88 NAMES OP PLACES 



localities it becomes necessary to indicate. The names 

 suggest themselves from some characteristic of the place, 

 some incident or occupation, or some individual who 

 dwells in or takes possession of a tract. They come 

 without conscious mental effort, such as all men in- 

 stinctively avoid ; they alter under use in compliance with 

 well-known phonetic laws arising out of the unconscious 

 avoidance of physical effort in pronunciation, to which all 

 of us unwittingly yield. 



Second, in nine cases out of ten place-names are 

 compound — made up, that is, of a substantive and a 

 qualitative. The first denotes the generic object — a hill, 

 a house, a river, a what not; the second some specific 

 attribute, indicated by a man's name or perhaps a beast's, 

 or by an adjective denoting position, colour, contour, or 

 some other characteristic. When the distinction between 

 these two parts has been settled, the first obstacle to 

 understanding the name has been overcome. Well, the 

 stress invariably remains on the qualitative syllable, 

 which, in English names, usually precedes the other. 

 Thus you say Oxford, Milnthorpe, Hknbury — not Oxford, 

 Milnthorpe, or Hanbury. You may rely on the fidehty 

 with which this pronunciation has been handed down in 

 each locality from generation to generation : it is a sure 

 key to the construction of the name, though railways are 

 apt to import obscurity into it. Thus Carlisle — ccer 

 Llewellyd — being a Celtic name, has the qualitative last. 

 The natives call it correctly CarUsle, but railway-people 

 and southerners may be heard talking of it as Carlisle — 

 which is wrong. 



As has been said, according to a general rule, the Celtic 

 qualitative follows the substantive ; but there are ex- 



