Key is in NARA RG92E623

I show here sentences extracted from C1 at the hand of DOI(Beale papers version).
1. Abolish absolute and Britain, but consent disposed elected forms.
2. And be bands bodies causes created. Causes colonies deriving direct ends.
3. Another colonies becomes creator. Declare design. Elect(ed) legislation.
4. Alter bodies changed. Direct equal from governments. Human is it. Just King laws mankind measures. Nature of history people.

Interesting point is that authors of puzzle tried to show us the full stop of sentences by repeating the same word. Nice idea, but it was difficult to realise, cos many words were mentioned in DOI only once.
So intead of FORMS, FOR was used. Intead of ENDS, AND was used, Instead of ELECTED, LEGISLATION was used. Only word PEOPLE was mentioned many times, so 4th section finished correctly - PEOPLE PEOPLE.
 

Have a nice day neighbor ). I'm from Latvia. And i'm going another way, so long way ). Decoding of Beale is in several "layers"\stages using Declaration. First stage contains the instructions for second stage, second stage contains the instructions for decoding(i think) "1" and "3". I'm on 2nd stage now. But big problem for me that the english is not my native language. I'm russian speaking. And i can't know how correctly was build some phrases in english.

P.S. I think that cipher not about the treasure. I guess the cipher creator is Thomas Jefferson, 3rd president ). And about the Lewis and Clark expedition started in St. Louis, MO.
 

Last edited:
Daniel Boone created the code.

One key clue to decipher was in his, "killed a bear" carving on a tree.
Since we started to talk about movies, I recommend to watch "The Good, the Bad and the Ugly (1966)".
Especially episode 1h10min47sec. Carlson talks.
-I hid the gold. The gold is safe. In the cemetery.
-Which grave? Have a name? Have a number?
Well, have you seen books in this movie? No books, No key.
Well, numbers are in cemetery. Of course, it is a most safe place numbers to hide.
 

Can someone English speaking tell me, can the phrase marked in red be correct? It's blow my mind. It seems to me it's the waste of words ).
"... people laws form formidable history the these province for that becomes it for us have that{or "that have"?} destruction example and a constitution sent into the off ..."

-------------------------------------
But I think my way is the only true one ).
 

Last edited:
Beale Papers, decrypted text: six feet below.
Common Idiom. It's a well-established idiom referencing death and burial.
It is not about the depth of the treasure, it is about Beale Papers link to cemeteries.
Another golden coin for a Cemetery version.
 

It is a one more hint.
Let's again take a look at Gillogly Keys. Location in B1 - 37. http://www.bealetreasurestory.com/id39.html
B1 Cipher Numbers: 758 485 604 230 436. Letter String Decoded with Beale Papers version of DOI is:
AAAAB

Correspond words are:
ABOLISHING A A ACCUSTOMED BODIES

Initially I thought it is a noise, and didn't pay attention. But now with help of AI I could say it is a meaning in these words.

Let's analyze 3 words here: abolishing accustomed bodies. Make questions and ask AI.
What is 'abolishing bodies' in context of 19th century USA?
What is 'accustomed bodies' in context of 19th century USA?
Answers of ChatGPT on 16:50 GMT 6/5/2024:

In the context of 19th century USA, "abolishing bodies" could refer to:
1. Abolitionist Organizations: Groups and societies dedicated to the abolition of slavery.
"Abolitionist bodies like the American Anti-Slavery Society played crucial roles in the fight against slavery."
2. Dismantling Institutions of Slavery: Efforts to dismantle the legal and social structures that supported slavery.
"The movement focused on abolishing bodies of law that upheld slavery, aiming for complete emancipation."
3. Ending Enslavement of Individuals: Freeing enslaved people, thus abolishing the control over their bodies.
"The Civil War and subsequent legislation were pivotal in abolishing the enslavement of African American bodies."

In the context of 19th century USA, "accustomed bodies" could refer to human bodies that have become habituated or adapted to certain conditions, environments, or practices of that time. Here are a few interpretations:
1. Enslaved Individuals: Enslaved people whose bodies were accustomed to the harsh physical labor and brutal conditions of slavery.
"The bodies of enslaved individuals were accustomed to the grueling labor demanded by plantation work, enduring severe physical and emotional hardship."
2. Industrial Workers: Factory workers and laborers whose bodies were adapted to the demanding and often dangerous conditions of early industrial work.
"As the Industrial Revolution progressed, the bodies of factory workers became accustomed to long hours and strenuous tasks in often hazardous environments."
3. Frontier Settlers: Pioneers and settlers who were used to the physical demands of life on the frontier, including farming, building, and defending their homesteads.
"Frontier settlers’ bodies were accustomed to the rigors of clearing land, constructing homes, and surviving in the wilderness."
4. Military Personnel: Soldiers who had adapted to the physical challenges and rigors of military life, particularly during the Civil War.
"During the Civil War, soldiers' bodies became accustomed to the harsh conditions of battlefields, long marches, and inadequate medical care."

