1911 Encyclopædia Britannica/Cryptography
CRYPTOGRAPHY (from Gr. κρύπτος, hidden, and γράφειν, to write), or writing in cipher, called also steganography (from Gr. στεγάνη, a covering), the art of writing in such a way as to be incomprehensible except to those who possess the key to the system employed. The unravelling of the writing is called deciphering. Cryptography having become a distinct art, Bacon (Lord Verulam) classed it (under the name ciphers) as a part of grammar. Secret modes of communication have been in use from the earliest times. The Lacedemonians had a method called the scytale, from the staff (σκυτάλη) employed in constructing and deciphering the message. When the Spartan ephors wished to forward their orders to their commanders abroad, they wound slantwise a narrow strip of parchment upon the σκυτάλη so that the edges met close together, and the message was then added in such a way that the centre of the line of writing was on the edges of the parchment. When unwound the scroll consisted of broken letters; and in that condition it was despatched to its destination, the general to whose hands it came deciphering it by means of a σκυτάλη exactly corresponding to that used by the ephors. Polybius has enumerated other methods of cryptography.
The art was in use also amongst the Romans. Upon the revival of letters methods of secret correspondence were introduced into private business, diplomacy, plots, &c.; and as the study of this art has always presented attractions to the ingenious, a curious body of literature has been the result.
John Trithemius (d. 1516), the abbot of Spanheim, was the first important writer on cryptography. His Polygraphia, published in 1518, has passed through many editions, and has supplied the basis upon which subsequent writers have worked. It was begun at the desire of the duke of Bavaria; but Trithemius did not at first intend to publish it, on the ground that it would be injurious to public interests. A Steganographia published at Lyons (? 1551) and later at Frankfort (1606), is also attributed to him. The next treatises of importance were those of Giovanni Battista della Porta, the Neapolitan mathematician, who wrote De furtivis litterarum notis, 1563; and of Blaise de Vigenere, whose Traité des chiffres appeared in Paris, 1587. Bacon proposed an ingenious system of cryptography on the plan of what is called the double cipher; but while thus lending to the art the influence of his great name, he gave an intimation as to the general opinion formed of it and as to the classes of men who used it. For when prosecuting the earl of Somerset in the matter of the poisoning of Overbury, he urged it as an aggravation of the crime that the earl and Overbury “had cyphers and jargons for the king and queen and all the great men,—things seldom used but either by princes and their ambassadors and ministers, or by such as work or practise against or, at least, upon princes.”
Other eminent Englishmen were afterwards connected with the art. John Wilkins, subsequently bishop of Chester, published in 1641 an anonymous treatise entitled Mercury, or The Secret and Swift Messenger,—a small but comprehensive work on the subject, and a timely gift to the diplomatists and leaders of the Civil War. The deciphering of many of the royalist papers of that period, such as the letters that fell into the hands of the parliament at the battle of Naseby, has by Henry Stubbe been charged on the celebrated mathematician Dr John Wallis (Athen. Oxon. iii. 1072), whose connexion with the subject of cipher-writing is referred to by himself in the Oxford edition of his mathematical works, 1689, p. 659; as also by John Davys. Dr Wallis elsewhere states that this art, formerly scarcely known to any but the secretaries of princes, &c., had grown very common and familiar during the civil commotions, “so that now there is scarce a person of quality but is more or less acquainted with it, and doth, as there is occasion, make use of it.” Subsequent writers on the subject are John Falconer (Cryptomenysis patefacta), 1685; John Davys (An Essay on the Art of Decyphering: in which is inserted a Discourse of Dr Wallis), 1737; Philip Thicknesse (A Treatise on the Art of Decyphering and of Writing in Cypher), 1772; William Blair (the writer of the comprehensive article “Cipher” in Rees’s Cyclopaedia), 1819; and G. von Marten (Cours diplomatique), 1801 (a fourth edition of which appeared in 1851). Perhaps the best modern work on this subject is the Kryptographik of J. L. Klüber (Tübingen, 1809), who was drawn into the investigation by inclination and official circumstances. In this work the different methods of cryptography are classified. Amongst others of lesser merit who have treated of this art may be named Gustavus Selenus (i.e. Augustus, duke of Brunswick), 1624; Cospi, translated by Niceron in 1641; the marquis of Worchester, 1659; Kircher, 1663; Schott, 1665; Ludwig Heinrich Hiller, 1682; Comiers; 1690; Baring, 1737; Conrad, 1739, &c. See also a paper on Elizabethan Cipher-books by A. J. Butler in the Bibliographical Society’s Transactions, London, 1901.
