Rhys — Irish Ogham Inscriptions. 301 



tion on the former is written in characters which may well be two 

 centuries older than those on the latter. According to the last descrip- 

 tion I have had from Dr. Ferguson of the former, I would read it 

 Uvanos Avi Evacattos. 



You expressed a wish to know to what language I think the 

 Ogams of Ireland belong. I do not recollect seeing a single instance 

 mentioned of the earlier class which struck me as being anything but 

 Early Irish ; and were there a complete vocabulary of the proper names 

 which occur in Irish MSS. and literature generally, I fancy that a 

 philologist would without difficulty identify 99 out of every 100 names 

 to be met with on Ogam-inscribed stones in Ireland. Some few names 

 in the latest specimens may betray Latin influence. 



And, as regards the suggested cryptic character of the inscriptions, 

 neither do I see any reasons to believe that the Ogam alphabet was in- 

 tended for cryptic purposes. It is possible, however, that it may have, 

 in the hands of pedants, been so applied just as it was growing obsolete. 

 I hope I shall not be giving offence in protesting against the theory that 

 the names occurring in Ogam were not really the names of the persons 

 intended, but were their real names disguised by formolad or similar 

 tricks of arbitrary change. As to formolad there seems to be some evi- 

 dence as to its being practised in the twelfth century ; but is there any 

 that it was in the ninth or in the sixth ? Moreover, if the Ogmic names 

 of Goidelic Celts received a finishing touch from jargon-makers, so did 

 those of their Kimric brothers in Wales and Cornwall ; so did those 

 of the Picts in Bede's time, and so did those of Gaul in Caesar's time ; 

 but that seems to me highly improbable. Take for instance the fol- 

 lowing names Gaulish, Conomaglus ; this is in the Irish of the ' Chronic. 

 Scotorum ' Conmdl, and in the ' Four Masters' Conmhal, and in modern 

 Welsh Cynfael : or take the early Welsh Maglocunum of Gildas, which 

 Bede gives still with an o between the / and the c as Meilochon. In 

 modern Welsh it becomes Maelgivn ; if you apply to these the ascer- 

 tained rules of phonology in Irish and Welsh you will find that the 

 later forms Conmdl and Maelgwn postulate Conomaglus and Maglocunus, 

 or similar forms of equal length as their real vernacular prototypes. 



For it must not be forgotten that the normal state of language is 

 that of change, and of change tending to the shortening of its words, 

 so it Avill not do to transpose a nineteenth century name, which is a com- 

 pound and measures two syllables, into the sixth century : instead of 

 two syllables, four would be more likely to represent its length then. 

 Thus suppose it were argued that the o in Dunocati in a British inscrip- 

 tion is the arbitrary insertion of a scribe, and we were obliged to 

 accept Duncati, then according to the rules of phonology we should 

 find the name assuming in modern Welsh the form Dinghad, and in 

 modern Irish Docadh ; but the actual forms are Dingad and Donnchadh 

 respectively. There is no escaping from these phonological laws, and 

 any one who wishes to maintain a theory of manufactured Ogmic 

 names will have to disprove all that Celtic philologists were supposed 

 to have made out. 



