The Acoustic Mirror

Donald Griffin, Bat Biosonar, and Doppler-Shift Compensation

Inside a physics laboratory at Harvard University in the spring of 1938, undergraduate researcher Donald Griffin carried a wire cage of little brown bats (Myotis lucifugus) toward an apparatus built by Professor George Washington Pierce. Pierce had invented a sonic detector around a piezoelectric Rochelle-salt crystal microphone capable of transducing acoustic pressure waves well above the 20-kilohertz ceiling of the human ear and heterodyning them down into audible clicks on a loudspeaker. To Griffin and Pierce, the laboratory room was silent. Yet the moment Griffin held the cage before the parabolic horn of the crystal transducer, the loudspeaker erupted in a machine-gun rattle of 50-kilohertz pulses.

That afternoon resolved an experimental mystery that had stood for 145 years. In 1793, the Italian physiologist Lazzaro Spallanzani had discovered at the University of Pavia that bats blinded with hot needles or opaque hoods flew effortlessly through darkened vaulted rooms, dodging silk threads hung from the ceiling and returning to their roosts with stomachs full of nocturnal insects. One year later in Geneva, the surgeon Louis Jurine reported that plugging a bat's ear canals with warm wax or tallow turned the agile flyer into a helpless creature that collided with every wall. Because no human observer heard a sound during bat flight, the French anatomist Georges Cuvier mocked the acoustic hypothesis in 1795, insisting that bats navigated instead through a heightened tactile sensitivity in the wing membranes—a consensus that froze sensory biology until Griffin approached Pierce's crystal microphone.

Working alongside neurophysiologist Robert Galambos between 1938 and 1940, Griffin strung vertical grids of 1-millimeter steel wires at 30-centimeter intervals across a Harvard test chamber. Bats with open ears and uncovered mouths missed 75 percent of the wires; bats whose ears were plugged or whose mouths were gently tied shut with thread crashed into two-thirds of the barriers, identical to random chance. When a bat approached a wire, the repetition rate of its laryngeal ultrasonic cries accelerated from 10 pulses per second during cruising flight to a terminal buzz of 200 pulses per second just before contact. Griffin coined the word echolocation in Science in 1944 to describe active biological sonar.

While Myotis emits brief, broadband frequency-modulated (FM) sweeps that drop from 80 kilohertz to 40 kilohertz in two milliseconds to measure target range from echo delay, a second class of bats solved a far harder physical problem—detecting a fluttering insect buried inside the dense acoustic clutter of forest leaves. In 1968 at the University of Tübingen, the German neuroethologist Hans-Ulrich Schnitzler trained greater horseshoe bats (Rhinolophus ferrumequinum) to fly across a six-meter chamber while a miniature radio microphone rode on the animal's head. Unlike Myotis, the horseshoe bat emits long, 50-millisecond constant-frequency (CF) calls through its horseshoe-shaped nasal baffle at a resting frequency of approximately 83.0 kilohertz.

When the bat launches into flight at velocity $v_{\text{bat}}$ through air of sound speed $c \approx 343\text{ m/s}$, the outgoing wave is compressed and the returning reflection from stationary foliage is compressed a second time, shifting the echo upward according to the relativistic two-way acoustic Doppler equation:

At a flight speed of 5 meters per second, an 83.0-kilohertz pulse returns from a tree trunk shifted upward by 2.4 kilohertz to 85.4 kilohertz. Anatomical dissections of the Rhinolophus inner ear by Gerhard Neuweiler revealed why this shift matters. Roughly half of the entire length of the bat's spiraled basilar membrane—and over 50 percent of its spiral ganglion auditory neurons—is dedicated to a razor-thin frequency band between 82.5 and 83.5 kilohertz. Within this acoustic fovea, thickened strips of the basilar membrane and specialized tectorial resonators act as ultra-high-$Q$ mechanical filters ($Q_{10\text{dB}} > 400$), capable of resolving frequency glints of just 20 hertz caused by the perpendicular stroke of a moth's wing.

If the returning echo drifted up to 85.4 kilohertz during flight, it would miss this high-resolution cochlear band entirely. Schnitzler proved in 1968—and confirmed in 1973 by placing bats inside a sealed wind tunnel filled with a helium-oxygen mixture to alter $c$—that the bat does not calculate airspeed from wind pressure on its skin. Instead, its auditory brainstem measures the frequency of the returning echo from the previous pulse and commands the cricothyroid laryngeal muscles to lower the outgoing pitch $f_{\text{emit}}$ by the exact kilohertz offset needed to keep $f_{\text{echo}}$ locked onto 83.0 kilohertz within 50 hertz. At the same time, two milliseconds before the larynx fires a 115-decibel pulse, the bat's tensor tympani and stapedius muscles contract to clamp the middle-ear ossicles, attenuating self-deafening bone conduction by 25 decibels, then relax two milliseconds after the pulse ends so the cochlea can register the micro-Doppler shimmer of a mosquito's wingbeat three meters away.