icles M, I, and S are respectively the _malleus_ (hammer),
_incus_ (anvil), and _stapes_ (stirrup). Each is attached by ligaments
to the walls of the middle ear. The tympanum moves the malleus, the
malleus the incus, and the incus the stapes, the last pressing into the
opening O of Fig. 133, which is scientifically known as the _fenestra
ovalis_, or oval window. As liquids are practically incompressible,
nature has made allowance for the squeezing in of the oval window
membrane, by providing a second opening, the round window, also covered
with a membrane. When the stapes pushes the oval membrane in, the round
membrane bulges out, its elasticity sufficing to put a certain pressure
on the perilymph (indicated by the dotted portion of the inner ear).

[Illustration: FIG. 134.--Diagrammatic section of the ear, showing the
various parts.]

The inner ear consists of two main parts, the _cochlea_--so called from
its resemblance in shape to a snail's shell--and the _semicircular
canals_. Each portion has its perilymph and endolymph, and contains a
number of the nerve-ends, which are, however, most numerous in the
cochlea. We do not know for certain what the functions of the canals and
the cochlea are; but it is probable that the former enables us to
distinguish between the _intensity_ or loudness of sounds and the
direction from which they come, while the latter enables us to determine
the _pitch_ of a note. In the cochlea are about 2,800 tiny nerve-ends,
called the _rods of Corti_. The normal ear has such a range as to give
about 33 rods to the semitone. The great scientist Helmholtz has
advanced the theory that these little rods are like tiny tuning-forks,
each responding to a note of a certain pitch; so that when a string of a
piano is sounded and the air vibrations are transmitted to the inner
ear, they affect only one of these rods and the part of the brain which
it serves, and we have the impression of one particular note. It has
been proved by experiment that a very sensitive ear can distinguish
between sounds varying in pitch by only 1/64th of a semitone, or but
half the range of any one Corti fibre. This difficulty Helmholtz gets
over by suggesting that in such an ear two adjacent fibres are affected,
but one more than the other.

A person who has a "good ear" for music is presumably one whose Corti
rods are very perfect. Unlucky people like the gentleman who could only
recognize one tune, and that because people to