94 Mr. Chonevix on a new Method 



is to say, detection is reduced to a mere guess. Th« expression 

 (m — /)" may be considered as a limit ; and a cipher which at- 

 tains it does all that can be done, and much more than is indis- 

 pensable for absolute secrecy. 



The ease with which this cipher may be used by those who 

 have the key, is the same as in all ciphers where a substitute is 

 employed for the true letter, and greater than in all ciphers 

 where that substitute is complicated. The most natural and 

 evident symbols which occur, are the letters of the alphabet, 

 with new values. They are so familiar to all persons as to be 

 executed without any perception of effort ; whereas, if new sym- 

 bols must be learned, innumerable inconveniencies may arise. 

 This, indeed, is a defect in the most ingenious and satisfactory 

 ciphers known, and particularly in one which is to be found in 

 JRees* Cydopcedia, Art. Cipher ; and which seems to possess the 

 requisite of secrecy, in a very eminent degree. The author pro- 

 poses the use of dots, or of lines, disposed according to a certain 

 law, above, upon, or below a horizontal line, reaching from left 

 to right of the page. But a cipherer is much more exposed to 

 commit mistakes in giving the due position to a dot, or the proper 

 length to a line, than in writing down a letter to which he has 

 long been habituated ; and indeed, upon the whole, no system 

 of symbols seems to unite so many advantages as the letters of 

 our common alphabet, to which new values are assigned. 



The author of the article just alluded to has, upon some occa- 

 sions, adopted arithmetical figures. These, though preferable 

 to dots and lines, are not so convenient as letters; for in their 

 single state they have but ten varieties of forms, and beyond that 

 number, two units must be used. Now this requires not only 

 double time and labour, but a double effort of attention, and 

 doubles the chances of committing mistakes, since the specifi- 

 cation of each letter depends upon the united power of two 

 symbols. 



It seems to be a radical fault in the cipher of that author, that 

 to express a single letter, he employs two symbols, and often 

 three ; or at least symbols composed of two or three parts, each 

 of which requires a separate operation both of time and of atten- 



