D. K. 27

418 CRYPTOGRAPHS AND CIPHERS [CH. XVIII
slips we have to try, and had the above message been three times as long, we could have solved the problem with half the trouble. The above example was not complicated by employing dummy letters or artificial alphabets: their use increases the difficulty of the decipherer, but if the message is a long one, the difficulties are not insuperable. Specialists, especially if working in combination, are said to select the right methods with almost uncanny quickness.
This chapter has already run to such a length that I cannot find space to describe more than one or two ciphers that appear in history.
It is said that Julius Caesar in making secret memoranda was accustomed to move every letter four places forward, writing d for a, e for h, &c. This would be a very easy instance of a cipher of the first type, but it may have been effective at that time. His nephew Augustus sometimes used a similar cipher, in which each letter was moved forward one place*.
Bacon proposed a cipher in which each letter was denoted by a group of five letters consisting of A and B only. Since there are 32 such groups, he had 6 symbols to spare, which he could use to separate words or to which he could assign special meanings. A message in this cipher would be five times as long as the original message. This may be compared with the far superior system of the five (or four) digit code- book system in use at the present time.
In the Morse code employed in telegraphy, as in the
Baconian system, only two signs are used, commonly a dot
or a short mark or a motion to the left, and a dash or a
long mark or a motion to the right. The Morse Alphabet is
as follows: a ( ), b ( • • ), c ( — ), d ( ), ^ (•)>
/( \9{ Xh(-'''\i{''),j( ),k( ),
l( ), ^( X^( )>o( \p( \q( ),
^( )> s('--),t{—), u{ ), v{ ), w ( ),x{ ),
* Of some of Caesar's correspondence, Suetonius says (cap. 56) si quis investigare et persequi velit, quartam elementorwn literani, id est, d pro a, et perinde reliquas commutet. And of Augustus he says (cap. 88) quoties autemper notas scribit, b pro a, c pro b, ac deinceps eadem ratione, sequentes literas ponit; pro X autem duplex a.
CH. XVIIl] CRYPTOGRAPHS AND CIPHERS 419
y ( ), z ( ). Since there are 30 possible per- mutations of two signs taken not more than four together,
this leaves four signals unemployed, ( ), ( ),
(• ), ( ), which might have been utilized for special
signals. In telegraphy there are also recognized signs or combinations for numerals, for the ends of words and messages, and for various calls between the sender and the recipient of a message.
Charles I used ciphers freely in important correspondence — the majority being of the second type. He was foolish enough to take a cabinet, containing many confidential notes in cipher, with him to the field of Naseby, where they fell into the hands of Fairfax*. In these papers each letter was repre- sented by a number. Clues were provided by the King who had written over the number the letter which it represented. Thus in two letters written in 1643, a is represented by 17 or 18, h by 13, c by 11 or 12, c^ by 5, g by 7 or 8 or 9 or 10, / by 15 or 16, g by 21, h by 31 or 32, i by 27 or 28, k by 25, I by 23 or 24, m by 42 or 44, n by 39 or 40 or 41, o by 35 or 36 or 37 or 38, p by 33 or 34, r by 50 or 51 or 52, 5 by 47 or 48, t by 45 or 46, u by 62 or 63, w by 58, and y by 74 or 77. Numbers of three digits were used to represent particular people or places. Thus 148 stood for France, 189 for the King, 260 for the Queen, 354 foi: Prince Rupert, and so on. Further, there were a few special symbols, thus ^1 stood for of, n\ for to, and /I for is. The numbers 2 to 4 and 65 to 72 were non- significant, and were to be struck out or neglected by the recipient of the message. Each symbol is separated from that which follows it by a full-stop.
A similar, though less elaborate, system was used by the French in the Peninsular War. An excellent illustration of the inherent defects of this method is to be found in the writings of the late Sir Charles Wheatstone. A paper in cipher, every page of which was initialed by Charles I, and countersigned by Lord Digby, was purchased some years
* First Report of the Royal Commisiion on Historical Mayiimcrijits, 1870, pp. 2, 4.
27—2
420 CR^PTOGRAPJdS AND CIPHERS [CH. XVllI
ago by the British Museum. It was believed to be a state paper of importance. It consists of a series of numbers with- out any clue to their meaning, or any indication of a division between words. The task of reading it was rendered the more difiScult by the supposition, which proved incorrect, that the document was in English ; but notwithstanding this, Sir Charles Wheatstone discovered the key*. In this cipher a was repre- sented by any of the numbers 12, 13, 14, 15, 16, or 17, 6 by 18, 19, and so on, while some 65 special words were represented by particular numbers: in all about 150 different symbols were used.
The famous diary of Samuel Pepys is commonly said to have been written in cipher, but in reality it is written in shorthand according to a system invented by T. Sheltonf. It is however somewhat difficult to read, for the vowels are usually omitted, and Pepys used some arbitrary signs for terminations, particles, and certain words — so far turning it into a cipher. Further, in certain places, where the matter is such that it can hardly be expressed with decency, he changed from English to a foreign language, or inserted non-significant letters. Shelton's system had been forgotten when attention was first attracted to the diary. Accordingly we may say that, to those who first tried to read it, it was written in cipher, but Pepy's contemporaries would have properly described it as being written in shorthand, though with a few modifications of his own invention.
A system of shorthand specially invented for the purpose is a true cipher. Such a system in which the letters were represented by four strokes varying in length and position was employed by Charles I. Another such system in which each letter is represented either by a dot or by a line of constant length was used by the Earl of Glamorgan, better known by his subsequent title of Marquis of Worcester, in 1645; each of these
* The document, its translation, and the key used, are given iu Wheatstone 's Scientific Papers, London, 1879, pp. 321—341.
t Tachy-graphy, by T. Shelton. The earliest edition I have seen is dated 1641. A somewhat similar system by W. Cartwright was issued by J. Eich under the title Semographie, London, 1644.
CH. XVIIl] CRYPTOGRAPHS AND CIPHERS 421
was a cipher of the first type arirl iiad the defects inherent in almost every cipher of this kind: in fact Glamorgan's letter was deciphered, and the system was discovered by H. Dircks*. Obsolete systems of shorthand f may be thus used as ciphers.
It is always difficult to read a very short message in cipher, since necessarily the clues are few in number. When the Chevalier de Rohan was sent to the Bastille, on suspicion of treason, there was no evidence against him except what might be extracted from Monsieur Latruaumont. The latter died without making any admission. De Rohan's friends had ar- ranged with him to communicate the result of Latruaumont's examination, and accordingly in sending him some fresh body linen they wrote on one of the shirts Mg dulhxcclgu ghj yontj, Im ct idgc alj. For twenty-four hours de Rohan pored over the message, but, failing to read it, he admitted his guilt, and was executed November 27, 1674 The cipher is a simple one of the first type, but the communication is so short that unless the key were known it would not be easy to read it. Had de Rohan suspected that the second word was pvisonnier, it would have given him 7 out of the 12 letters used, and as the first and third words suggest the symbols used for I and t, he could hardly have failed to read the message.
Marie Antoinette used what was in effect a St Cyr cipher, consisting of 11 substitution alphabets employed in succession. The first alphabet was n, o, p, z, a, h, I, m\ the next,
0, p, q, m, n; the next /j, q, r, n, o; and so on. An
expert would easily read a message in this cipher.
One of the systems in use to-day is the five digit code-