On the Jrt of conveying Secret and Swift Inielligenct.' "' 32 j 



mediately felefted ; the ferie&of feveh pages, which one of thefe rulers opens, is cut, liice 

 the alphabet of a ledger, at the edge in feven divifions. By thefe means the page con- 

 taining the hundred which is wanted, is inltantly found.. In the page thus found the 

 tens, from ten to feventy inclufive, are divided from each other fo as to be inliantly.dif- 

 tinguifhable, and the units under each diviiion are in like manner eauly felefled. 



Plate III. is-a fpecimen of the firft page of the vocabulary ;. and though it is but one- 

 fifth of the real fize, it is fufficiently diitinft. It^ is divided into eight clafTes f ail the: 

 dalles are numbered downwards feriatim from i to 77, omitting cyphers or zero, and 

 eights and nines. When once the clafs required is afcertained, any number on the page 

 can be found immediately :— As for inftance, the reader will cafily fcle£t clafs iv. . 

 number 36^, or clafs vii. number 77, and fo of the refL 



Nothing remains to be explained but the manner in which the clafs in each page Is 

 pointed out by the machinery. For this purpofe, before the pointers are turned to any fet 

 of figures, the polfiter that reprefents thoufands is turned to the claf, that is wanted ; as 

 foon as the correfpondent anfwers this fignal, thousands is returned to O, and inftantly 

 all the pointers are moved to the places which denote the figures required for any word , 

 or fentence. 



When the cl&fs is thus afcertained, an index, which fiides on' the mairogany covcp of 

 the book, is fet to the column belonging to this clafs; the nunbr of thoufands is then: 

 opened by the ruler, as foon as it is read off by the teiefcope ; the number of hundreds 

 IS opened by the pages when they are cut away, and the number of tens and units is feen. 

 on the page. As the pointers are moved in fucceffion from-thoufands to. unito, the dif- 

 ferent divifions of the book- can be opened as fait as the pointers are moved.. The order 

 of this book might be reverfed with apparent advantage, by dividing the bsok-into clafles 

 by the mahogany rulers, &c. Uut I prefer, for reafons which- it would bs tedious to infill 

 upon, the arrangement which I have followed. 



As fecrefy is an obje6l of the greatell confequence, I (half endeavour to point out in a 

 few words the fuperiority of this mode of communication over any alphabetical arrange- 

 ment, not only in point of expedition, but of concealment. 



Although the common alphabet-may be varied at pleafure,' and any arbitrary figns.may 

 be employed to convey the powers of each letter, yet, by-certain rules, any, of thefe arrange- 

 ments may be deciphered. Whoever fees the movements of the French telegraph. (I'mean 

 of that which is commonly known as fuch) may unfold the intelligence which it, conveys, 

 by merely marking down the changes which he fees, and putting them into. the biiads of 

 a decipherer. The rules for deciphering depend upon the ufual arrangements bf letters. 

 In our language a fingle letter my ft be A or I. The proportions which exift between 

 words of one, two, thrccj and any greater number of letters, are clafled in- catalogues; 

 and from thefe the monofyllables of any cipher are eafily obtained, and from the. letters 

 of thefe' menofyllables the letters of longer vfords are difcovered. By fimilar rules, fomc 

 of which are very ingenious, and which depend upon the general philofophy of language, 

 any alphabetical cipher may be e;ifiiy unfolded. But thefe rules, except a very fewof them^ 

 are ufelefs, when we employ ciphers which denote entire words. Here the moll obvious 

 means of difcovety may be avoided, by omitting thofe common words which occur fo 

 frequ'titly in every language, the, and-, that, to, &g. But,, fuppofing that from its. fre^ 

 aijent.iccurrepce any particular word ihould be difcovered, no progrcfs can be made from 



tkefft. 



