book cover
From FLYNN's August 27, 1927

SOLVING CIPHER SECRETS

Edited by M. E. Ohaver
PRESENTING A COMPLETE AND FINAL EXPLANATION OF CIPHER No. 44
OF THE JULY 16 ISSUE, AND A NEW QUESTION AND ANSWER CIPHER

I

N reconstructing the key to cipher No. 44 ( J. McA. T.), which was published at full length in the July 16 issue, and as below with a description of the system in the length of the groups.

Message:  IWOND   ERWHE   THERT
Cipher:   VZPOD   BWEZL   WBTLO

  HESIM   PLEAN   DDIRE   CTMET
  ACXPC   PRHGE   GBVXN   QMMMD

  HODOF   ENCIP   HERME   NTHER
  ZYXTF   RCGKS   AVIYE   OAEXY

  EUSED   ISREA   DILYD   ISCOV
  ICGHD   IBHUK   KLDLO   IUICT

  ERABL   EITHI   NKNOT
  BALAE   BWEQS   NZFET

Given the translation, this is easily done by factoring intervals between recurrent message groups which give rise to repeated cipher letters at respectively similar intervals in the message and cryptogram.

Here, for example, ER occurs at the sixth and seventy-sixth letters of the message, producing B's in the cipher, with an interval of seventy in both cases. Similarly, HE, at the sixteenth and forty-sixth letters, and the resultant A's, both occur at an interval of thirty letters. The only odd factor in common to these two numbers is 5, indicating groups of that length in encipherment.

Dividing into groups of this length, letters having the same units in common may now be discovered by comparing message and cipher letters in each group. In the first group, for example, the first units of the letters I, O, and D—message—are the same, respectively, as the first units of V, Z, and P—cipher. And the second units of I, O, and D are the same as the second units, respectively, of P, 0, and D—cipher.

By treating all the groups in this way, and placing letters with the same first units in rows, and those with the same second units in columns, the following tabulations are obtained:

 (Rows)     (Columns)
I V B E R  I│E│T│M│W
O Z H N A  P│L│O│C│Z
D p G Y K  X│G│D│K│R
T W M X F  S│N│V│A│Y
S C Q U L  H│F│Q│B│U

It is now possible to combine these squares into one, by rearranging the letters in the rows and columns. In doing this, observe that only the order of the letters in any row or column is changed.

(Reconstructed Key)  (Original Key)
  y x v w z             1 2 3 4 6
v I V B E R           1 G D K Y P
w P D K G Y           2 F T M W X
x X T M F W           3 E V B R I
y S Q C L U           4 N O A Z H
z H O A N Z           5 L Q C U S

Assigning the values v, w, x, y, z, to the rows, the next step is to find the values for the columns. And this may be done by finding letters which have the same units in first and second places. Groups of the present length afford four such points of comparison.

In the first group, for example, the first units of W and N—message—are the same, respectively, as the second units of V and Z—cipher—thus fixing the values x and z, respectively, for columns two and five. Similarly, the second units of W and N are the same, respectively, as the firpt units of O and D—cipher—fixing the valuess and w for columns five and four. The other values may be found by the same method.

A given key in this cipher can be transposed in fourteen thousand four hundred different ways, and yet give identical results in use. Note here, for example, that our correspondent's key—arranged in square form—and the reconstructed key produce exactly the same cryptogram.

(Reconstructed Key)    (Original Key(
I W O N D               I W O N D
v x z z w               3 2 4 4 1
y z x w x               5 4 2 1 2
V Z P O D               V Z P O D

Without the translation, group lengths could be determined by analyzing intervals between letters or series of letters. And predominating letters would be those using predominating units. Here, for example, the predominating unit 3, used forty-five times, occurs in five out of six of the predominating letters, E, B, A, C, I, L. Positions of letters in groups could also be turned to account.

 Did you get No. 56, last week's "question and answer" cryptogram? Here is the translation:

Question: What is the portrait parle, or
speaking portrait system?
Answer: This is a system of standardizing
personal characteristics so that they may be
accurately described in words.

No. 57 (F. H. Newell) conveyed the message: "I urge you to go away at once: you are in peril. " The keys were GO and PERIL. The alphabet was a straight A= 1 ... Z = 26, with & for zero. To encipher, first substitute numbers for letters of the message (a) and key (b); take the differences (c) between these numbers; and substitute letters, for numbers at (d).

(a)   I  U  R  G  E  Y  O  U
      9  21 18 7  5  25 15 21
(b)   P  E  R  I  L  P  E  R
      16  5 18  9 12 16  5 18
      ———————————————————————
(c)    7 16  0  2  7  9 10  3
(d)    G  P  &  B  G  I  J  C

It is possible for some of the numbers to offer a choice of two letters in deciphering, depending upon whether the message number was larger or smaller than its key number by the given difference.

Now for this week's ciphers. No. 59 is another "question and answer" cryptogram. Two alphabets are used; one for the question, and the other for the answer. There are thus really two ciphers here, the solution to either one of which will help in solving the other.

In No. 60 you are being offered a fine example of transposition cipher. This is not too difficult; but be careful! Some traps have been laid for the unwary to stumble into!

No. 61 is based on a key sentence which must contain all the letters in the alphabet, the symbol for a given letter being merely the two letters before and after it. For example, with the sentence, "Pack my box with five dozen liquor jugs," A would be represented by PC; B, by YO; C, by AK; and so on. But in this case a much longer key sentence has been used, providing several substitutes for the frequently used letters.

CIPHER No. 59.

Question: KSOJ  AT  ABFGHJALM
CMOTNBABR?

Answer: SVC EBHQCFF  HU  BCDFHJ-
KJM  UBHZ  D  EDBS  SH  D  IVHYC  UBHZ
EDBSKQOYDBF  SH  MCJCBDYF  HB
UBHZ  SVC  KJTKNKTODY  SH  SVC
OJKNCBFDY.

CIPHER No. 60 (M. Walker, Akron, Ohio).

EDIIH   NFCOR   NAWLQ   OESUS
NCCEF   TOENV   RLCCI   EEURN
RNDSE   EBIEV   DYUOI   ARFMQ
QGNK

CIPHER No. 61 (J. Levine, Los Angeles, Calif.).

SQEIV   METNO   MNUBI   KHEIK
WUEEA   SOTSX   UBDZE   TXUPR
UDUDF   HTEZS   UDENF   HASEN
ADEIE   POTNO   TEJOO   TUDOA
TESQN   OVMAS   DZEIQ   lEIHE
LJFHB   AVMEE   IKSXO   ATENM
ADNOT   EUCEI   KOHEE   IEIOT
XUZS