Olden — The Number Tivo in Irish Proper Names. 637 



The first question is, how far back can we trace this custom. Dr. 

 Joyce says we find it in the earliest MSS., but I think we can trace it 

 back farther than any existing manuscript. In the fifth, century, as 

 is well known, a colony from Ireland settled in Argyleshire, and it 

 occurred to me that if the custom existed in Ireland at that time the 

 emigrants would probably have taken it with them, and we should 

 find similar names there at the present day. To ascertain whether 

 this was so or not, I wrote to Mr. Donald Mackinnon, Professor of 

 Gaelic in the Uniyersity of Edinburgh. In his reply he says, *'^a, 

 two, is quite common in place-names in the Scottish Highlands," and 

 he mentioned a few names that occurred to him offhand. These are 

 Acha{dh)-dd-dalaich in Colonsay, Acha{dh)-dd-I)omhmiill in Eoss- 

 shire, Dun da laimhe in Inrern ess-shire, Dun-da-ramh in Argyleshire. 

 He most kindly ojGPered to supply me with, a list of such, names, but I 

 did not think it right to give him the trouble, as all I wanted was 

 the fact of their occurrence. 



Eut can we trace the custom still further back ? I think we can, 

 for we may argue in this way, if the immigrants to Scotland in th.e 

 fifth century took it with them to that country, it was possible that 

 the original settlers in Irehmd in the remote past may have brought 

 it with them from the Continent of Europe, and if so that such, names 

 might still be found abroad. 



It is generally allowtd that the branch of the Celtic race which 

 settled in Ireland were the earliest of the Aryan peoples who appeared 

 in Europe, and for this reason it is that the names of many of the 

 permanent features of the landscape throughout Europe are of Gaelic 

 origin, having been imposed by them. 



All local names recognised as Celtic abroad are given in the 

 " Alt-Celtischer Sprachschatz " of Alfred Holder; and on consulting 

 this I found one which seemed to throw light on the subject. The 

 word Conddte, the penultimate being long, appears to have our word 

 da as part of its composition, and this derives confirmation from the 

 fact that the name occiirs twenty-seven times in mediceval documents 

 in France, and always with the meaning of a "confluence of hvo 

 rivers." M. Arbois de Jubainville, however, proposes a derivation 

 for it of a different kind. He regards it as equivalent to the Greek 

 cruV^ecns, and composed of the particle co7i and '^datis, originally *dotis 

 or ""dhotis, Greek ^ecrts, originally +^€tis, the root being dJie, ' to place or 

 lay.' I have not been able to find that o-w^eo-ts is ever applied to a 

 confluence of rivers ; but apart from this the proposed derivation is 

 highly conjectural, and requires many assumptions, while it fails to 



K.I.A. PEOC, SER, III., VOL. IV. 2 Z 



