30 BROOMALL : 



inju'ry and his'tory from histo'ry. A preference is usually indi- 

 cated in the dictionary by placing the supposed better pronun- 

 ciation before the other. The lexicographer, in giving " abdo' - 

 men ox ab'domen " indicates, by placing it first, that he pre- 

 fers abdo' men, the older, classically derived pronunciation, 

 founded upon a L,atin long o, and therefore accented. His 

 predecessors admitted only abdo'^nen ; he is now forced to 

 acknowledge ab' domen also ; his successors may give simply 

 ab'domen , just as he already gives only in' jury and his'tory. 

 These changes of preference in variant pronunciations, traced 

 in successive dictionaries, would show the gradual predomi- 

 nance of popular forms, such as are continually changing 

 our own old words and forcing new words to accommodate 

 themselves to English phonetic laws. But the present inves- 

 tigation is of a narrower scope. The specimens to be consid- 

 ered are those of variant accentuation acknowledged by the 

 lexicographer as simultaneously extant. Any lexicographic 

 authority would suffice for Ihis purpose, and the Century is 

 selected for our consideration because it is both scholarly and 

 popular. 



In the Century Dictionary two different pronunciations 

 are given to 2234 words. This count includes derivatives as 

 well as originals. Thus, where "' deco'rons or dec'orous'' is 

 authorized, and also '' deco'rously ox dec'oronsly,'' the adverb 

 is counted as an additional case of variant pronunciation. 

 Derivatives and originals do not always suffer the same 

 uncertain t}' of pronunciation, and therefore, where it does 

 occur, it should be counted. Again, where two different 

 spellings, either causing or accompanying two correspondingly 

 different pronunciations, seem to represent an incipient differ- 

 entiation of meaning, the forms are not strictly one word and 

 are not counted as instances of variant pronunciation for our 

 present purpose. Thus tJursh and thrash, although etymolo- 

 gically the same word, are not counted because they are now 

 receiving a difference of application, since we thresh wheat 

 and thrash persons. The development, in such a case, has 



