The use of thawed frozen temporal bones offers advantages over fresh bones
in the study of middle-ear and inner-ear mechanical function. We show, howe
ver, that freezing and thawing can cause a reduction in the magnitude of th
e input impedance of the stapes and cochlea Z(SC) in unfixed temporal bones
from human cadavers of as much as a factor of 3-10 over the frequency rang
e 25 Hz-7 kHz. Z(SC) is considered to be the sum of the impedances of the a
nnular ligament Z(S) and the cochlea Z(C) and has been shown to be controll
ed by Z(S) below 1 kHz and by Z(C) at higher frequencies [Merchant et al.,
1996. Hear. Res. 97; 30-45]. Experiments in which the inner ear was opened,
drained, and refilled identified two mechanisms by which freezing and thaw
ing can cause a reduction in the magnitude of Z(SC) (\Z(SC)\) Freezing can
allow air to enter the inner ear, with the result that \Z(C)\ is reduced ab
ove about 1 kHz; and freezing can reduce \Z(S)\, which causes a reduction i
n \Z(SC)\ below 1 kHz. Changes in the phase angle of Z(SC) induced by freez
ing were small and were consistent with changes in \Z(SC)\. Removing air fr
om the inner ear returned Z(C) to near its value in fresh bones, but \Z(SC)
\ remained lower in some thawed bones by a factor of 2-3. Investigations of
middle-ear function for which Z(SC) is critical should use fresh temporal
bones only or should allow for the possible reduction in \Z(SC)\ in thawed
frozen bones. (C) 2000 Elsevier Science B.V. All rights reserved.