Slavery is a topic of first Gillogly Key. And I know why. There is interesting work by Tim Dennee. 'African-American Civilians Interred in Section 27 of Arlington National Cemetery, 1864-1867.'
He writes: One large plot is given over to the graves of ‘contrabands,’ runaway slaves who died under the protection of the Union forces around Washington.
Well, here I must again insert a picture of Section 27. So, I repeat again. Beale papers is a brilliant paperwork of brilliant patriot about his brilliant country.
section 27 up to 1263.jpg
 

Let's talk again about Gillogly Strings.
Strings are here:
758 485 604 230 436
150 251 284 308 231 124 211 486 225 401
25 485 18 436 65 84 200 283 118 320 138
24 283 134 92 63 246 486
147 436 195 320 37 122 113 6 140 8 120 305 42 58 461 44 106 301 13 408
Well, there are 53 numbers of all 5 Gillogly strings. What I did?
I took all 53 numbers; also I have divided entire 1st ciphertext into 10 sets by 53 numbers. I got totally 11 sets. I plot these 11 sets as Sorted Value Plot.
The result is here:
Sorted Value Plot2.png

Sorted Value Plot1.png
We see that Gillogly Strings numbers line is very different than others. Obviously it comes from different source than other numbers. Well, my recommendation, if somebody wants to do a brute force attack on ciphertext 1, it is better to remove from consideration a Gillogly Strings numbers.
 

Last edited:
I show here sentences extracted from C1 at the hand of DOI(Beale papers version).
1. Abolish absolute and Britain, but consent disposed elected forms.
2. And be bands bodies causes created. Causes colonies deriving direct ends.
3. Another colonies becomes creator. Declare design. Elect(ed) legislation.
4. Alter bodies changed. Direct equal from governments. Human is it. Just King laws mankind measures. Nature of history people.

Interesting point is that authors of puzzle tried to show us the full stop of sentences by repeating the same word. Nice idea, but it was difficult to realise, cos many words were mentioned in DOI only once.
So intead of FORMS, FOR was used. Intead of ENDS, AND was used, Instead of ELECTED, LEGISLATION was used. Only word PEOPLE was mentioned many times, so 4th section finished correctly - PEOPLE PEOPLE.
Yes, you can pick any subject, any topic, and can create strings of cleartext relating to that "chosen subject or topic with any random set of numbers."

Per example, (which you won't do) grab any book off of the shelf and you can "manufacture" similar seemingly related strings of cleartext.

This same type of thing is nothing new to the Beale story, all been done before time and time again.
 

Yes, you can pick any subject, any topic, and can create strings of cleartext relating to that "chosen subject or topic with any random set of numbers."

I want to emphasize two points.
1. These 53 letters weren't provided by me. It was extracted by Mr.Gillogly.
2. I would like you to create an algorithm, based on it extract some data out of Beale ciphers or letters, and let AI evaluate it like "Sentences are grammatically correct and semantically meaningful, although they are somewhat unusual. However, it is still important to note that the probability of the sentence being generated accidentally is still quite low". We will look at it :-)
 

Here I put some more reliable info regarding Gillogly strings. Wanted to back my assumptions not based on pictures, but on numbers, on math function used to compare few data sets - Coefficient of Variation (CV). So, CV < 1 are considered low-variance. Look at calculations - Gillogly numbers (and words) come from other space.
1000002650.png
 

Other than proving "randoms numbers can be random" I'm not sure what you're trying to establish?
 

Do you know what is Coefficient of Variation (CV)?
In essence, it's a highly subjective means for calculating "probability" in unknown means.

"Unknown Means" is the key phrase here as I could simply, at complete random, arrange a wide margin of meaningless random numbers on a sheet of paper (just like the alleged Beale codes) and then watch people try to apply "Coefficient of Variation" to arrive at a number of "suggestive probabilities."

"They are random numbers with no know or identifiable means to begin with."
 

Do you know what is Coefficient of Variation (CV)?
And I'll even go one further. What happens if those alleged clear text are written, in say, Hebrew, Russian, Dutch, or even French-creole? What's this do for your "Coefficient Variation?"
 

In essence, it's a highly subjective means for calculating "probability" in unknown means.

"Unknown Means" is the key phrase here as I could simply, at complete random, arrange a wide margin of meaningless random numbers on a sheet of paper (just like the alleged Beale codes) and then watch people try to apply "Coefficient of Variation" to arrive at a number of "suggestive probabilities."

"They are random numbers with no know or identifiable means to begin with."

I see, I you do not understand The Coefficient of Variation. The aim of it is not about of probabilities. The aim of Coefficient of Variation is to expose a distinction of one dataset from another dataset or group of datasets.
Your expression "try to apply Coefficient of Variation" sounds funny to me, cos I do not "play" with random numbers. Mr.Gillogly got these 53 numbers. Well, the Sorted Value Plot together with Coefficient of Variation of Gillogly numbers gives us the hint that these numbers are not random.
The answer is in NARA RG92E623. Only this book could show, whether am I right or not.

