ial modification of the male voice seldom employed now, and almost
never except by tenors. With this writer, "falsetto" as applied to
female voices replaces "middle," in the commoner usage.

[Illustration: FIG. 50. Tabular representation of Madame Seiler's
division of the register.]

Garcia, also, recognized five registers. Behnke, a teacher of singing,
who practised laryngoscopy and auto-laryngoscopy in the investigation
of the registers, used "lower thick," "upper thick," "lower thin,"
"upper thin," and "small," as answering to the "first chest," "second
chest," etc., of Madame Seiler and others.

Nearly all writers have used the term "break" to indicate the point at
which a new register begins. Behnke held that the break between the
thick and the thin register occurred in _both_ sexes at about
[Illustration: f' f-sharp']. The vocal bands in this part of the scale
vibrate in their entire breadth, and the series of tones above the
point just referred to is produced by a new mechanism, but one which
is the same for all voices and both sexes--_i.e._, only the inner
edges of the vocal bands vibrate.

According to Behnke, the male voice has but two registers, the thick
and the thin, but the female voice three, the thick, the thin, and the
small. These terms were not original with Behnke, but had been used
earlier by Curwen.

Behnke was emphatic on one point, to which we would call special
attention, in his own words: "If there is _straining_ anywhere, it is
during the attempt to carry the mechanism of the upper thick beyond
its natural limit."

Mackenzie (afterwards Sir Morell Mackenzie) held that "It is certain
that however over-refined musicians may multiply the 'registers' of
the voice, physiologically there are but two--_i.e._, 'chest' and
'head,' the falsetto of the man answering to the head production of
women."

According to the same author, "The essential factor in chest
production is the long reed, whilst the essential factor in head
delivery is the short reed." The terms "long reed" and "short reed"
were the equivalents of Madame Seiler's "glottis" and "ligamentous
glottis" respectively. Mackenzie held that the cartilaginous
(inter-arytenoid) glottis is generally open in the lower and gently
closed in the upper tones of the chest register, while a segment of
the ligamentous glottis (vocal bands proper) is tightly closed in the
head voice.

As the result of the examination of 50 persons gifted with fine