voices, 42 of whom were "trained" singers and 8 "natural" singers,
Mackenzie formulated his conclusions as follows:

1. In tenor voices the whole glottis may be open to [Illustration: g a
b] and not unfrequently to [Illustration: g']. Beyond this point there
is closure of the cartilaginous glottis. Sometimes the whole glottis
is open throughout.

2. In barytone voices the whole glottis is often open to
[Illustration: a b], and occasionally to [Illustration: c']. Beyond
this point the cartilaginous glottis is closed, except in rare cases.

3. In bass voices the whole glottis is sometimes open to
[Illustration: g b]. Beyond this point, except in a few instances, the
cartilaginous glottis is gradually closed.

4. In sopranos and mezzo-sopranos the whole glottis is sometimes open
to [Illustration: f' g'], often to [Illustration: c''], beyond which
the cartilaginous glottis is usually closed. The glottis is sometimes
closed throughout the scale, and in one case it was open throughout.

5. In contralto voices the whole of the glottis is often open to
[Illustration: f' g'], beyond which the cartilaginous portion is
closed.

6. In the head voice of women and the falsetto voice of men
"stop-closure" (_i.e._, closure so tight that the cords in this region
do not vibrate) always takes place in the posterior portion of the
ligamentous glottis, and sometimes at the anterior part also.

This writer also held that "Boys who sing alto always use the chest
register." He was of opinion that "The quality of the voice generally,
but not always, indicates which mechanism is being used."

The views of the author, published at a former period, and based on
the special examination of a large number of persons with the
laryngoscope, etc., and on auto-laryngoscopy, may be briefly stated as
follows:

A nomenclature for the registers involving no theory would be best,
such, for example, as _lower_, _middle_, and _upper_ registers. Mandl,
who recognized only two registers, spoke of them as "lower" and
"upper," equivalent to "chest" and "head," as commonly used.

The author examined with the laryngoscope 50 persons, who might (with
Gruetzner) be divided into "trained singers," "natural singers," and
"non-singers." The whole glottis was found to be open in all voices in
the lowest tones of the chest register, and this condition obtained up
to about [Illustration: f-sharp' g'], beyond which another mechanism
came into play, except in rar