26 Hi neks on Secret Writing. 



and the latter served to unravel whatever intricacy was attached 

 to the former. I believe that, with a different and unknown 

 key, it would be possible to decipher a long specimen of this 

 sort of writing, but certainly not one of a few sentences only, 

 or which concealed only the important words of the writing. 

 I must however own, that I think still greater secrecy would be 

 attained if, of the 81 characters, only about 40 were used to 

 express the letters of the alphabet, and the remaining half ap- 

 plied to the syllables and short words of most frequent recur- 

 rence. The cipher of 729 characters has certainly no need of 

 any additional contrivances to increase its secrecy, as with a 

 new key, even arranged on a similar plan to the present one, 

 which is by no means necessary, detection is entirely out of the 

 question. But wherefore use above 700 literal characters, 

 when 100 is sufficient to produce absolute secrecy ; and even 

 half that number, if syllabic. characters be interspersed? The 

 cipher is, beyond a doubt, sufficiently copious, without in the 

 smallest degree diminishing its secrecy, to express by separate 

 characters all the particles, pronouns, auxiliary verbs, and ter- 

 minations, as well those that belong to the grammar, as those 

 that make up the great body of the dictionary, and will have a 

 number yet to spare for the proper names, and other words, 

 that may be beforehand selected, as specially likely to occur in 

 the correspondence. But to employ a cipher, possessing such 

 means as this, in the manner Rlr. B. has done, to denote merely 

 the letters of the alphabet, appears to me a waste of time and 

 trouble, to as little purpose as if he had turned the force of a 

 steam-engine to draw a cork, or to crack a nut. What I have 

 objected to the dot writing of 81 characters, will apply with 

 apply with equal force to the figure-cipher which admits 100 

 characters, each composed of two figures. But, in his appli- 

 cation of this method, Mr. B. Jias adopted a very happy con- 

 trivance, the description of which leads me back to a subject I 

 promised to return to — the additional secrecy produced by the 

 employment of complex characters, above what the same num- 

 ber of simple ones would create. This additional secrecy con- 

 sists in this, tliat a person uuucquainted with the key can never 



