361 



divided them into two classes, broad and slender. The broad, 

 a, o, u, are put first; the slender, e, i, last. At whatever time 

 this distinction had its rise, it was not by any means strictly ob- 

 served by the earliest writers of this country. Frequent vio- 

 lations of it are to be found in the orthography of the Irish 

 passages in the Book of Armagh/ and of the names which 

 occur in the most ancient inscriptions. 



There is scarcely any particular in the foregoing account 

 of the Ogham alphabet which does not indicate a connexion 

 between it and the Runic alphabets, especially the later and 

 more developed ones, such as were used by the Anglo- 

 Saxons, and were constructed by persons acquainted with 

 the Roman letters. 



The most ancient Runic alphabet was commonly divided 

 into three groups of letters (citter) ; thus, f, u, th, o, r, k — h, 

 n, i, a, s — f, b, I, m, o ; and there existed an almost infinite 

 variety of cryptic alphabets, all founded upon this one princi- 

 ple, that the symbol for any letter indicated, in the first in- 

 stance, to which of these three groups it belonged, and, in the 

 next, the place which it held in that group. 



No better instance can be given than the following alpha- 

 bet, described by Liljegren in his Otimlcira, p. 50 : 



+ += -^ ^ -fc -1 4- =+ ^ ^ ^ 4- ^ + ^ ^ 



Here we see not only an exemplification of the principle on 

 which the Ogham alphabet is constructed, but even a develop- 

 ment of it in a form very nearly the same as that of the 

 Ogham itself. Goransson, in his Bautil, p. 232, gives a figure 

 of an ancient monument, on which occur a few words written 

 in these Ogham-like Runes, the remaining part of the in- 

 scription being in Runes of the common form. 



Other Runic alphabets were formed by repeating the initial 

 letter of each group a different number of times, to denote each 



