206 £ssaif Introductort/ to the Archeeology of the fVest of England, 



consider the Welsh to be the representative, remained longer in tlie East, 

 and never extended their possessions beyond Gaul and South Britain. It 

 has indeed been sliewn, that a dialect closely resembling the Erse has 

 contributed with the Greek to form the Latin tongue, but the points of 

 difference and coincidence between the Celtic dialects have not yet been 

 sufficiently examined into, to throw light upon the whole hypothesis. 



The narrow limits of this essay, independently of more cogent reasons, 

 will not admit of the detail requisite for comparing satisfactorily the extant 

 Celtic dialects, so that we must content ourselves with a few observations 

 respecting their differences, a method of examination, the reverse of Dr. 

 Prichard's, but which it is obvious must be adopted, when the growth of 

 dialects from a parent stock is to be estimated. 



The changes, to which Celtic in common with other languages is liable, 

 are of three kinds, literal, ideographical, and grammatical, the former affect- 

 ing the orthography and signification of the word, the second its sense or 

 meaning, the latter its inflections. 



Thus words obsolete in some dialects are retained in others: Kail, crafty, 

 Maur, great, and others, are retained in Welsh but are obsolete in other 

 dialects. 



Another class are transpositions due to careless pronunciation, Brecon 

 for Brecknock, etc. Another, transpositions of whole syllables, Kyn-vael 

 for Mael-gun ; lltyd for Tydvil. 



Sometimes letters are omitted, and changes made to facilitate pronun- 

 ciation : thus Cassibelau becomes Cassivelan. These are partly grammatical, 

 or the initial syllable may be varied, as in the Welsh Guan, week, which be- 

 comes in Erse, Anvan. Though the difference in most cases, as perhaps in 

 this, depends upon a separate prefix. 



Nations usually adapt the terminations of foreign words to the rules of 

 their own language. The Romans thus altered British words, and the 

 Britons Roman. Some of those Greek or Latin technical terms, which, 

 from their connexion with Christianity, became suddenly adopted into a 

 great many languages, display this rule remarkably. Thus iKiaKo-n-og, or 

 Episcopus, becomes in Italian, Vescovo ; in Spanish, Obispo; in French, 

 Evesque, or Eveque ; in Hungarian, Pispok ; in German, Bischoff; in Eng- 

 lish and Dutch; Bishop ; in Welsh, Esgob ; and in Erse, Easbog. 



But by far the most extensive class of literal differences, are those which 

 arise from the permutation of one letter into another. Thus all letters in 

 the articulation of which the same parts of the organs of voice are con- 

 cerned, are mutable into each other, labials into labials, etc. 



Also certain letters of one class in one language, are mutable into certain 

 letters of a different class in another. Thus the labial P, in Welsh, becomes 

 in Erse, C or K : Paul, a pole, becoming Kfial j Pen, a head, Kean, etc. ; 

 just as the Greek \vkoq, a wolf, becomes in Latin, Lupus ; or k, in the 

 Ionic and /Eolic, became not unfrequently n in the other Grecian dialects : 



