by mail. The development of the telegraph complicated the private transmission of messages, because anyone along the line of communication could tap the line and intercept the series of dots and dashes that were used to transmit messages over the wires. In wartime, messages traveling hundreds of miles were often intercepted by enemy agents.
Anson Stager (photo from Library of Congress)
The fact that Confederates could often intercept Union messages sent by telegraph made encoding important messages critical to the success of the Union war effort. At the outbreak of the American Civil War, the governor of Ohio asked Anson Stager, the civilian superintendent of the Western Union Company, to manage telegraph operations in southern Ohio and along its border with (West) Virginia. Stager developed a cipher, or code, with which the governor of Ohio could communicate secretly with the governors of Indiana and Illinois. General George B. McClellan also used the cipher when he was leading a Union army in western Virginia. When Stager became head of the Military Telegraph Department in Washington in October 1861, the Union Army put the components of Stager’s cipher into widespread use.
The key to the success of Stager’s cipher was absolute secrecy. Typically, fewer than a dozen people had access to any cipher. These small books became the most important piece of property traveling with an army. Even commanding generals and other telegraphers did not have access to the cipher code. Large armies in the field generated thousands of telegrams, and only a minority of the most important were sent in cipher. The cipher telegrapher had to encode and decode any message sent in cipher himself. Once encoded, he could have another telegrapher send the message because the other telegrapher was only sending a series of words, without understanding their meaning. Similarly, a telegrapher could receive a telegram in cipher, then pass it along to the cipher telegrapher for decoding.
In addition, the War Department used multiple ciphers simultaneously, so different armies communicated highly sensitive materials to Washington using different codes. Only the cipher telegraphers in the War Department knew which codes were used by which armies. If a cipher telegrapher were captured by the enemy, as happened at least once, the telegraphers switched to a different cipher to communicate with that army.
Like messages for thousands of years, Union cipher telegrams used word substitution to obscure the meaning of the message. Stager developed code words for times of day; important political and military leaders, from both the Union and the Confederacy; cities, states, and rivers; and important military terms. Substituting arbitraries, or code words, for key words in a message made them difficult to understand. In addition, Stager added arbitraries for punctuation. In Cipher No. 1, for example, “pedlar” and “Pekin” were arbitraries for a comma, and “star” was an arbitrary for the word “interrogation” or a question mark, depending on context. “Unity,” “zodiac,” and “zebra” were arbitraries for a period.
For frequently mentioned persons, multiple arbitraries referred to a single person. For example, in Cipher 9, used from late 1862 to mid-1864, the arbitraries for President Abraham Lincoln were “Adam” or “Asia.” In Cipher 1, used from mid-1863 through the end of the war, arbitraries for Lincoln included “Bologna,” “Bolivia,” “Ida,” “ink,” “Irving,” “ingress,” “ingrate,” and “ingot.” Only the cipher telegrapher would know that any of those eight arbitraries referred to President Lincoln.
The use of arbitraries in cipher messages provided a great deal of security for important messages, but arbitraries might only replace 10 to 20 percent of the key nouns and verbs in a message. Through careful study, Confederates might have been able to decipher one or more of the Union ciphers, but Stager’s system had another feature that made cracking the Union ciphers much more difficult.
That is our next blog post!
Decoding the Civil War 