._, he must really hear it; (2) observe the shape of the
resonance-chambers; (3) try to produce the same shape of his own, and
under the guidance of his ear and his eye (watching the mouth of the
teacher) so utter the sound correctly. This sound should be fixed in
the mind, and the ear trained by comparing it with other sounds, as
the wise teacher will do, and require imitations. Any language can be
pronounced correctly in a short time, if this method be followed. It
is, indeed, the only one that rests on science and common sense. The
student when away from the teacher, after he has once learned to form
the vowels correctly, should practise with a hand-glass before him for
some time, at least.

The learning of a new language is the acquiring of a new mouth, or, at
all events, entirely new methods of using the old one. In reality,
however, this is not so fully the case as it at first seems. In all
the languages one wishes to acquire, the same vowels occur, and for
the learner it is often a question of lower or higher pitch, or
greater or less breadth, though all this involves the formation of new
habits and the fighting of old ones, and often in the case of the
adult the struggle is a long-continued and severe one. Some nations
speak at a lower pitch than others, and if a foreigner enunciate ever
so well, yet at the pitch of his own and not that of the new language,
his utterance may seem foreign. The Germans speak at a much lower
pitch than Americans, and their tongue, even when grammatically spoken
by the latter, is apt to have a sort of foreign flavor. It slightly
disturbs the listener, who is not accustomed to hear his mother-tongue
transposed into another key, so to speak.

We have known a learner to derive great benefit from having it pointed
out to him that certain of his vowel sounds would at once cease to be
incorrect if their pitch were altered. Of course, in doing this, there
were at once many changes made in the resonance-chambers, in order to
get the changed pitch. Pitch, accent, and duration of the sound throw
much light on the subject of dialect, as a little analysis of Irish or
Scotch will show.

Consonants are, as we have already said, noisy nuisances for the
singer, but indispensable for word-formation, and so for human
intercourse. Each has also its own pitch, and investigators have come
to a measurable degree of agreement on this subject.

To illustrate: Madame Seiler found that _r_ and _s_ are se