OBS. 15.—­In all late editions of L. Murray’s Grammar, and many modifications of it, accent is defined thus:  “Accent is the laying of a peculiar stress of the voice, on a certain letter OR syllable in a word, that it may be better heard than the rest, or distinguished from them; as, in the word presume, the stress of the voice must be on the letter u, AND [the] second syllable, sume, which takes the accent.”—­Murray’s Gram., 8vo, p. 235; 12mo, 188; 18mo, 57; Alger’s, 72; Bacon’s, 52; Comly’s, 168; Cooper’s, 176; Davenport’s, 121; Felton’s, 134; Frost’s El., 50; Fisk’s, 32; Merchant’s, 145; Parker and Fox’s, iii, 44; Pond’s, 197; Putnam’s, 96; Russell’s, 106; R.  O. Smith’s, 186.  Here we see a curious jumble of the common idea of accent, as “stress laid on some particular syllable of a word,” with Sheridan’s doctrine of accenting always “a particular letter of a syllable,”—­an idle doctrine, contrived solely for the accommodation of short quantity with long, under the accent.  When this definition was adopted, Murray’s scheme of quantity was also revised, and materially altered.  The principles of his main text, to which his copiers all confine themselves, then took the following form: 

“The quantity of a syllable, is that time which is occupied in pronouncing it.  It is considered as LONG or SHORT.

“A vowel or syllable is long, when the accent is on the vowel; which occasions it to be slowly joined in pronunciation with the following letters:  as, ‘F=all, b=ale, m=o=od, h=o=use, f=eature.’

“A syllable is short, when the accent is on the consonant; which occasions the vowel to be quickly joined to the succeeding letter:  as, ‘ant, b=onnet, h=ung~er.’

“A long syllable generally requires double the time of a short one in pronouncing it:  thus, ‘M=ate’ and ‘N=ote’ should be pronounced as slowly again as ‘M~at’ and ‘N~ot.’”—­Murray’s Gram., 8vo, p. 239; 12mo, 192; 18mo, 57; Alger’s, 72; D.  C. Allen’s, 86; Bacon’s, 52; Comly’s, 168; Cooper’s, 176; Cutler’s, 165; Davenport’s, 121; Felton’s, 134; Frost’s El., 50; Fisk’s, 32; Maltby’s, 115; Parker and Fox’s, iii, 47; Pond’s, 198; S.  Putnam’s, 96; R.  C. Smith’s, 187; Rev. T. Smith’s, 68.

Here we see a revival and an abundant propagation of Sheridan’s erroneous doctrine, that our accent produces both short quantity and long, according to its seat; and since none of all these grammars, but the first two of Murray’s, give any other rules for the discrimination of quantities, we must infer, that these were judged sufficient.  Now, of all the principles on which any have ever pretended