How Sherman created C1

Manual typesetting: 101

   To begin, we need a primer on how type was set by hand. The images and technical information below are from the Melbourne Museum of Printing website.

Sherman's press at the Lynchburg Virginian probably looked like this:
	To create a page of print, type was set by hand, letter by letter into a frame called a "galley". In this image, note the empty squares. These are "formes" used to frame a block of text, or an advertisement. It is important to note that each iota of space had to be filled, including blank areas, to prevent the type from wobbling or falling over when pressure was applied in printing.
There were different sizes and fonts, and pieces of different sizes for blank areas, as well as lines. Individual pieces were called "sorts". These sorts varied in width according to the letter. For example, i and l were narrower than r and s, and w and m were wider still. Blanks for spaces also had varying widths to allow for "justification", eg. to make straight edges on both sides of the text.
The sorts were kept in a "typecase". The typesetter used a "composing stick" as shown here, to place the sorts he selected from the typecase to create text, then he dropped the selected text into the forme.

The Lewxian Extensions

    As the last step before we join Sherman in his composing room, I present below the Lewxian extensions. This discovery is by Robert Lewxian, who refers to them as "expanded Gillogly strings", but their importance is worthy of their own name.

    The decryption of C1 below is made using the Hart Papers version of the DOI using the first letter of the words as in the C2 decryption.

    Note in particular the longer words, including AFRICA, CHRIST, MORAL, FAITH, BOWL, PROTECT,  POLITIC, PROMPT, and TRANSIT. The significance of these words is that they are largely intact. A doubter may be tempted to conclude that these are simply accidental anagrams, but that is negated by their obvious relation in subject matter, suggesting a common origin.

    There are several shorter words including ACT, AVIS, DOT, FOE, HOT, MOB, PAW, PEA, SAT, TAP, TAR, TEA, TEN, and WENT, but the shorter the word, the more likely it is to be coincidental.

    The lower case x's are due to code values higher than the DOI word count.

SCSxETFAxGCDOTTUCWOTWTAAIWDBIIDTTxWTTAABBPLAAABWCTLTFIFLKILP

EAABPWCHOTOAPPPMORALANHAABBCCACDDEAOSDSFHNTFTATPOCACBC

DDLBERIFEBTHIFOEHUUBTTTTTIHPAOAASSTAATTOMTAPOAAAROMPJDRAxxT

SBCOBDAAACPNRBABFDEFGHIIJKLMMNOHPPAWTACMOBLSOESSOAVISPFT

AOTBTFTHFOAOGHWTENALCAASAATTAROSLTAWGFESAUWAOLTTAHHTTAS

OTTEAFAASCSTAIFRCABTOTLHHDTNHWTSTEAIEOAASTWTTSOITSSTAAOPIW

CPCWSOTTIOIESITTDATTPIUFSFRFABPTCCOAITNATTOSTSTFxxATDATWTATT

OCWTOMPATSOTECTTTOTBSOGCWCDROLITIBHPWAAExBTSTAFAEWCAxC

BOWLTSOACTEWTAFOAITHTTTTOSHRISTETOECUSCxRAIHRLWSTRASNITPC

BFAEFTB

How Sherman created C1

    Let us now spy upon John William Sherman working in his composing room circa November 1884.

    He is in the midst of the most tedious and boring job in printing: disassembling the galleys from yesterday's edition and returning the sorts into the typecase, one by one, thousands of them. As he is doing this, his mind wanders to his other project, the dime novel for the families bereaved by the great fire. His story needs another cipher to go with the two he already has. Why not now?

    He takes a standard forme from those he uses for display ads. It is very similar to the one he will use in two years for The Beale Papers advertisement. In so doing, he pre-determines the length of his new cipher: 13 lines of 40 characters, for a total of 520. It is slightly wider than high.

