From FLYNN's January 14, 1928

HOW TO DETERMINE THE LITERAL KEYS FROM WHICH THE NUMERICAL

KEYS TO TRANSPOSITION CIPHERS ARE DERIVED, WITH SOLUTION TO No. 115

KEYS TO TRANSPOSITION CIPHERS ARE DERIVED, WITH SOLUTION TO No. 115

THE various methods of solving single and double transposition ciphers recently discussed in this department have so far not considered the determination of the literal keys from which the numerical keys to these ciphers are derived. A few words on this interesting feature of the solution should therefore not be amiss.

As our readers have already been informed, a numerical key of this type if formed
by numbering the letters of the literal key from *1* up, according to their relative
positions in the alphabet, taking repeated letters, if any, from left to right.

In the key word COMMUNICATION, for example,
*A* would be numbered *1*. Next, since there
are no *B*'s, the two *C*'s would be numbered *2* and *3*, left to right, respectively.
The *I*'s, next in alphabetical order, would similarly be assigned the numbers
*4* and
*5*. The whole key, treated in the same manner, stands as follows:

C O M M U N I C A T I O N 2 -10-6 -7 -13-8 -4 -3 -1 -12-5 -11-9

The preparation of a numerical key from the literal key is thus a very simple matter. But the determination of a literal key from a given numerical key is not so easy. Numerous methods may be applied, depending upon the length of the key, letter frequencies, sequences, terminations, et cetera. But each key will present its individual problems to tax the ingenuity of the solver.

A good way to begin, especially with longer keys, is to find the alphabetical limits for each number. In the above key, for example, it will thus be seen that the first key number, 2, must represent some letter between B and U, inclusive; that 10 must stand for a letter between E and X, inclusive; and so on.

2 -10-6 -7 -13-8 -4 -3 -1 -12-5 -11-9 B E D D G D C B A F C E D U X W W Z W V U T Y V X W

These limits will, of course, decrease as the key increases in length. But in any event, the process may be further simplified by guessing one or more letters. Here, for example, looking for a probable suffix, we find that ION—probably SIGN or TION—fits like an old shoe, at the same time regulating the limits in the remaining columns as follows:

2 -10-6 -7 -13-8 -4 -3 -1 -12-5 -11-9 BOJ J T J C BASIONC K KUK DCBTD L L V L E D O EMMW M F E D F N N XNG F E G Y H G F H ZIK G I I H I

The literal key may now be further developed, by trying various combinations of
letters in the several columns. Thus, *13* probably stands for the vowel,
*U*, since
the adjacent columns contain only consonants. Again, if *2* and *6* are given the values
*C* and *M*, respectively, we have the prefix *COM*. The whole key may be evolved in the
same manner.

Short keys are often capable of more than one interpretation. Thus, the key to Cipher No. 90, of November 5, was CIPHER. But one reader submitted the key DORMER. Both of these words give rise to the same numerical key, 1-4-5-3-2-6. And without doubt there are still other words that would do the same thing.

To afford the reader some fun in experimenting with keys of this sort, we are offering
three examples under Cipher No. 121. The first of these *(a)* may provide several
answers. The second *(b)* should prove no more difficult than the example in this
article. The third key *(c)* is practically of message length, and in some ways should
prove most interesting of all. At any rate, fans, search out these keys, and send
us your findings.

Did you get all tangled up in last week's No. 116? Here is the solution:
*Q*: What
is a labyrinth? *A*: A place full of intricate windings which render it difficult
to find the way between the entrance and central compartment.

No. 117, by Paul Napier, conveyed the message, "Solving ciphers is a fascinating
task." This system used the well known Vigenère square of twenty-six alphabets
with the numbers *1* to *26* down the side, and—beginning with A=*13*—across
the top.

Each pair of numbers acted as symbol for a letter. Thus,
*14-19* stood for *S*, found
at the intersection of the *14th* row and the *19th* column. An interesting peculiarity
of this system is that it permits a repetition of numbers in the cryptogram quite
baffling to the uninitiated, perhaps, but offering no difficulty to one having the
key. How did you make out with it?

No. 118, double transposition cipher, used the key word SKIRMISH—7-4-2-6-5-3-8-1—and conveyed the message: "Frustrate the enemy's attempts to organize retreat.s under cover of darkness." Through the omission of the null required to fill out the last cipher group of five letters, this cryptogram formed an eight by eight square, easily decipherable by the method given last week.

Double transposition Cipher No. 115, of two weeks ago, used the key word FIELDPIECE—6-7-3-9-2-10-8-4-1-5. The message: "Scouting planes report reenforcement of enemy line on five mile front extending from railway junction west to river." This cryptogram formed a ten by ten square, also decipherable by last week's method.

Turning now to this week's ciphers, try James McGuire's No. 119 if you want something good, but not too difficult in a straight substitution cipher. Then let Mr. Bellamy keep you guessing for awhile with the apparent words ITEM, MERIT, DAT.A,, et cetera, scattered throughout his transposition Cipher No. 120. We have already covered No. 121.

CIPHER No 119 (James McGuire, Benton, Illinois).

4628164648503730582739391048482759164628 5048471710571957101710384716584048164747 19501830

CIPHER No. 120 (Arthur Bellamy, Boston, Massachusetts).

ITEMT TIIAE SMERI TGROT STAOH ZDATA HNNAA EISOP SCSGS WREHP ICMTZ FOEME HEHSI EGASS

CIPHER No. 121.

(a)4-3-1-6-5-2.(b)2-8-11-6-10-5-3-9-1-7-4-12.(c)52-21-7-8-62-27-33-53-22-1-54-35-9-37- 4-40-34-28-63-10-49-2-18-55-11-45-56- 23-12-36-57-24-13-20-41-42-5-29-50-43- 19-58-30-38-59-14-46-47-15-6-64-31-60- 25-61-26-16-32-48-3-44-39-17-51.

Whoop 'em up, fans! Keep your answers coming. And send in your new ciphers, too! But remember to include full explanations with them. The solutions to this week's ciphers will appear next week.

Letters from cipher fans indicate that Solving Cipher Secrets is still one of the most popular departments of the magazine, and is still growing in the scope of its interest. And all of you know that the more readers who make their mterest active, the more readers there will be to enjoy the department to the full.