inty, that _a_ is a short letter, and _b_ a long one;
the former a vowel, the latter a consonant: and so of others. Yet as he may
doubt whether _t_ is a long letter or a short one, so he may be puzzled to
say whether _w_ and _y_, as heard in _we_ and _ye_, are vowels or
consonants: but neither of these difficulties should impair his confidence
in any of his other decisions. If he attain by observation and practice a
clear and perfect pronunciation of the letters, he will be able to class
them for himself with as much accuracy as he will find in books.

OBS. 3.--Grammarians have generally agreed that every letter is either a
vowel or a consonant; and also that there are among the latter some
semivowels, some mutes, some aspirates, some liquids, some sharps, some
flats, some labials, some dentals, some nasals, some palatals, and perhaps
yet other species; but in enumerating the letters which belong to these
several classes, they disagree so much as to make it no easy matter to
ascertain what particular classification is best supported by their
authority. I have adopted what I conceive to be the best authorized, and at
the same time the most intelligible. He that dislikes the scheme, may do
better, if he can. But let him with modesty determine what sort of
discoveries may render our ancient authorities questionable. Aristotle,
three hundred and thirty years before Christ, divided the Greek letters
into _vowels, semivowels_, and _mutes_, and declared that no syllable could
be formed without a vowel. In the opinion of some neoterics, it has been
reserved to our age, to detect the fallacy of this. But I would fain
believe that the Stagirite knew as well what he was saying, as did Dr.
James Rush, when, in 1827, he declared the doctrine of vowels and
consonants to be "a misrepresentation." The latter philosopher resolves the
letters into "_tonics, subtonics_, and _atonics_;" and avers that
"consonants alone may form syllables." Indeed, I cannot but think the
ancient doctrine better. For, to say that "consonants alone may form
syllables," is as much as to say that consonants are not consonants, but
vowels! To be consistent, the attempters of this reformation should never
speak of vowels or consonants, semivowels or mutes; because they judge the
terms inappropriate, and the classification absurd. They should therefore
adhere strictly to their "tonics, subtonics, and atonics;" which classes,
though apparently the same as vowels, sem