92 Mr. Chenevix on a new Method 



greater variety of values than appears to have been hitlierto 

 accomplished in any other system. 



The English alphabet is composed of twenty-six letters ; u, 

 V, w, andi, j, being considered as having distinct functions. 

 No cipher can be complete if each of the twenty-six letters is 

 not represented. But, if the value of the symbol never changes, 

 immediate detection, as every cipherer well knows, is inevitable. 

 Even should their value alter, and return to be the same, at the 

 end of a certain period or revolution, the objection is still the 

 same, with this modification merely, that it is diminished in pro- 

 portion to the length of the period. Thus the period of a cipher 

 for an alphabet of twenty-six letters, each having its variable 

 symbol, offers 26 loci, if so they may be termed; and the diffi- 

 culty of unravelling it, compared to the difficulty of unravelling 

 a cipher in which the value of the symbols is invariable, is, as it 

 •were, multiplied by 26. But even this security is not sufficient 

 in practice. 



To avoid the existence of a period, or so to lengthen the 

 revolution as to make it practically infinite, is the evident mode 

 to be followed, in order to accomplish secrecy ; and such is the 

 principle adopted in this cipher. For this purpose four addi- 

 tional symbols have been introduced, which in no manner com- 

 plicate either the theory or the practice, but which most amply 

 produce the desired effect. By their assistance the cipher is 

 composed of 26 original, and of four additional symbols ; conse- 

 quently the period of recurrent values must extend at least to 

 thirty ; but, such is the power of the four additional signs, aided 

 by other contrivances, that many new combinations interrupt 

 the period, even when but one single key, with all its variations, 

 is used ; and as 26 letters admit of a variety of permutations 

 which would be expressed by quadrillions, and have for type, all 

 the languages of the earth, it follows that the number of keys of 

 which this cipher is capable, would also be expressed by qua- 

 drillions ; and that these quadisillions, multiplied by the length 

 of the period of each key, with all the variable values of its sym- 

 bols, gives the number of loci contained in one entire period or 