>And I'll even go one further. What happens if those alleged clear text are written, in say, Hebrew, Russian, Dutch, or even French-creole? What's this do for your "Coefficient Variation?"
I can't understand this question. After decryption text in any language is acceptable.
 

I see, I you do not understand The Coefficient of Variation. The aim of it is not about of probabilities. The aim of Coefficient of Variation is to expose a distinction of one dataset from another dataset or group of datasets.
Your expression "try to apply Coefficient of Variation" sounds funny to me, cos I do not "play" with random numbers. Mr.Gillogly got these 53 numbers. Well, the Sorted Value Plot together with Coefficient of Variation of Gillogly numbers gives us the hint that these numbers are not random.
The answer is in NARA RG92E623. Only this book could show, whether am I right or not.

>And I'll even go one further. What happens if those alleged clear text are written, in say, Hebrew, Russian, Dutch, or even French-creole? What's this do for your "Coefficient Variation?"
I can't understand this question. After decryption text in any language is acceptable.
You do realize that you're practicing "theory", yes? The problem with "in theory" processes is that they are subject to just that, "in theory." And you failed to reference the "means" that there is any actual clear text to be had at all in the remaining ciphers? Without that "means" then the door becomes wide open to multiple "in theory" summaries depending "who is practicing that in theory process." The simple fact there exist no such evidences to suggest that either of the remaining ciphers actually contain clear text opens the door to individual summations, of various processes, concluding that they do......lol

The only possible way "your theoretical process" would stand any chance at all is if you were provided with a known keyword that exist in the alleged undecoded cipher.

"What is the point of seeking reality if we are to deny those realities that we don't want to recognize?"
 

I see, I you do not understand The Coefficient of Variation. The aim of it is not about of probabilities. The aim of Coefficient of Variation is to expose a distinction of one dataset from another dataset or group of datasets.
Your expression "try to apply Coefficient of Variation" sounds funny to me, cos I do not "play" with random numbers. Mr.Gillogly got these 53 numbers. Well, the Sorted Value Plot together with Coefficient of Variation of Gillogly numbers gives us the hint that these numbers are not random.
The answer is in NARA RG92E623. Only this book could show, whether am I right or not.

>And I'll even go one further. What happens if those alleged clear text are written, in say, Hebrew, Russian, Dutch, or even French-creole? What's this do for your "Coefficient Variation?"
I can't understand this question. After decryption text in any language is acceptable.

Look, you're completely missing the reality of the situation. Do you even realize just how many ways numbers can be used in the creating of a cipher? “A lot of different ways!” This is your first problem right out of the gate.

Take c1, per example. It contains a much wider range of code and also 19 four-digit codes. It could “very-easily” be those 19four-digit codes represent words that form a single sentence, all the other remaining code simply being random ghost code.

The problem is that “you are assuming way too much” in the “practicing of your in-theory process that just by its very nature the practice contains many uncertainties and flaws when there exist, and here it comes, no known means of the applied coding process.” And there could be dozens, if not hundreds of different ways, to use numbers in the drafting of a cipher. And the two remaining ciphers already possess strong suggestive evidence that they have indeed been drafted differently.

Without the “means” you are simply left to take wild stab, or “desired stab” as to the applied coding process right from the very start.

And I might add, since you doubted my knowledge of these things, this was just pulled from online resources,

“The coefficient of variation (CV) is the ratio of the standard deviation to the mean. The higher the coefficient of variation, the greater the level of dispersion around the mean. It is generally expressed as a percentage.” Or, in the case of uncertain or unknown conclusion, such as the remaining two ciphers, “probability in theory.”
 

I see, I you do not understand The Coefficient of Variation. The aim of it is not about of probabilities. The aim of Coefficient of Variation is to expose a distinction of one dataset from another dataset or group of datasets.
Your expression "try to apply Coefficient of Variation" sounds funny to me, cos I do not "play" with random numbers. Mr.Gillogly got these 53 numbers. Well, the Sorted Value Plot together with Coefficient of Variation of Gillogly numbers gives us the hint that these numbers are not random.
The answer is in NARA RG92E623. Only this book could show, whether am I right or not.

>And I'll even go one further. What happens if those alleged clear text are written, in say, Hebrew, Russian, Dutch, or even French-creole? What's this do for your "Coefficient Variation?"
I can't understand this question. After decryption text in any language is acceptable.
And one last thing when suggesting that, "data suggest that it "probably" contains a clear text." You'd better know exactly what language you're dealing with first because spelling (another assumed means) obviously isn't the same in every language .....lol
 

Top Member Reactions

Users who are viewing this thread

Back
Top