icy," &c. You write with twelve letters only, as in the subjoined table, in which however I have not followed his ways of diversifying.

In this table, z1, for instance, represents 1st M, and 2nd R or S; to distinguish whether R or S, he has recourse to a supplementary contrivance by nulls.

f. 242. v. He refers to table at 200., and says the three letters abc, (which there represent I) may be replaced by a single character; for this table represents in another column letters by dots. Thus T is.. .; D . . . ; or if we will we may put o's for dots; so that D=0 0 0 and T=00 000 o; and the spaces may be filled up by a slightly varied o. Thus D=00000, T=00000000, and thus the whole cipher will apparently consist of o's.

.....

The transition from this to Bacon's cipher is so easy that the credit given to him must be reduced.

LIST of WORKS in GENERAL

PUBLISHED BY

LITERATURE,

Messrs. LONGMAN, BROWN, GREEN, LONGMANS, and ROBERTS, 39, PATERNOSTER ROW, LONDON.

alphabets are given by Porta, and therefore the secret communication can consist only of sixty letters. It is worth remarking that when Porta wrote it was usual to put the sign of the cross at the head of an ordinary epistle. The first of his alphabets corresponds not to a series of words but to two and twenty different modifications of the figure of a cross, and his second alphabet similarly corresponds to two and twenty different modifications of the introductory flourish. His sixtieth alphabet is of the same kind. We see here perhaps whence Bacon derived his idea of giving significance to seemingly accidental modifications of the characters of ordinary writing.

The idea of a biliteral alphabet, which Bacon seems to claim as his own, is employed, though in a different manner, by Porta. His method is in effect this. He reduces the alphabet to sixteen letters, and then takes the eight different arrangements aaa, aba, &c., to represent them; each arrangement representing two letters indifferently the ambiguity arising from hence he seems to disregard. In this manner he reduces any given word or sentence to a succession of a's and b's. At this point his method, of which he has given several modifications, departs wholly from Bacon's. Let us suppose the biliteral series to commence with aababb. A word of two syllables and beginning with a indicates that two a's commence the series; any monosyllable will serve to show that one b follows, another that it is succeeded by one a, and then any dissyllable will stand for bb. Thus Amo te mi fili or Amat qui non sapit will represent the biliteral arrangement aababb, and so on on a larger scale. Porta's method is therefore not, like Bacon's, a method scribendi omnia per omnia, but only omnia per multa. Still the analogy of the two methods is to be remarked: both aim at concealing that there is any but the obvious meaning, and both depend essentially on rcpresenting all letters by combinations of two only. See the De oc. Lit. Signis. v. c. 3.

The Polygraphia of Trithemius (dedicated to Maximilian in 15081) consists of six books. The first four contain extensive tables constituting four different ciphræ verborum; the first and second of which are significant, and relate, the former to the second person of the Trinity, and the latter to the Blessed Virgin. The fifth and sixth books are of less importance. Trithemius, written in the cipher of the second book, becomes "Charitatem pudicissimæ Virginis Mariæ productricis coexistentis verbi, robustissimi commilitonis mei dilectissimi devotissime benedicamus; vivificatrix omnium," &c.

The edition of 1600 is that I use.

Traicté des Chiffres, ou secrètes manières d'escrire, par Blaise de Vigenère, Bourbonnois. (Paris, 1587.)

This work is described by the author as what he had saved of his work "Du Secrétaire," written in Italy in 1567 and 68. The two first books were stolen at Turin in 1569. The third is the foundation of the present work. (v. f. 285. verso.) He says he had revealed nothing of its contents.

The two authors whom he chiefly mentions are Trithemius and Porta; that is, modern authors; for there is a great deal said of the Cabala. The key ciphers of which Porta speaks he ascribes to a certain Belasio, who employed it as early as 1549: Porta's book not being published until 1563, "auquel il a inseré ce chiffre sans faire mention dont il le tenoit." Porta's book, he goes on to say, was not en vente until 1568. The invention was ascribed to Belasio by the grand vicar of St. Peter at Rome, who had great skill in deciphering. (f. 35. rect. and 37. verso.)

At f. 199. Vigenère gives an account of ciphers in which letters are represented by combinations of other letters,-which Porta had already done, but which he varies in a number of ways.

f. 200. A table where the twenty-three letters of the alphabet, and four other characters are represented by combinations of abc. D (e. gr.)=aaa, S=bac, &c.)

f. 201. A smaller table where an alphabet of twenty-one letters is similarly represented.

f. 202. An alphabet of twenty letters represented by binary combinations of five letters, a=ED, &c.

f. 202. goes on to what Bacon speaks of, a cipher within a cipher. You write in a common cipher with an alphabet of eighteen letters; the cipher being such that the five vowels are used as nulls; then by the last cipher these five vowels are made significant, and give the hidden sense. He seems to speak of this as his own.

After mentioning a cipher described by Cardan, he goes on, f. 205. to Porta's ciphers by transposition, &c.

At f. 240. he shows how characters may be multiplied by different ways of writing them; which Porta had not done.

f. 241. An alphabet and &, each character written in four ways. f. 241. verso, An application of these variations.

f. 242. He remarks that a great variety of uses may be made of this idea, and gives some.

f. 244. He goes on "De ce même retranchement et de la variété de figure, part une autre invention encore d'un chiffre carré à double entente, le plus exquis de tous ceux qui ayent esté decouvers jusqu'à

icy," &c. You write with twelve letters only, as in the subjoined table, in which however I have not followed his ways of diversifying.

In this table, z1, for instance, represents 1st M, and 2nd R or S; to distinguish whether R or S, he has recourse to a supplementary contrivance by nulls.

f. 242. v. He refers to table at 200., and says the three letters abc, (which there represent I) may be replaced by a single character; for this table represents in another column letters by dots. Thus Tis.. ... .; D. • ; or if we will we may put o's for dots; so that D=0 0 0 and T=00 000 o; and the spaces may be filled up by a slightly varied o. Thus D=00000, T=00000000, and thus the whole cipher will apparently consist of o's.

The transition from this to Bacon's cipher is so easy that the credit given to him must be reduced.