d behind
the nose, toward the cavities of the head, of a strong but very
elastic rubber ball, which I fill like a balloon with my breath
streaming up far back of it. And this filling keeps on in even
measure. That is, the branch stream of the breath, which flows into
the head cavities, must be free to flow very strongly without
hindrance. (See Plate B.)

I can increase the size of this ball above, to a pear shape, as soon
as I think of singing higher; and, indeed, I heighten the form
_before_ I go on from the tone just sung, making it, so to speak,
_higher_ in that way, and thus keep the form, that is, the
"propagation form," ready for the next higher tone, which I can now
reach easily, as long as no interruption in the stream of breath
against the mucous membrane can take place. For this reason the breath
must _never be held back_, but must always be emitted in a more and
more powerful stream. The higher the tone, the more numerous are the
vibrations, the more rapidly the whirling currents circulate, and the
more unchangeable must the form be.

Catarrh often dries up the mucous membrane; then the tones are
inclined to break off. At such times one must sing with peculiar
circumspection, and with an especially powerful stream of breath
behind the tone: it is better to take breath frequently. In a
descending scale or figure I must, on the contrary, preserve very
carefully the form taken for the highest tone. I must not go higher,
nor yet, under any circumstances, lower, but must imagine that I
remain at the same pitch, and must suggest to myself that I am
striking the same tone again. The form may gradually be a little
modified at the upper end: that is, the soft palate is lowered very
carefully behind the nose: keeping almost always to the form employed
for the highest tone, sing the figure to its end, toward the nose,
with the help of the vowel _oo_. (This auxiliary vowel _oo_ means
nothing more than that the larynx is slowly lowered in position.)

When this happens, the resonance of the head cavities is diminished,
that of the palate increased; for the soft palate sinks, and the
pillars of the fauces are raised more and more. Yet the head tone must
not be entirely free from palatal resonance. Both remain to the last
breath united, mutually supporting each other in ascending and
descending passages, and alternately but inaudibly increasing and
diminishing.

These things go to make up the form:--

The raising and