44
Henry C. Bennet-Clark
the 12-cycle song pulse is excited by two or three impulses, the first of which
occurs at the start of the pulse with further impulses at intervals of two or three
cycles. In crickets such as Gryllus campestris, the sound pulse is sustained on a
cycle-by-cycle basis by the catch and release of the teeth of file on one wing by the
plectrum or scraper on the edge of the other wing (see reviews by Bennet-Clark
1989, 1999); between 35 and 65 %. of the 100 or more teeth on the file are
engaged to make the pulse (Leroy 1966).
Whether resonant or not, the muscle-powered vibrating structure appears to
have two major roles: to act as a frequency multiplier and as a determinant of the
sound frequency. However, this primary vibrator may not be the effective sound
radiator: sound radiated from the outside of the tymbal of the cicada C.
australasiae is 13 dB quieter than that radiated via its tympana (or ear drums). For
optimum impedance matching between the muscle power source and the vibrating
structure, it is mechanically feasible to drive stiff structures with a large mass
because the muscle can produce large pressures.
Thus, the primary vibrating structure may be a small thick structure, with large
mass and high stiffness, rather than one that is lighter and less stiff with the same
resonant frequency. Examples of this are cicada tymbals or vertebrate larynges and
syrinxes, none of which act as the primary or sole sound radiating surface. In many
animals there is an additional stage of transduction and impedance matching (Sect.
3.4).
3.4 Problems of Sound Radiation
Impedance matching between source and medium is well known from acoustical
engineering (Olson 1957). The specific acoustic resistance of a source depends
on its configuration: whether it is a freely-suspended piston, a piston in the side of
a closed box, a pulsating sphere etc.; sources of all these types are found in
animals.
The relationship between the radius of a source and its specific radiation
resistance relative to that of the surrounding medium is shown in Fig. I for two
types of source: a monopole source or a vibrating piston in an infinite baffle (Fig.
IA) and a dipole source or a free-edged piston vibrating normal to its plane (Fig.
IB). Because there is no back-to-front leakage with the monopole source, its
specific acoustic resistance decreases less rapidly with decreasing size than that of
the free edge piston, where back-to-front leakage becomes worse as the source
becomes smaller.
There will be good impedance matching between the source and its
surroundings if the radius of a momopole exceeds 0.16 of the sound wavelength A,
or, if a dipole, it exceeds 0.24 A (Fig. I C). Table 2 gives examples of these types of
source found in animals.
Henry C. Bennet-Clark
the 12-cycle song pulse is excited by two or three impulses, the first of which
occurs at the start of the pulse with further impulses at intervals of two or three
cycles. In crickets such as Gryllus campestris, the sound pulse is sustained on a
cycle-by-cycle basis by the catch and release of the teeth of file on one wing by the
plectrum or scraper on the edge of the other wing (see reviews by Bennet-Clark
1989, 1999); between 35 and 65 %. of the 100 or more teeth on the file are
engaged to make the pulse (Leroy 1966).
Whether resonant or not, the muscle-powered vibrating structure appears to
have two major roles: to act as a frequency multiplier and as a determinant of the
sound frequency. However, this primary vibrator may not be the effective sound
radiator: sound radiated from the outside of the tymbal of the cicada C.
australasiae is 13 dB quieter than that radiated via its tympana (or ear drums). For
optimum impedance matching between the muscle power source and the vibrating
structure, it is mechanically feasible to drive stiff structures with a large mass
because the muscle can produce large pressures.
Thus, the primary vibrating structure may be a small thick structure, with large
mass and high stiffness, rather than one that is lighter and less stiff with the same
resonant frequency. Examples of this are cicada tymbals or vertebrate larynges and
syrinxes, none of which act as the primary or sole sound radiating surface. In many
animals there is an additional stage of transduction and impedance matching (Sect.
3.4).
3.4 Problems of Sound Radiation
Impedance matching between source and medium is well known from acoustical
engineering (Olson 1957). The specific acoustic resistance of a source depends
on its configuration: whether it is a freely-suspended piston, a piston in the side of
a closed box, a pulsating sphere etc.; sources of all these types are found in
animals.
The relationship between the radius of a source and its specific radiation
resistance relative to that of the surrounding medium is shown in Fig. I for two
types of source: a monopole source or a vibrating piston in an infinite baffle (Fig.
IA) and a dipole source or a free-edged piston vibrating normal to its plane (Fig.
IB). Because there is no back-to-front leakage with the monopole source, its
specific acoustic resistance decreases less rapidly with decreasing size than that of
the free edge piston, where back-to-front leakage becomes worse as the source
becomes smaller.
There will be good impedance matching between the source and its
surroundings if the radius of a momopole exceeds 0.16 of the sound wavelength A,
or, if a dipole, it exceeds 0.24 A (Fig. I C). Table 2 gives examples of these types of
source found in animals.