SOLVING CIPHER SECRETS

I N last week's article, stress was laid on the element of brevity in cipher construction.

This week, the third of the ten propositions published in the May 28 issue—to the effect that the security or use of a cipher should not be dependent upon any limitations in the text of the message material— will be briefly considered. While this proposition is so self-evident as hardly to require any further comment, a few stray remarks may not be altogether out of order.

That a substitution cipher, for example, must be capable of communicating any text, does not necessarily mean that it must provide special symbols for figures, punctuation marks, or other characters. In most ciphers these are usually spelled out in full.

On the other hand, there are ciphers which make provisions for these characters. Bazeries, the French author on cryptography, for example, uses "K" and "KK" in one of his ciphers as indicators with other letters, as follows:

    A  B  C  D  E  F  G  H  I  J
K   .  ,  ;  :  '  !  ?  -  ¶  ()  K
KK  0  1  2  3  4  5  6  7  8  9   KK

According to this plan, which is suggestive of the "shifts" of the five-unit telegraph code, a period would be written KAK. Similarlv, the number 2500 would become KKCFAAKK. Either of which could then be enciphered in the regular manner with the rest of the message without materially increasing its length, or requiring any additional symbols.

To take up another angle of the above proposition, a cipher crops up occasionally which would limit the choice of words in preparing a message; suggesting, for example, that common short words be avoided, or that synonyms be used instead of repeated words.

The purpose of this advice is usually to render unauthorized decipherment more difficult by affording less material for analysis, upsetting frequency tables, and so on.

But such suggestions are somewhat impractical, since it might not always be convenient or possible to avoid the "telltale" combinations. Or they might be included inadvertently by the writer of the message.

While ciphers intended for general use should thus preferably allow complete freedom of expression, there are important systems, for especial purposes, which do not include this feature. Take telegraph and cable codes, for example, where the user is limited to the words, terms, or phrases, printed or written in the vocabulary.

And "Calloway's code," in O. Henry's "Whirligigs," might also be mentioned as a curious example illustrating how an extremely limited system, created in an emergency, might still be good enough to fulfill its intended purpose.

In this week's ciphers, readers are offered several interesting novelties.

No. 27 uses the Selenus alphabet exactly as given below, but without the dot (.) separators between symbols, as in the original cipher. To discover the correct divisions, even with the alphabet, should provide a fascinating, and by no means easy, problem.

No. 28, in straight letter substitution, uses the Bazeries shifts, shown above, for punctuation and figures. Find the substitute for the indicator, K, and you have a good start toward solution.

C. E. R., Hudson, Massachusetts, submits No. 29 with the following note: "I am sending you a cipher that I would like to see tried on the other boys. They have given me a lot of trouble, and I would like to get back at them."

This cipher is extremely simple; but it is our humble opinion that you will have to wait until next week's issue to know the solution.

CIPHER No. 27.

-----   U--U-   -UUUU   ----U
UUU--   --UUU   --UU-   -UUUU
--UUU   ----U   -----   -----
UUUU-   ----U   ----U   UU-U-
---UU   UU-UU   ----U   --UUU
UU--U   -----   ---U-   -----
----U   --U-U   UU---   -U

CIPHER No. 28.

KUHQT  ROPKU  PXRUY  JOJUX
JKWRN  PPIJW  DKAJQ  NOLJO
BHDDW  JIJDL  KPIJK  LMNKO
PJOVK  PAAWK  AAPRU  HZIPA
KAWJP  IJOJB  HPIRN  PYKHD
AKA

CIPHER No. 29.

DS  ZK  ZXO  YU  VWA  ZY  MCCH  DZ
YDO  ZE  CY  ZX  WUE  G  BZ  BJ  WZO
ZN  WYD, OD  WOCCB  AZ  L  WCR  GS,
ZY  YK  OF  RCCH RCC  ZXA  YJGGC  L
YW  OAGH  AY  ZOX  YK  WWA  XW?

There's the lot, fans! Do your level best, now; and write us how you make out. Answers to all three will be published in the next issue of FLYNN'S WEEKLY.

Last week's cipher No. 25, as previously stated, conveyed two messages. If "put to the torture" the receiver could reveal the "false" message, NO LIQUOR WILL BE SHIPPED THIS WEEK, formed by reading only the odd numbered letters.

But if the "inquisitors" should be skeptical, as they usually are, and should decide to toast the unfortunate correspondent on a hot grid, or throw him upon a bed of spikes, the chances are that he would come across with the following "true" explanation:

Take the letters by twos; add the numerical values—A=1 ... Z=26—of each pair; and subtract 26 from any sums in excess of that number. The resulting numbers will then represent the alphabetical places of the letters of the real message.

Whereupon the gleeful inquisitors would soon be in possession of the actual message: WE ARE SENDING THREE CASKS TUESDAY.

Cipher No. 26 employed the following straight substitution alphabet of Gustavus Selenus:

The message: THE SIGNS USED IN THIS CIPHER ARE SUGGESTIVE OF THE DOTS AND DASHES OF THE MORSE CODE.

Before closing shop, just a few words about ciphers submitted by readers. Now that solutions to all ciphers are being published in the next following issues, it is essential that ciphers be accompanied by the explanations and solutions.

We would also like to caution the fans about making their ciphers too complicated. We have scores of ciphers on hand that are too complicated to use. It may be all well enough to make a cipher difficult; but, to be practicable, it should at the same time be easy to use and describe.

However, we believe the majority of our readers prefer easier ciphers. Not too easy, perhaps, but not so difficult as to be tedious.

How about it?