book cover
From FLYNN's June 4, 1927


Edited by M. E. Ohaver

A CIPHER conforming to the requirements of all ten propositions in last week's department might, at first thought, seem as elusive as a will-o'-the-wisp.

Upon closer consideration, however, such a cipher might not be relegated altogether to the realms of the unattainable. The first proposition has already been dealt with in the issue of March 12. And the second, which requires a cipher to express a given message in the shortest practicable cryptogram, will be briefly covered here.

The relative length of a message in its clear and enciphered form has an important bearing on the rapidity, economy, and consequent value of the cipher.

In most ciphers of recognized merit, the message and cryptogram are substantially of the same length. And, as a general rule, anything tending to lengthen a cipher should be avoided.

In literal ciphers, for example, even the word space is usually dispensed with, since it would increase the number of characters by about a fifth, with a consequent slowing up in the processes of encipherment, transmission, and decipherment. The high speed printing telegraph cipher used during the World War may be mentioned, however, as an important system employing the word space.

It would ordinarily be better if a cipher message were even shorter than its translation. And in this connection code books come up for favorable mention.

Code message No. 3, in the April 23 issue, for example, expressed a ten-word message of sixty-one letters by means of only fifty letters in code. Phrase codes, like the Western Union Telegraph code, and the A. B. C. code, in which a single code word may signify a whole sentence, are especially brief forms of cipher.

There may be times, admittedly, when the length of the cryptogram might work no particular hardship even if it were several times longer than the message. But such a system could hardly be put to any general use.

It would not be practicable commercially over telegraph or cable, for example, on account of the needless additional cost of transmission. Similarly, its use as a military cipher might result in serious delay or disaster.

While brevity in a cipher is thus greatly to be desired, nevertheless a longer cipher, on proper occasion, might not be without its own peculiar advantages.

For example, the additional bulk in No. 25, below, permits the cipher to have two readings which may be quite different. If the recipient is "put to the torture" he can explain that the meaning is shown by the odd letters, which read: NO LIQUOR WILL BE SHIPPED THIS WEEK. But how about the real, or inner, message? Can you find it?

CIPHER No. 25 (M. Walker, Akron, Ohio).


No. 26 uses an alphabet of short curved and straight lines—here represented by U's and hyphens—found in an old work on cryptography by Gustavus Selenus, published in 1624. No particular advantage is gained in this instance by the additional length; but it is interesting for several reasons, one of which you will find in the message. The use of a dot after each symbol is in accordance with the original.

CIPHER No. 26.

--UU.  UUU.  --U.  UUU--.  ----U.
-----.  --U--.  UUU--.  UU--.  UUU--
--U.  UU.  ----U.  --U--.  --UU.  UUU.
----U.  UUU--.  ---. ----U.  ----.
UUU.  --U.  ---U.  -.  ---U.  --U.
UUU--.  UU--.  -----. -----. --U.
UUU--.  --UU.  ----U.  UU--.  --U.
U--U.  U--.  --UU.  UUU.  --U.  UU.
U--U.  --UU.  UUU--.  -.  --U--.  UU.
UU.  -.  UUU--.  UUU.  --U.  UUU--.
U--U.  U--.  --UU.  UUU.  --U.  U----.
U--U.  ---U.  UUU--.  --U.  ---.
U--U.  UU.  --U.

The "foreign language" cipher of William Blair—last week's No. 23—used a multisubstitutional alphabet in which each letter has one or more syllabic equivalents. As much of this alphabet as Blair used in his example is subjoined:

A al, do, lp N gi, in, ni, sp, yi
B au O ra, xa, za
C co P br, lo
D du, ng, th, ul Q  
E el, fa, ka, ma, pa, va R on, ri, sc, si, ti, zi
F mo S or, ro, xo
G   T ho
H di, hy, qui U jo, qua
I ga, na, ya V fu, mu
J   W que
K X bu
L bi, jy Y ne, ye
M my Z

There seems to be no special order in this alphabet unless it was Blair's intention for any symbol and its reversal to have the same meaning. This is suggested by the use of in and ni for N, and or and ro for S.

In enciphering, substitutes were selected and grouped with a view to forming pronounceable cipher "words." Some of the symbols are fragmented; for example, ma—E, divided between the second and third groups, as here shown.

Cryptogram:  Si-ka jy-ga-m  a  fu-va  ...
Message:     R E   L  I  E     V  E   ...

Blair followed the original punctuation. The full translation was given last week, and can easily be worked out with the above alphabet.

The alphabet and translation of Blair's example of "plain ciphering "—last week's No. 24—follow:

    A   B   C   D   E   F   G   H   I
    31  32  33  34  35  36  37  38  39

J   K   L   M   N   O   P   Q   R   S
10  11  12  13  14  15  16  17  18  19

T   U-V W   X   Y   Z
20  21  22  23  24  25

Message: I have entrusted this bearer with many things as to the managing of the affair requisite for you to know. Remember that on no account you fail to meet me at nine to-morrow's night, for we must not now delay the surprise of the castle when things are just ripe, for execution.

Blair intended this cryptogram for the beginner, and it is full of instructive situations. The first word, "39," for example, is probably A or I. And near the middle occur the groups, "15, 14.  14, 15," where just a little experimenting with reversible two-letter words will lead to "ON NO" as the most likely translation. Also, seven short words begin with "20, 30," suggesting that they are common words beginning with TH.

A few suppositions of this kind quickly lead to a discovery of the methodical arrangement of the alphabet, and to a filling in of the remaining letters without the need of finding them by the usual methods.

The answers and full explanations of this week's ciphers will be published next week.