and if the cipher is Vigenère, the key-letters which produce these deciphered columns will have to run backward in the alphabet. These can be added at the tops or bottoms of their columns, and can, if desired, be written in red ink, or otherwise distinguished.
Fig. 113 shows what modifications would be necessary if the sheets were being prepared for one of the Beauforts. For the variant Beaufort, the only difference lies in the fact that key-letters must progress in the same alphabetical direction as their decipherments. With the true Beaufort, however, the making of an A-decipherment does not mean a simple copying of cryptogram letters, as in the other two ciphers; this A-decipherment must first be made; after that, the series of normal alphabets can be extended as before, and the key-letters will progress in the same alphabetical direction as their deciphered columns.
Now, assuming that these sheets have actually been prepared, say on quadrille paper, the various columns of decipherment may be examined, and a check-mark placed beside each column in which the series of letters appears to represent a “good” decipherment. With longer columns, those may be checked which contain the largest percentages of letters E T A O N I R S H, without too many of the letters J K Q X Z; with shorter columns, perhaps those are “best” in which any repeated letters are chiefly vowels, it being remembered that when the cryptogram contains repeated sequences, as well as repeated single letters, the possible identity of these repeated digrams or trigrams must also be taken into consideration. With all of the apparently good columns checked for attention, sheet 1 may be creased vertically so as to place any desired column on the extreme right, and this column may then be laid directly against any desired column of sheet 2 for an observation of the resulting digrams. If these appear to be satisfactory, then sheet 2 may also be creased vertically, and the series of apparently good digrams may be laid directly against any desired column of sheet 3 for an observation of the resulting trigrams. And so on, if desired, to a possible sheet 4, or 5, or 6, though, as a rule, the first three sheets will be found sufficient. While the method, as indicated, is intended to be mechanical, that is, largely visual, it would be possible, where uncertainty exists between two given combinations, to copy these and subject them to a digram test. But this should not be necessary in a case where key-letters, as well as their
| Figure 113 If column Z U Z O E were VARIANT: If column Z U Z O E were BEAUFORT: KEYS: a b c d e f g ..... KEYS: a b c d e f g ..... Z A B C D E F ..... Z - B C D E F G H ..... U V W X Y Z A ..... U - G H I J K L M ..... Z A B C D E F ..... Z - B C D E F G H ..... O P Q R S T U ..... O - M N O P Q R S ..... E F G H I J K ..... E - W X Y Z A B C ..... |
deciphered columns, are expected to set up good combinations in order to form a plaintext key-word.
An interesting version of this method, as shown by Admiral Elliott Snow, included the following variations: To begin with, in extending the alphabets, the decryptor omits altogether the letters J K Q X Z, and perhaps one or two others of extremely low frequency, simply leaving the blank spaces which indicate their alphabetical positions. This makes the work more rapid, and, in addition, the presence of these blank spaces in any column of decipherment, advertises at once that the column is probably not a very good one. But Admiral Snow’s columns were not columns; they were rows. A given series of letters: as Z U Z O E of our foregoing sheet 1, is laid out horizontally, and its decipherments are extended vertically. The spacing on each row is arranged to correspond with the period; that is, the letters Z U Z O E, instead of being continuous, are spaced six columns apart if the period is 6, and their decipherments, of course, are spaced in the same way. The sheets may now be creased horizontally between rows, and one sheet placed against another in such a way that the resulting digrams are all standing on diagonals, but have appeared at exactly their cryptogram distance apart. The student should experiment with both arrangements and decide which one he likes.
It has been pointed out by C. A. Castle, another of our members, that the foregoing method will find its chief application, not on a single cryptogram, but as applied to a case which, so far, we have not considered in connection with the substitution ciphers: One in which the decryptor has in his possession five or six cryptograms, all very brief, but all enciphered with the same key. Here, we have the common practical case, to be handled in somewhat the same way as the last of our transposition examples; the cryptograms can be written one below another, thus forming a series of columns in which every column has been enciphered with the same cipher alphabet. If this case happens to involve a comparatively short period, it is possible to take intervals between repeated sequences found in two different cryptograms, using the intervals indicated by the number of columns between the first letter of one sequence and the first letter of its repetition. Castle’s example, however, was not based on a short key, but upon an extremely long one, and his five or six messages were merely fragments, each one of which was known to be the beginning of an English sentence. In the English language, about half of all initial letters used are found in the group T A O S H I and more than another one-fourth are found in the group W C B P F D M. Thus, having a series of beginnings in which the first column will include only initial letters, the number of truly acceptable decipherments on any sheet 1 will usually be quite limited. In addition, with vowels known to have a fondness for second and third positions in words, there should be little difficulty in selecting decipherments from sheets 2 and 3.
While we have described this device as having been written out on sheets of paper, there are many persons who prefer to have at hand a series of cardboard strips which will set up the “sheets” mechanically. If each of the strips carries
| Figure 114 One Form of "DECRYPTING SLIDE" C. STANLEY LAMB For VIGENÈRE, the "Decrypting Alphabet" runs backward: | a z y x w v u t s r q p o n m l k j i h g f e d c b | ( 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 ... ( 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 ... ( 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 ... | 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 | As this is shown, it has been set for the decipherment of a trigram H D G, and every possible decipherment can be read from the slide without changing its present adjustment. The entire list of frequent trigrams can be used as trial keys: Trial Keys: T H E A N D T H A E N T I O N T I O F O R (Etc.) H D G Fragment of True Key: O W C H Q D O W G D Q N Z P T O V S C P P (Etc.) |
the normal alphabet written twice in succession, it is possible to adjust five of the strips so as to place the letters Z U Z O E one below another in the form of a column and automatically set up the other 25 columns. The strips can be loose, or may form part of a slide. Slides, in fact, may be used for many purposes, and are well worth preparing for any kind of cipher which the decryptor expects to encounter a great many times. The members of the American Cryptogram Association, who solve a great many Vigenères, Beauforts, and so on, as a matter of recreation, have practically all “invented” slides (or tableaux) which will, to some extent, do away with the irksome task of carrying out a trigram-search. These are prepared in various ways, and variously used, though the principle for all is about the same as that indicated in Fig. 114. They are usually referred to as decrypting slides, and the single stationary alphabet, sometimes a list of key-letters and sometimes not, will be called “the decrypting alphabet.” C. Stanley Lamb, who is by no means the only “inventor” of the device illustrated, has this in several different forms, according to the purpose for which he intends to use it. Notice that the card, as we have placed it, shows the stationary single alphabet running contrary to the others, for use on the Vigenère cipher, and that this card need merely be reversed in order to have a single stationary alphabet running parallel to the others, for use on the two Beauforts. As to the sliding double alphabets, there may be as many of these as the operator feels like setting up; if the device is being used to assist in the trigram search, three will be needed.