Tympanic‐membrane (TM) vibrations were studied in fresh human‐cadaver specimens by means of time‐averaged holography. As with an earlier study in cats [S. M. Khanna and J. Tonndorf, J. Acoust. Soc. Amer.51, 1904 (1972)], the findings supported the curved‐membrane hypothesis of Helmholtz. Despite a careful search for experimental artifacts, no satisfactory explanation was found to account for the fact that the present results are at variance with the stiff‐plate concept of von Békésy. With respect to their quantitative evaluation, the human‐cadaver data were not as clear‐cut as the earlier cat data on account of anatomical differences. A number of points, which on first glance do not seem to agree with Helmholtz's concept are being discussed in the light of present or other recent findings, and explanations are being offered. These topics include: (1) the compliance of the TM; (2) the effects of differential air pressure across the TM upon middle‐ear transmission; (3) the effects of TM perforations; and (4) the fact that Helmholtz's cadaver experiment, which was thought to support his hypothesis, could not be duplicated.