Haigh — Earliest Inscribed Monuments. 435 



Hie) is much shorter than the first, to keep clear of the m ; for if it 

 had been of the same length it would have run into it. 



At Crickhowel, too, although the two inscriptions commemorate 

 the same person, the Ogham was written first ; for the legs of the n 

 are spread out wide to keep clear of the two scores of the l. Indeed 

 the amplification of the Latin legend, as compared with the Ogham, 

 is an indication of afterthought, and the same remark will apply to 

 the inscriptions which supply paternity, omitted in the Ogham. My 

 conclusion, therefore, would be, that these monuments belong to a 

 people who ordinarily used Ogham writing ; that Latin inscriptions 

 were sometimes added for the sake of the people amongst whom they 

 dwelt ; and that these people were more numerous than the Ogham- 

 writing people, because their monuments are more numerous than 

 those inscribed with Oghams. It would be quite natural that the 

 Ogham- writing people, after long residence in these districts, should 

 adopt the writing of the country and lay aside their own : thus, 

 m^ep^i alone is written on the monument at S. Florence, but the 

 monument of his son from Pant y Polion has a Latin legend only ; and 

 en&b&ppi alone appears on the Buckland stone, but the epitaph of 

 his son is in Latin. On the other hand, the occasional use of the 

 Ogham by the Latin-writing people may easily be supposed. 



The Gaulish inscription at Todi, giving dbutignos = dkttti filitjs, 

 has taught us that -gnos is patronymic. So also teaches the Irish 

 Ogham legend at Monataggart, -o^tA^ni m&<ji t>&bi. ignos and 

 agnos have become -en and -an in later speech ; and indeed the early 

 monuments exhibit the first traces of this change in names, such as 

 Dpoc&ni, sagrani, and sevemni fili seveei at Llan Newycld. But 

 even the fuller form is contracted from ganos or genos, of which we 

 have interesting examples in brocagan, and in the genitive utogeni. 

 The St. Dogmaels monument presents the form -&mni by the side 

 of -&111, as it were a link in the connexion, already supposed, be- 

 tween the Greek a/xvos and Latin agnus, (unless we admit the like- 

 lihood that an Ogham score has been omitted, and that psjgp&mni 

 stands for 'p&^p&^ni). 2 One other form of patronymic these legends 

 supply, -ing or -eng, in ercilingi, evolengi, identical with the common 

 Teutonic form. 



The discovery of the S. Dogmaels monument, giving m&<|i = fili, s 

 fully confirmed Dr. Graves' theory, already universally accepted, as 

 to the sense of m&cji, genitive of uiA^-p. The plural nominative of 

 this word seems to occur in the Ardmore legend — 



coinuc<s-p e;c;&mon&"p tu^u-oeccAf m&qi T>ot&ci bi^oe^obi ; 



and the genitive plural in one of those at Ballintaggart, 



T\\Ms m&<\& ni.ML&pii, 



2 See note (J) added in press. 3 See note (e) added in prefs. 



SEE. II., VOL. I., POL. LIT. AND AXTIQ. 3 If 



