The Biological Bolometer

Theodore Bullock, Pit Vipers, and Thermal Membrane Imaging

Inside an experimental animal room at the American Museum of Natural History in New York in 1936, herpetologists Gladstone Koffler Noble and Arthur Schmidt set out to isolate how crotaline snakes—rattlesnakes, copperheads, and moccasins—strike warm-blooded rodents in pitch darkness. One by one, they eliminated the known reptilian senses of timber rattlesnakes (Crotalus horridus) and prairie rattlesnakes (Crotalus viridis). They covered the snakes' eyes with strips of opaque black adhesive tape, plugged their nostrils with cotton and collodion, and amputated or sealed the forked tips of their tongues to cut off chemical delivery to the vomeronasal Jacobson's organ in the roof of the mouth. Yet when Noble and Schmidt swung a freshly killed rat suspended on a string past the cage, the blind, deaf, anosmic snake coiled and struck the carcass squarely in midair.

Next, Noble and Schmidt replaced the dead rat with two identical glass lightbulbs wrapped in black cloth so no visible light could escape. One bulb was left cold at room temperature; the other was warmed internally by an electrical resistor to 37 degrees Celsius. Whenever the warm cloth-wrapped bulb swung within twenty centimeters of the snake's snout, the rattlesnake struck it with millimeter precision, whereas it ignored the cold bulb completely. Only when the researchers plugged the two deep cavities situated between the eye and the nostril on each side of the viper's face—the loreal pits, which anatomists since Edward Tyson in 1683 had dismissed as auditory passages or tear ducts—did the strikes cease.

Sixteen years later, in 1952 at the University of California, Los Angeles, neurophysiologist Theodore Holmes Bullock and herpetologist Raymond Cowles placed microelectrodes directly onto single myelinated fibers of the ophthalmic and maxillary branches of the trigeminal nerve innervating the loreal pit of rattlesnakes. In total thermal uniformity, the trigeminal fibers were not silent; they fired a steady, rhythmic background stream of 15 to 25 action potentials per second. When Bullock held a neutral glass rod two degrees cooler than ambient room air thirty centimeters in front of the pit, the background firing stopped instantly. When he brought a human hand or a warm water tube thirty centimeters before the pit—delivering a long-wave infrared flux peaking near 10 micrometers wavelength that raised the temperature of the pit tissue by just 0.003 degrees Celsius—the trigeminal nerve fiber surged to 60 to 100 spikes per second within fifty milliseconds.

Unlike vertebrate rods and cones, which rely on a single high-energy visible photon (1.8 to 3.1 electron-volts) to isomerize 11-cis-retinal in rhodopsin, a 10-micrometer thermal infrared photon carries only 0.12 electron-volts of energy—far too weak to break or rearrange a covalent bond above thermal background noise. The pit viper therefore cannot detect individual infrared photons photochemically. Instead, its loreal pit operates as a biological bolometer, absorbing broadband thermal radiation issued according to the Stefan-Boltzmann law and converting the resulting micro-kelvin tissue heating into electrical spikes:

For nearly sixty years after Bullock's recordings, the molecular identity of the ion channel that converts a one-millikelvin temperature step into a depolarizing current remained unknown. In 2010 at the University of California, San Francisco, physiologist David Julius and postdoctoral fellow Elena Gracheva solved the puzzle by sequencing the transcriptome of rattlesnake and python trigeminal ganglia and comparing it with non-pit snakes. Whereas dorsal root ganglia throughout the snake's body express normal pain and temperature receptors, the trigeminal neurons projecting to the facial pit showed a 400-fold enrichment of a single transient receptor potential channel: TRPA1. In humans and mice, TRPA1 is the chemical irritant receptor activated by mustard oil, tear gas, and wasabi, with no sensitivity to gentle warmth. In pit vipers (and independently in boid snakes with labial pits), evolutionary amino-acid substitutions in the N-terminal ankyrin repeat domains shifted the thermal activation threshold of TRPA1 down to 28.0 degrees Celsius with an ultra-steep thermodynamic temperature coefficient ($Q_{10} > 30$).

Why doesn't the snake's own body heat blind its bolometer, or leave lingering thermal afterimages across the receptor sheet? Electron microscopy of the loreal organ by Richard Goris and colleagues revealed an engineering design identical to cooled aerospace infrared focal-plane arrays. Rather than resting against the solid, heat-sinking bone of the skull, the sensory epithelium—packed with flattened terminal nerve masses (TNM) rich in mitochondria—is suspended in midair as a drumhead membrane just 15 micrometers thick, stretched between an outer chamber open to the night air and a sealed inner air cavity vented by a microscopic pore to equalize atmospheric pressure. Because air has twenty-five times lower thermal conductivity than water or bone, virtually zero absorbed infrared energy leaks into the skull; the entire radiative flux goes into heating the 15-micrometer membrane. Simultaneously, an intricate mesh of capillary blood vessels threading directly between the terminal nerve masses acts as a liquid-cooled radiator, flushing absorbed heat away within 50 to 100 milliseconds so that a running mouse leaves a crisp moving trajectory rather than a blurred thermal streak across the optic tectum.