101 



the gold diggings of the island, and furnishing ornaments of that metal 

 conformable to their state of culture. In my view it is to the nearer 

 Iberians, rather than to the distant Phoenicians, we are to look for the 

 chief pioneers of commercial intercourse in those epochs. 



To arrive at a sound judgment on this subject, a series of researches 

 is required. 



One most important branch is the collection and analysis of the 

 topographical names in Ireland, to be obtained from the Ordnance 

 Survey, and other authorities. Every name should be investigated, 

 even the names of fields. Undoubtedly this topographical nomen- 

 clature will be found to be almost without exception Hiberno- Celtic, 

 and much of it modern; but in investigation it will yield results 

 illustrating the Celtic occupation, and even in that respect the 

 anterior possession by another race. 



I have observed it is a law in topographical nomenclature that 

 where a race, altogether foreign in language, enters a country, it 

 applies a system of terms to the settlements of the formerly ex- 

 isting rejected race. ' This is what we observe in England, where 

 words purely English or Anglo-Saxon give tens of thousands of 

 evidences of Roman occupation, even to the names of wells. This 

 nomenclature follows a law conforming to that applied by the Ger- 

 manic population to the Roman colonies on the Rhine, and their 

 outliers. Thus such a term as Cold Harbour will be found ex- 

 tensively distributed in England, the Netherlands, and Western 

 Germania. The same law is found in Asia Minor in its application 

 by the Turks to the sites of Greek cities and establishments, 

 where we have Ax Hissae and Esxi Hissar, representing the Whit- 

 Chester and Old Chester of the Anglo-Saxons. 



The words must be carefully analyzed and classified, compounds 

 being entered under each of their elements. The classification will 

 include the names of each class of object, as rivers, hills, towns, home- 

 steads, fields, wells, &c, and it will distribute each root into its own class. 

 It is then necessary to eliminate all the modern names, and carefully 

 examine what are recognized as more ancient names. All names 

 occurring since the English Settlement must be excluded, and the 

 ancient residuum carefully studied. It will most likely be found that 

 certain terms occur more or less in groups, and the details of situation 

 will afford ground for identification. 



It will most probably result that there is a residuum, containing 

 first Celtic words, expressive of anterior settlement ; and, secondly, of 

 words doubtfully Celtic, or other than Celtic. 



In my opinion the names of the great rivers in Ireland, claimed as 

 Celtic, are not Celtic, but conform to the names of rivers found in the 

 non- Celtic or Iberian area. The determination of this point is very 

 desirable ; for it has generally been assumed that the names of the great 

 rivers of north- western Europe are Celtic ; but the explanation of the 

 names of the rivers of Spain, Italy, and Asia Minor, has to be settled 



