Nine Photons in the Dark

1. Probability Geometry: One Photon per Rod

In their classic 1942 study, Selig Hecht, Simon Shlaer, and Maurice Pirenne projected 1-millisecond monochromatic light flashes (510 nm) onto a retinal region 20 degrees temporal to the fovea, where rod density peaks.

Of the 54 to 148 quanta reaching the cornea, 4% reflect off the corneal interface, roughly 50% are absorbed or scattered by ocular media (cornea, lens, and humors), and 80% of the remainder pass between outer segments without striking a chromophore. Ultimately, only 5 to 14 photons (9 on average) are absorbed by 11-cis-retinal molecules in rhodopsin.

For the calculation below, 9 is a representative value within the 5–14 absorbed-photon range. These photons fall across a retinal patch containing approximately 500 rods. Applying the Poisson distribution for an average density of μ = 9 / 500 = 0.018 photons per cell:

In this model, the probability that a rod receives at least two quanta is 0.016%. In a pool of 500 rods, the expected number of double-hit cells is 0.08 (effectively zero). The vast majority of the remaining activated rods receive one quantum.

The result shows that an individual rod can respond to a single quantum, while the estimated threshold for whole-eye perception depends on retinal circuits and signal integration that filter spontaneous thermal isomerizations of rhodopsin.

2. Enzymatic Cascade: From Femtoseconds to Picoamperes

Observe how a single quantum of light converts into a measurable electrical signal across the rod outer segment membrane.

The biochemical amplification unfolds across four sequential steps:

1. Retinal Isomerization (200 fs): The photon strikes 11-cis-retinal bound to lysine 296 in rhodopsin. The C11=C12 double bond rotates in roughly 200 femtoseconds, producing the active Metarhodopsin II (R*) state.
2. G-Protein Activation (Transducin): During its 100-millisecond active lifetime before phosphorylation by rhodopsin kinase, one R* molecule activates transducin successively. Modern experiments in intact mouse rods estimate roughly 12–14 effective transducin–PDE6 complexes per R*. The historical 500–800 transducin figure comes from older models and should not be treated as a general human value.
3. Catalytic cGMP Hydrolysis: Each active transducin–PDE6 complex contributes to the breakdown of cyclic GMP (cGMP) into 5'-GMP. In the same type of rod, the modern downstream estimate is on the order of 2,000 cGMP molecules hydrolyzed per activated R*; historical amphibian models produced much larger values.
4. CNG Channel Closure: The local drop in cGMP causes it to dissociate from cyclic nucleotide-gated cation channels. Closing approximately 1,000 channels blocks the influx of 10,000,000 Na⁺ and Ca²⁺ ions per second, generating a photocurrent of roughly 1.0 picoampere. Baylor, Lamb, and Yau measured this response in toad rods (Bufo marinus) in 1979.

3. Bioenergetic Paradox: Cells Working in Darkness

Most neurons remain at a negative resting potential and depolarize when stimulated. Photoreceptors operate in reverse.

In complete darkness, CNG channels stay open. A steady inward current of 35 picoamperes (the dark current) enters the outer segment, keeping the cell depolarized at −40 mV and maintaining tonic glutamate release at the synapse.

To prevent osmotic swelling and maintain electrochemical gradients, Na⁺/K⁺-ATPase pumps in the inner segment constantly extrude sodium ions. Each rod burns over 73 million ATP molecules per second merely to sustain this waiting state.

The human retina maintains a specific oxygen consumption rate of approximately 300 µmol O₂ per gram per hour — 2.5 times higher than the cerebral cortex. Light does not energize the cell: light acts as an off-switch that halts the current, hyperpolarizes the membrane to −70 mV, and grants the rod a metabolic pause.

4. Dark Adaptation Chronograph: The Kohlrausch Break

Transitioning from sunlight to nighttime darkness proceeds through two distinct biochemical regimes identified by Arnt Kohlrausch in 1922.

During the first 7 minutes, cones rapidly regenerate their iodopsin photopigments, reducing the threshold by 2 to 3 orders of magnitude before plateauing.

Between minutes 7 and 10, the Kohlrausch break occurs: rods, which regenerate rhodopsin far more slowly through the retinal pigment epithelium retinoid cycle, surpass cone sensitivity and assume control of vision. By minute 40, light sensitivity increases by over 100,000-fold relative to the bleached daylight state.

5. How Far Away Can You See a Candle?

A common pop-science myth claims the human eye can see a candle flame from 30 miles (48 km). Strict photonic calculations show atmospheric scattering and the Hecht quantum threshold enforce much tighter boundaries.

A 1-candela flame radiates approximately 1.89 × 10¹⁶ photons per second isotropically. The estimate below uses 9 photons as a representative value within Hecht's range of 5–14 absorbed photons in the retina's 100-millisecond integration window.

In a vacuum, a dilated 8 mm pupil intercepts this threshold at a maximum distance of 9.2 kilometers. In Earth's atmosphere, Rayleigh scattering and aerosol extinction compress this limit to approximately 4.3 kilometers on a clear rural night and 2.9 kilometers in mild urban haze.

Methods & Reference Literature

The quantitative models and biophysical parameters in this analysis derive from peer-reviewed retinal electrophysiology and quantum optics literature:

• Hecht, S., Shlaer, S., & Pirenne, M. H. (1942). Energy, quanta, and vision. Journal of General Physiology, 25(6), 819–840. [Experimental derivation of the 5–14 absorbed quanta threshold].
• Baylor, D. A., Lamb, T. D., & Yau, K. W. (1979). The membrane current of single rod outer segments. Journal of Physiology, 288(1), 589–611. [Photocurrent measured in toad rods, with a unit response of roughly 1 pA].
• Yue, W. W. S. et al. (2019). Elementary response triggered by transducin in retinal rods. Proceedings of the National Academy of Sciences, 116(11), 5144–5153. [Modern amplification estimate in intact rods].
• Stryer, L. (1986). Cyclic GMP cascade of vision. Annual Review of Neuroscience, 9(1), 87–119. [Kinetics of the rhodopsin-transducin-PDE6 enzymatic cascade].
• Hagins, W. A., Penn, R. D., & Yoshikami, S. (1970). Dark current and photocurrent in retinal rods. Biophysical Journal, 10(5), 380–412. [Discovery of the dark current and photoreceptor bioenergetics].
• Ames, A., Barbour, B., & Tolman, K. (1992). Energy metabolism of rabbit retina: role of fatty acids and glucose. Journal of Physiology, 451(1), 383–404. [Measurement of 300 µmol O₂/g·h retinal metabolic rate].
• Kohlrausch, A. (1922). Untersuchungen mit farbigen Schwellenprüflichtern über den Dunkeladaptationsverlauf des normalen Auges. Pflügers Archiv, 196(1), 113–117. [Identification of the rod-cone break].
• Tinsley, J. N. et al. (2016). Direct detection of a single photon by humans. Nature Communications, 7, 12172. [Modern psychophysical validation of single-photon responses].