Hiticks on Seoi-et Writing. 21 



Art. III. On Secret Writing, in reply to Mr. Chenevix's 

 Challenge. Bi/ the Rev. Edward Hincks, A.M., and 

 formerly Felloiv of Trinity College, Dublin. 



The expedients which have been adopted by persons who 

 have occasion to carry on a correspondence necessary to be 

 concealed, and yet liable to be intercepted, have been very 

 numerous. More than one might be pointed out, by which ab- 

 solute inscrutability is attained with a very trifling expenditure 

 of trouble : but as our regularly-bred physicians refuse to ad- 

 minister the nostrum of a quack, even in diseases where its 

 healing efficacy is acknowledged by every one but themselves ; 

 so professional cipherers have always exhibited an unwillino-ness 

 to have secrecy obtabed by any other means than a literal 

 cipher. I employ this term to distinguish from sr/llabic and 

 verbal ciphers those in which the letters of the sentence to be 

 concealed are separately expressed by characters, either simple 

 or complex, in the cipher. I am net, of course, sufficiently ac- 

 quainted with the diplomacy of the present day, to be able to 

 state whether secretaries for foreign affairs, and their agents, 

 are so precise in their ideas as to decline to be secret, unless 

 they can be so secundum artem ; but if they be, they are cer- 

 tainly more scrupulous than their predecessors of the 17th 

 century were ; and yet I much doubt if they can better baffle 

 those whom violence or treachery might put in possession of 

 their despatches. 



My intention, in the present paper, is to expose the futility of 

 literal ciphers ; among the rest those of Mr. Blair, (Rees's 

 Cyclopmdia, Article Cipher,) and of Mr. Chenevix, [Journal of 

 Science, &c., No. XIX.) I do not mean to assert, that no literal 

 cipher can be contrived so as to be sufficiently secret ; but that 

 this cannot be effected without a key, so extensive and com- 

 plicated that it might be applied with equal facility to syllables, 

 as to letters. And if so, a syllabic cipher is unquestionably 

 preferable; inasmuch as, being equally secret, it is written and 

 read with about one-third part of the labour. 



Of literal ciphers thoco are two kinds ; one, in which the sig- 



