“…The samples were chosen to ensure that the slit widths and array periodicities remained subwavelength to incident sound (100w < λ 0 < 300w, and 8Λ < λ 0 < 14Λ), avoiding any diffractive phenomena. The measurements were taken under different temperatures, pressures, and humidities, the latter two having negligible effect [24]. A temperature change causes a small but systematic frequency shift of 1.64 × 10 −3 f Hz −1 K [23], and thus all data were normalized to 293.15 K. Small systematic changes in all slit widths are included to account for the fact that, for the slit array, bowing of the slats will give somewhat wider gaps than the measured spacers, while for the single slits compression of the spacers by the weight of the sample and its baffle reduced the width to below that intended.…”