In Fig. 88, we have the principle of the sliding device by means of which this encipherment is accomplished. The Saint-Cyr slide is very easily prepared of cardboard, or of any other flexible and fairly strong material, but may also be prepared of wood, or may be set up for any temporary purpose on two strips of paper. Its details, also, may be varied to suit the operator’s own convenience. As shown, however, the upper and single alphabet, which is the plaintext one, is written on a card, and slots will be cut in this card at two points: Just below and to the left of A; and just below and to the right of Z. This plaintext alphabet is considered stationary.
| Figure 88 THE SAINT-CYR SLIDE ┌─────────────────────────────────────────────────────┐ │ A B C D E F G H I J K L M N O P Q R S T U V W X Y Z │ ┌───────┴┬───────────────────────────────────────────────────┬┴──────────── │ A B C D│E F G H I J K L M N O P Q R S T U V W X Y Z A B C D│E F G H I J.. └───────┬┴───────────────────────────────────────────────────┴┬─────────────┘ └─────────────────────────────────────────────────────┘ (To Y.) |
The lower and double alphabet, which is to furnish all of the substitutes, is written on a long narrow strip, the two ends of which may be inserted into the slots of the other card. This strip, or slide, may then be moved back and forth at will. However prepared, the spacing must be uniform throughout both alphabets. The Saint-Cyr cipher also makes use of a key-word in which each letter is the key to a cipher alphabet, and which is applied exactly as in Fig. 86 or Fig. 87. To apply the key-letter B, we adjust the slide in such a way that the B of the sliding alphabet will stand directly beneath A of the stationary one. This gives us exactly the same set-up which we used in [Chapter IX] for cases of simple substitution; that is, we have a plaintext alphabet with a cipher alphabet standing just below it; each plaintext letter is standing directly above its substitute, and each substitute directly beneath its original. The cipher alphabet just referred to, in which key-letter B, found in the sliding alphabet, is standing directly below the index-letter, A, found in the stationary alphabet, is identical with the B-alphabet of the Vigenère tableau, and is even called by the same name. Should we move the sliding alphabet, so as to place key-letter C directly beneath index-letter A, we reproduce the C-alphabet of the Vigenère cipher, again called by the same name. In the figure, we have the E-alphabet in position, with key-letter E standing directly beneath index-letter A. And since the sliding alphabet may be placed in 26 different positions, each time reproducing one of the Vigenère cipher alphabets, having the same key and the same name, it appears that our Saint-Cyr “cipher” is merely a duplication of the Vigenère. The chances are, then, that even though we call our cipher by its original name, and even make references to its tableau, our actual work of encipherment and decipherment will have been accomplished by means of the more convenient and rapid Saint-Cyr slide. But where a slide is possible, a cipher disk is also possible, and many will prefer to use the disk.
To prepare one of these, we might proceed as follows: First, cut out from cardboard (or other desired material) a pair of disks, one smaller than the other. Divide the peripheries of both disks into 26 equal segments, and write the 26 letters of the alphabet in a circle around both of the peripheries, causing both alphabets to run in the same direction. Place the smaller disk on top of the larger; and, finally, stick a drawing pin through the exact center of both disks, to serve as a pivot. The smaller disk may now be rotated to 26 different positions, so that any desired key-letter can be caused to stand beside index-letter A of the outer disk, and will place in position the cipher alphabet of which it is the key. The use of this revolving alphabet in place of a sliding one does away with the necessity for doubling its length.
Now let us examine carefully Fig. 89, with its two examples of decipherment. At (a) of this figure, a short cryptogram fragment, beginning T I Q. . . . , is being deciphered with the original key-word, BED, and is bringing out the message, SEND SUPPLIES. . . . . This, of course, is to be expected of any cipher. But at (b), it is this message fragment, SEND SUPPLIES, which is acting as a trial key; exactly the same process is being used as if applying the true key, and this decipherment is bringing out the original key, repeating over and over. The Vigenère cipher, then, works equally well in reverse, and in this respect it differs from some of its kindred ciphers. To understand this peculiarity, we have merely to consider the tableau. Concerning this we have said that the horizontal alphabet which stands across the top is the plaintext alphabet, and that the vertical one at the left is merely a list of keys. Suppose we decide to look at it the other way round, and say that the vertical alphabet at the left is the plaintext one, and that all 26 of the cipher alphabets are standing on end with their key-letters at the top, so that the horizontal alphabet, written across the top, is merely a list of these keys. Will there be any difference in the encipherment? Might the slide, also, be prepared in a vertical position? Does it make any difference in the results whether we encipher plaintext SEN by key BED, or encipher plaintext BED by key SEN?
One road to decryptment, then, is clearly indicated. If we have a probable word, we may use this word exactly as if it were the key, and, if it is actually present, it will bring out the true key. Or, if we have no probable word, we may try probable sequences, or make use of the trigram list. Here, however, we have two separate cases: The simplest, in which the probable word is long enough to bring out the key-word repeating; and the most difficult, in which the sequence, or probable word, is very short, and will bring out only a short fragment of the key-word.
The simpler case is readily explained. We have, say, a cryptogram beginning U S Z H L W D B P B G G F S. . . , in which we suspect the presence of the word SUPPLIES. We decipher the first eight letters, using this probable word as a trial key, and obtain a jumbled series of letters C Y K S A O Z J, which is not satisfactory.
| Figure 89 (a) Deciphering with the KEY: Key: B E D B E D B E D B E D....... CRYPTOGRAM: T I Q E W X Q T O J I V....... Plaintext: S E N D S U P P L I E S....... (b) Deciphering with the MESSAGE: Trial Key: S E N D S U P P L I E S....... CRYPTOGRAM: T I Q E W X Q T O J I V....... True Key: B E D B E D B E D B E D....... |
We leave off the first cryptogram-letter, U, and decipher the next eight, obtaining another jumbled series of letters A F S W L V X X. We start again at the third letter, then at the fourth letter, and still there is no information. But at the fifth trial, beginning at the fifth cryptogram-letter, we obtain a series T C O M E T C O, and this is satisfactory, not necessarily because we have recognized the word COMET, though this, of course, is a very desirable happening, but because the last three letters, T C O, are repeating the first three. The series is beginning over. The student should practice doing this, using both the tableau and the slide (or disk), until he is sure that he understands the process. The exact details of his work are immaterial; if he is sure that his key will be a recognizable word, it will be satisfactory to make decipherments directly on the cryptogram, erasing as he goes. Sometimes, however, the key is incoherent, or apparently so, and a jumbled series like C Y K S A O Z J might actually be the correct key; for this reason, it is well to follow a routine of some kind which will preserve all of the decipherments. One such plan is illustrated in Fig. 90.