River-Names. . 161 



signifies water. It is found in the Latin as aqua. As illus- 

 trations of a similar change of consonants we may compare 

 the Greek 1917709 and iiivri with the Latin equus and quinque. 

 The tennination " on " or " en " is prohably expressive of a 

 distinct unity, in fact the mode of distinguishing the concrete 

 from the abstract/ so that the word Av-on signifies literally 

 " a river/^ According to the oldest orthography, Lhuyd 

 says, ^«;-on was written Am-on, so that we see its connection 

 with the Latin " am-nis " a river.' 



The following names of places are derived from the various 

 forms of this River-Name : — 



Avon. The appellation given to two hamlets, one close to Foxham, 

 on the banks of the Upper Avon : the other by Old Sarum, 

 situated on the Lower Avon. 



Up- Avon ") Two villages a few miles apart from each other. 



Nether- Avon j situated on the Lower Avon. 



AvENiNG. Close by Tetbury, near the site of the upper Avon, on 

 the Gloucestershire border. In an old Charter^ of A.D. 896, 

 we have lands near this place alluded to as " sum to ^fen- 

 ingum," i.e., " some at Avening.-'-' The Anglo-Saxon form 

 can only mean literally " the dwellers, or clan, settled near 

 the Avon " (or river) . 



' It is certainly so used in Welsh : thus hrwyn means rushes, whilst hrwyn- 

 en means "'a rush." — In like manner derw is the uak, or oak-trees, whilst 

 Deru)-en signifies " an oak," 



' Lhuyd gives a number of examples shewing how, in cases where words are 

 common to the Latin and "Welsh, the letter m in the former, becomes v in the 

 latter : — e g., Lat. elewen-tum, Welch, elren : — Lat. firnius, Welsh, fyr« : — 

 Lat : — termin-us, Welsh, tertyn. A suggestion has been made that in the 

 word Ampnet, (formerly written Amenie and Amhei,) a name given to 

 several parishes on the Gloiicestershire border, this change of consonants has 

 taken place, and the root of the word brought more like the Latin amnis. There 

 are in the immediate vicinity two of the great Roman roads, and other names 

 that denote Roman occupation, so that thftre is nothing improbable in the 

 conjecture. 



' Cod. Dipl. 1073. The other places named, such as Roddanbeorg (Rod- 

 borough), and Dornbyrig (Thornbury), prove that Avening is the place alluded 

 to. See also Thorpe's Dipl. Angl. p. 139. 