Schemes of cryptography are endless in their variety. Bacon lays down the following as the “virtues” to be looked for in them:—“that they be not laborious to write and read; that they be impossible to decipher; and, in some cases, that they be without suspicion.” These principles are more or less disregarded by all the modes that have been advanced, including that of Bacon himself, which has been unduly extolled by his admirers as “one of the most ingenious methods of writing in cypher, and the most difficult to be decyphered, of any yet contrived” (Thicknesse, p. 13).
The simplest and commonest of all the ciphers is that in which the writer selects in place of the proper letters certain other letters in regular advance. This method of transposition was used by Julius Caesar. He, “per quartam elementorum literam,” wrote d for a, e for b, and so on. There are instances of this arrangement in the Jewish rabbis, and even in the sacred writers. An illustration of it occurs in Jeremiah (xxv. 26), where the prophet, to conceal the meaning of his prediction from all but the initiated, writes Sheshak instead of Babel (Babylon), the place meant; i.e. in place of using the second and twelfth letters of the Hebrew alphabet (b, b, l) from the beginning, he wrote the second and twelfth (sh, sh, k) from the end. To this kind of cipher-writing Buxtorf gives the name Athbash (from a the first letter of the Hebrew alphabet, and th the last; b the second from the beginning, and h the second from the end). Another Jewish cabalism of like nature was called Albam; of which an example is in Isaiah vii. 6, where Tabeal is written for Remaliah. In its adaptation to English this method of transposition, of which there are many modifications, is comparatively easy to decipher. A rough key may be derived from an examination of the respective quantities of letters in a type-founder’s bill, or a printer’s “case.” The decipherer’s first business is to classify the letters of the secret message in the order of their frequency. The letter that occurs oftenest is e; and the next in order of frequency is t. The following groups come after these, separated from each other by degrees of decreasing recurrence:—a, o, n, i; r, s, h; d, l; c, w, u, m; f, y, g, p, b; v, k; x, q, j, z. All the single letters must be a, I or O. Letters occurring together are ee, oo, ff, ll, ss, &c. The commonest words of two letters are (roughly arranged in the order of their frequency) of, to, in, it, is, be, he, by, or, as, at, an, so, &c. The commonest words of three letters are the and and (in great excess), for, are, but, all, not, &c.; and of four letters—that, with, from, have, this, they, &c. Familiarity with the composition of the language will suggest numerous other points that are of value to the decipherer. He may obtain other hints from Poe’s tale called The Gold Bug. As to messages in the continental languages constructed upon this system of transposition, rules for deciphering may be derived from Breithaupt’s Ars decifratoria (1737), and other treatises.
Bacon remarks that though ciphers were commonly in letters and alphabets yet they might be in words. Upon this basis codes have been constructed, classified words taken from dictionaries being made to represent complete ideas. In recent years such codes have been adapted by merchants and others to communications by telegraph, and have served the purpose not only of keeping business affairs private, but also of reducing the excessive cost of telegraphic messages to distant markets. Obviously this class of ciphers presents greater difficulties to the skill of the decipherer.
Figures and other characters have been also used as letters; and with them ranges of numerals have been combined as the representatives of syllables, parts of words, words themselves, and complete phrases. Under this head must be placed the despatches of Giovanni Michael, the Venetian ambassador to England in the reign of Queen Mary, documents which have only of late years been deciphered. Many of the private letters and papers from the pen of Charles I. and his queen, who were adepts in the use of ciphers, are of the same description. One of that monarch’s letters, a document of considerable interest, consisting entirely of numerals purposely complicated, was in 1858 deciphered by Professor Wheatstone, the inventor of the ingenious crypto-machine, and printed by the Philobiblon Society. Other letters of the like character have been published in the First Report of the Royal Commission on Historical Manuscripts (1870). In the second and subsequent reports of the same commission several keys to ciphers have been catalogued, which seem to refer themselves to the methods of cryptography under notice. In this connexion also should be mentioned the “characters,” which the diarist Pepys drew up when clerk to Sir George Downing and secretary to the earl of Sandwich and to the admiralty, and which are frequently mentioned in his journal. Pepys describes one of them as “a great large character,” over which he spent much time, but which was at length finished, 25th April 1660; “it being,” says he, “very handsomely done and a very good one in itself, but that not truly alphabetical.”