    Now he must fill this in. With thumb and forefinger, he lifts strings of sorts from yesterday's galley and puts them in the forme. The page he is lifting from contains the words listed above, but he is careful to split up the main words. This all very quick and without regard to, or knowledge of, cryptography. When he is almost finished, there are gaps here and there that need to be filled. He takes a handful of the letter A from the typecase and fills in the gaps. Also a few Ts, Ss, and Ds. Finally, there is one long gap remaining. With his composing stick he selects letters in alphabetical sequence from the typecase and drops them in the forme, but they are not snug enough. He inserts an extra I, M, and H to solidify the line, thereby making "ABFDEFGHIIJKLMMNOHP". Thus were created the Gillogly strings.

    Finally, he prints a galley proof of this "cleartext" and brings it home to encode it with the help of the table of DOI letters and word count he made previously. These numbers are later typeset as Cipher 1 in The Beale Papers.

Conclusive proof

    The preceding is a hypothesis on the creation of C1. I now propose a simple method by which this can be proven conclusively.

   In Page 24 of this website we state that The Lynchburg Virginian newspaper published display advertisements for The Beale Papers 85 times, starting April 15 and ending August 27, 1886. The dates and page numbers of these ads are well documented. This information is the result of painstaking research of the microfilm copies of the paper by Richard H. Greaves. These microfilms are available to researchers at the Jones Memorial Library, Lynchburg, VA.

    The final proof requires further detailed review of the microfilms from May 30, 1883, the date of the Lynchburg fire, following which the pamphlet project was conceived, and up to April 15, 1885, when the booklet appeared for sale. Some time in the middle of that date range is most probable. The researcher would look for a page, and perhaps a single article, which includes several of these words: AFRICA, CHRIST, MORAL, FAITH, BOWL, POLITICAL, PROMPT, TRANSIT. I have listed the larger words; shorter ones are detailed above.

    Finding such a page is half the battle. The final proof would lie in a comparison of the rest of the text of that page versus the rest of the C1 text. Other similarities and letter sequences should be evident which would identify that page as the absolute source of C1.

Progressive insight

   After reading the page above, Steve Margolis of California suggests the following:

    The word list proposed may be invalid because they are spread over different lines. The creator of the ciphers was in a hurry, which is why he took type from other galleys and put it on his composing stick. Having worked with typesetting equipment, I know that once you create a line, you add a slug to seperate it from the following line (as illustrated by the yellow lines below). It is very difficult to go back to a completed line and add type, so jumbling the letters between different lines would not make sense from a timesaving point of view.

    Working from the assumption that only the lines were jumbled, I took the Lewxian Extensions and came up with these anagrams, comprised only from letters on the same line:

    I therefore suggest the ciphers were constructed from stories, advertisements, and articles that ran around Christmas or Easter, hence the religious theme.

Comments

   Steve's amazing anagrams conceal a potential flaw. We are working with the ciphers from a manual transcription by Clayton Hart, then typewritten by George Hart, and the graphic used by Lewxian was from a book on the Beale codes. We are far from certain that the line breaks, which are fundamental to this idea, are the same as the original pamphlet. Nonetheless, it is significant that those lengthy words on a common theme can be anagramed from adjacent letters.

    Given the discovery of the Lewxian Extensions, this hypothesis is so simple and so logical as to be almost certain. This leaves C3 as the only dim hope left for treasure hunters. So far C3 has resisted all attempts at explanation.

Continuity

    During the winter of 2007/8, R. H. Greaves searched The Lynchburg Virginian microfilms for an article containing the words listed above. The search was both incomplete and inconclusive. Some issues of the newspaper are missing and probably forever lost. The microfilms can be viewed at the following facilities:

• Roanoke City Library

• Jones Memorial Library, Lynchburg

• University of Virginia, Charlottesville

• Richmond State Library, Richmond

    The films can also be obtained on loan from the Richmond State Library and Church of Latter Day Saints libraries thoughout the country.

Back to homepage