book cover
From FLYNN's July 9, 1927

SOLVING CIPHER SECRETS

Edited by M. E. Ohaver
AN ILLUMINATING DISCUSSION ON WHY CIPHERS SHOULD BE UNDECIPHERABLE
WITHOUT THE KEY—ALSO, NEWS ABOUT WINNERS IN THE "CRYPT" CONTEST

WITH reference to No. 7 of the ten cipher essentials published in the May 28 installment of this department, it is a self-evident proposition that a cipher, to afford complete security, should be absolutely undecipherable without the key.

To permit free use of the cipher, it should not only be safe for single messages, but also proof against methods which combine a number of cryptograms for analysis. And this state of absolute security should persist even if the decipherer is in full possession' of the method and equipment.

Kerckhoffs, speaking of military ciphers, says: "It should cause no inconvenience if the apparatus and methods should fall into the hands of the enemy." And for obvious reasons this statement applies equally as well to ciphers for other purposes where the methods and equipment are liable to seizure, exposure, or loss in any way.

This is not saying that available ciphers actually possess all these qualifications. For example, the unauthorized sight of a code book, with opportunity to copy or memorize certain portions of it, may be sufficient to compromise the whole code.

Also, there may be cases where absolute security, though desirable, may not be necessary. Or, because added security in a cipher usually means added complexity, it may not always be practicable to use the safer systems.

The hurry, excitement, and mental strain under which military ciphers must be used, for instance, precludes the employment of the more complex "hand methods "in warfare. Fortunately, a military cipher is only required to offer enough resistance to decipherment so that any information in intercepted communications will be of little or no tactical value to the enemy by the time translation has been effected.

This does not mean that undecipherable systems would be unfit for military service; but that available systems which are simple enough to use under the adverse conditions may not be such as to guarantee absolute security. Many military ciphers are thus only "delaying" ciphers. They possess only the requisite margin of safety, and will break under sufficient strain. Some ciphers of this kind will be described in early issues. The "Playfair "is a famous example of them.

It may be possible to devise ciphers the security of which depends upon keeping the methods and apparatus secret. And such systems may be more or less effective in a limited way.

Finally, if for any reason it is desired to use a purposely weak cipher with the expectation that it be solved, there are plenty of ciphers to choose from. The security of ciphers, as covered by No. 8 of the above mentioned essentials, will be further considered next week.

Last week's cipher No. 36 used the simple substitution alphabet, BFJNR VZEKP UAHOW DMXIT LCYGS Q, formed by taking B and counting out every fourth letter thereafter in rotation. The message: "Tropical king's jewels quickly become exhausted; lazy, jealous wives ferociously attack royal hoard, dancing indecorously."

To decipher No. 37, first transcribe it by alternate verticals—left to right—into a rectangle of five lines by fifteen columns, the first line of which will begin ATEON ... The message: "All ciphers can be solved by perseverance, but I could not prove the truth of that statement," can then be read by alternate diagonals, beginning at the upper right-hand corner.

No. 38—modified Bazeries—used an alphabet based on the key letters NB—142, or ONE HUNDRED FORTY-TWO—according to the plan described last week, excepting that the key letters were reversed in prefixing them to the cryptogram, and the letters O-R-B-M-Z, forming one of the diagonals in the key, were used as nulls instead of the vowels. The message: "The enemy has been heavily reinforced on a five-mile front and is preparing to advance to-morrow morning."

The first of this week's problems, No. 39, is a simple substitution cipher. The alphabet is based on a key word, which you should try to find. The man referred to in the message is badly in need of it.

Cipher No. 40 uses a key word, and can be worked on the Vigenere table—see issue of February 20, 1926—by using the block alphabet for the message instead of the cipher. It not being practicable to reprint the table here, the following method of working the same cipher is appended:

(a)  C  O  M  E   A  T    O  N  C  E
(b)  2  14 12 4   O  19   14 13 2  4
(c)  L  A  W  L   A  W    L  A  W  L
(d)  11 0  22 11  O  22   11 O  22 11
(e)  17 14 16 19  O  23   3  13 6  19
(f)  R  O  Q  T   A  X    D  N  G  T

For example, to encipher the message (a), using the key word LAW, (c), first substitute at (b) the numerical values from a, straight.A=0, B=1 ... Z=25 alphabet. Subtract the key numbers (d) from these numbers, adding 26 whenever necessary. The resulting differences (e) are then transformed into the letters (f) at the cryptogram, using the same alphabet.

No. 41 is a key-phrase cipher of the type described in FLYNN'S WEEKLY for January 22, 1927. The key to this cipher consists of a twenty-six-letter phrase or sentence, written letter for letter under the twenty-six letters of the normal alphabet; any message letter being represented in cipher by the corresponding letter of the key phrase.

Any cipher letter which is repeated in the key phrase acts as substitute for the several message letters occurring above it in the message alphabet. As a result of this the key-phrase cipher is much harder to solve than.,the simple substitution cipher.

To make the present example still more difficult, the normal word divisions have been discarded in favor of the grouping by fives; and a foreign, instead of English, key phrase has been employed.

CIPHER No. 39.

NMP  NMG  HEHML  NB
NHO  LT  AQMBLCO.

CIPHER No. 40 (James Ricker, Pittsburgh, Pa.).

OZL   NIRRNV   CDFNFL   SAIIBN
SUE   BWEICD   RJPU   GKEEH   LI
FYS   UHU   YDRM   HPFNU   WOG
TDL   ZIUHU   LJOZ   AA   CUYT
TD   ZVFJFF   DJLU   X   GAUF
HKA   AAI   TGBS   NATRH   TVKO
POP   J   WDOY   GM   JN

CIPHER No. 41.

EEERE  FESNR  IIMTE  MRRSR
SELTE  SRERS  SIEES  SIERS
EEERS  SERNI  TREEM  EIMIT
SERRS  NRSEN  RSSIT  SEFER
SITSM  ESSES  EEEER  RRSES
NETSE  SOIET  REETN  SSIME
ISENS  NEEEM  SIESE  FENMR
RPJTS  TMNRE  EEMMI  SINRT
EESSI  NMMIS  STEEI  EEPTN
SNRSE  E

The answers to Nos. 39 and 40 will be published next week. To give the fans a better chance at No. 41, its solution wiH be withheld for two or three weeks.

Just a word about new ciphers. In submitting original ciphers for publication be sure to include the full explanations.

Incidentally, the list of prize winners in the "crypt" contest of March 19 will also be published next week.