Shorthand marks and other arbitrary characters have also been largely imported into cryptographic systems to represent both letters and words, but more commonly the latter. This plan is said to have been first put into use by the old Roman poet Ennius. It formed the basis of the method of Cicero’s freedman, Tiro, who seems to have systematized the labours of his predecessors. A large quantity of these characters have been engraved in Gruter’s Inscriptiones. The correspondence of Charlemagne was in part made up of marks of this nature. In Rees’s Cyclopaedia specimens were engraved of the cipher used by Cardinal Wolsey at the court of Vienna in 1524, of that used by Sir Thomas Smith at Paris in 1563, and of that of Sir Edward Stafford in 1586; in all of which arbitrary marks are introduced. The first English system of shorthand—Bright’s Characterie, 1588—almost belongs to the same category of ciphers. A favourite system of Charles I., used by him during the year 1646, was one made up of an alphabet of twenty-four letters, which were represented by four simple strokes varied in length, slope and position. This alphabet is engraved in Clive’s Linear System of Shorthand (1830), having been found amongst the royal manuscripts in the British Museum. An interest attaches to this cipher from the fact that it was employed in the well-known letter addressed by the king to the earl of Glamorgan, in which the former made concessions to the Roman Catholics of Ireland.
Complications have been introduced into ciphers by the employment of “dummy” letters,—“nulls and insignificants,” as Bacon terms them. Other devices have been introduced to perplex the decipherer, such as spelling words backwards, making false divisions between words, &c. The greatest security against the decipherer has been found in the use of elaborate tables of letters, arranged in the form of the multiplication table, the message being constructed by the aid of preconcerted key-words. Details of the working of these ciphers may be found in the treatises named in this article. The deciphering of them is one of the most difficult of tasks. A method of this kind is explained in the Latin and English lives of Dr John Barwick, whose correspondence with Hyde, afterwards earl of Clarendon, was carried on in cryptography. In a letter dated 20th February 1659/60, Hyde, alluding to the skill of his political opponents in deciphering, says that “nobody needs to fear them, if they write carefully in good cyphers.” In his next he allays his correspondent’s apprehensiveness as to the deciphering of their letters.
“I confess to you, as I am sure no copy could be gotten of any of my cyphers from hence, so I did not think it probable that they could be got on your side the water. But I was as confident, till you tell me you believe it, that the devil himself cannot decypher a letter that is well written, or find that 100 stands for Sir H. Vane. I have heard of many of the pretenders to that skill, and have spoken with some of them, but have found them all to be mountebanks; nor did I ever hear that more of the King’s letters that were found at Naseby, than those which they found decyphered, or found the cyphers in which they were writ, were decyphered. And I very well remember that in the volume they published there was much left in cypher which could not be understood, and which I believe they would have explained if it had been in their power.”
An excellent modification of the key-word principle was constructed by Admiral Sir Francis Beaufort.
Ciphers have been constructed on the principle of altering the places of the letters without changing their powers. The message is first written Chinese-wise, upward and downward, and the letters are then combined in given rows from left to right. In the celebrated cipher used by the earl of Argyll when plotting against James II., he altered the positions of the words. Sentences of an indifferent nature were constructed, but the real meaning of the message was to be gathered from words, placed at certain intervals. This method, which is connected with the name of Cardan, is sometimes called the trellis or cardboard cipher.
The wheel-cipher, which is an Italian invention, the string-cipher, the circle-cipher and many others are fully explained, with the necessary diagrams, in the authorities named above—more particularly by Klüber in his Kryptographik. (J. E. B.)