How this instrument works
A refractive index is a ratio of speeds: n = c ⁄ v, where v is how fast a wavefront advances through a material and c is its speed in vacuum. Water's 1.333 means light crawls across a swimming pool at roughly 225,000 km/s. What survives a boundary untouched is not angle itself but a wave's component along that surface — crests arriving on one side must line up with crests departing beyond it, which forces n·sin θ to match. That single condition, written once per medium, is your whole formula.
This relation is considerably older than its name. Ibn Sahl, working in Baghdad around 984, set it down as a ratio of hypotenuses in a manuscript on burning mirrors and lenses, and used it to shape aplanatic surfaces. Thomas Harriot measured it in England around 1602 and published nothing. Willebrord Snel van Royen rediscovered it at Leiden in 1621 and also left it sitting in his papers, where Huygens later found it. Descartes got a sine form into print first, in La Dioptrique of 1637, which is why French classrooms still call it loi de Descartes. Fermat then derived all of it in 1662 from one premise — light takes whichever route is quickest between two points — and that derivation remains its most satisfying justification, showing why any bending happens at all.
Direction is all you get here. How much light crosses instead of bouncing back belongs to Fresnel, and that result assumes a smooth interface, isotropic non-magnetic media, and one wavelength at a time. Index shifts with colour: crown glass runs about 0.013 higher in blue at 486 nm than in red at 656 nm, so a quoted n is really n at a stated line, conventionally sodium's D line near 589 nm at 20 °C. Calcite splits an arriving ray in two and only half of it complies. Push far enough and this equation simply has no answer — once n₁·sin θ₁ exceeds n₂ every photon turns back, and this sheet says so rather than inventing a number.
- Enter Refractive index, first medium — whatever the light starts in. Vacuum is exactly 1, air 1.000293, water 1.333, crown glass near 1.52, diamond 2.417.
- Set Angle of incidence, measured from the normal (the perpendicular to the surface), never from the surface itself. Switch its unit between degrees and radians as needed.
- Enter Refractive index, second medium, the material being entered. Both indices are pure ratios, so neither carries a unit.
- Read Angle of refraction (deg), returned to four decimals. It tilts toward the normal when the second index is larger, away from it when smaller.
- A total internal reflection warning in place of a number means you are past the critical angle for that pair and no refracted ray exists.
Worked example — a laser into water at 30°
Clamp a green pointer above an aquarium and tilt it until a protractor reads 30° away from vertical, vertical being the normal to a flat water surface. Put 1 into Refractive index, first medium (air, close enough to vacuum at three decimals), 30 into Angle of incidence, and 1.333 into Refractive index, second medium. Angle of refraction (deg) returns 22.0301. Follow the arithmetic yourself: sin 30° = 0.5, divided by 1.333 gives 0.37509, and taking arcsin of that lands on 22.0301 degrees.
Those eight missing degrees explain why a pool floor sits visibly higher than it really is. Rays leaving the water bend away from the normal, your eye traces them back along straighter lines, and depth collapses to about three quarters of true — 2.0 m of water reads as roughly 1.50 m. Spearfishers compensate by aiming beneath the fish they can see. Swap the two indices, feed 22.0301 back as the incidence angle, and 30° comes out again; the path runs just as well in reverse.
Questions
Do I measure the angle from the surface or from the normal?
From the normal — the line perpendicular to the interface. This is far and away the most common mistake with Snell's law, and it quietly substitutes the complement of your answer: a ray 30° from the normal sits 60° from the surface. Grazing incidence therefore means θ₁ close to 90°, not close to zero, and a beam striking dead-on is 0°. If your result bends the wrong way, suspect this before anything else.
What does the total internal reflection warning mean?
No refracted ray exists, so there is no angle to report. It arises only going from higher index to lower — glass into air, water into air — once incidence passes the critical value arcsin(n₂ ⁄ n₁). Water to air gives 48.6°, ordinary glass 41.8°, and diamond a mere 24.4°, which is precisely why a brilliant cut bounces light around inside several times before letting it escape. Optical fibre lives on the same effect.
Does refractive index depend on colour?
Yes, and that dispersion is what makes prisms and rainbows possible. Index drops as wavelength grows across the visible band, so blue bends harder than red — crown glass sits near 1.5240 at 486 nm against 1.5110 at 656 nm. Tabulated values assume a reference line unless stated, usually sodium's D line at 589.3 nm and 20 °C, written as n-D-20. Mix figures measured at different lines and expect disagreement in the third decimal.
How much light actually crosses the boundary?
Snell's law is silent on that; it fixes direction and nothing more. Dividing energy between transmitted and reflected beams is the job of the Fresnel equations, which also care about polarisation. For a rough anchor, an air-to-glass face near normal incidence reflects about 4% and passes 96%, while the reflected fraction climbs steeply beyond 60° until a surface viewed at a glancing angle behaves like a mirror. Wet roads at night demonstrate it nightly.
Where do people measure refractive index in practice?
On a refractometer, typically an Abbe instrument reading a critical angle off a thin film of sample. Winemakers read sugar in grape must with one, beekeepers check honey moisture, machine shops verify cutting-fluid concentration, and clinical labs estimate urine specific gravity. Analytical chemists quote index beside melting point as an identifying constant, since a pure liquid reproduces its value reliably to four decimals.
Can an index be below 1, or negative?
Below 1 happens routinely. That means phase velocity above c, which carries no information and violates nothing — X-rays see most solids at around 0.99999, and grazing-incidence X-ray telescope mirrors exploit exactly that. Negative index occurs nowhere in nature. Engineered metamaterials, predicted by Veselago in 1968 and demonstrated at microwave frequencies in 2001, bend a beam to the wrong side of the normal entirely. This sheet assumes ordinary positive media and requires both indices above zero.