73 



ing points of marked similarity. Nay, more, it seems that we 

 might thus ascertain whether any given writing, of which 

 there existed considerable remains, were phonetic or idea- 

 graphic. In the former case, it is evident that the distribution 

 of characters in any two passages of equal and considerable 

 length would be very similar ; in the latter, we could not 

 count upon so great uniformity, seeing that the connexion of 

 ideas is so much less regular than the sequence of sounds. 



In proof of the actual efficiency of his method Mr. Graves 

 exhibited two tables, one of which was founded upon an analy- 

 sis of a small number of Runic inscriptions, the other being 

 made from a short passage of Icelandic of the thirteenth century. 

 A cursory inspection of these tables would be sufficient to 

 enable a decipherer to assign their proper powers to the 

 Runic letters, supposing that their values had not been other- 

 wise known. Mr. Graves also exhibited tables formed from 

 the analysis of passages in the Irish language, contained in 

 the Book of Armagh, and written, as he believes, according 

 to the orthography of the seventh century. These were the 

 tables employed by him in determining the powers of the 

 Ogham characters ; but he reserved the statement of the re- 

 sults arrived at in that research for a communication which he 

 hoped very soon to make to the Academy. 



VOL. IV. 



