Saramsha

No. 063  ·  18 November 2026

The Angle Every Drop Agrees On

A rainbow isn't one beam of coloured light — it's sunlight passing through millions of individual raindrops, each one bending it by a slightly different amount depending on exactly where the light enters it. Drag the entry point and watch most exit angles scatter wide, except for one, where they all pile up at once. That pile-up is the rainbow.

red (n 1.331) violet (n 1.344) the ray you're dragging

About this piece

No single ray makes a rainbow. A pile-up of them does

Picture one spherical raindrop hit by a single ray of sunlight, entering somewhere between its centre line and its very edge. The ray refracts going in, bends toward the normal because it's slowing down in water; crosses the drop; reflects once off the inside of the far wall, the way light always partly reflects off any boundary; then refracts again on the way back out, speeding up and bending away from the normal. The total angle by which that ray has been turned from its original direction — its deviation — depends on exactly where it entered. A ray that grazes near the centre line comes almost straight back. A ray that clips the very edge barely turns at all. Somewhere in between, it turns out, the deviation doesn't change smoothly the whole way: it dips to a shallow minimum around 138° and climbs back up on either side. Near that minimum, a whole range of different entry points all produce almost exactly the same exit angle — mathematically, the deviation's rate of change passes through zero there, a textbook calculus turning point known in optics as a caustic.

That bunching is the entire rainbow. Away from the minimum, light leaving a raindrop fans out thinly across many directions and is too faint, from any one angle, to see. Right at the minimum, rays that entered at measurably different points all leave travelling in essentially the same direction, so the light intensity at that one angle is far higher than anywhere nearby. Flip the deviation around — measured instead from the point directly opposite the sun, the antisolar point, which is where your own shadow's head would fall if the ground were a screen — and that pile-up angle becomes 180° − 138° ≈ 42°. Every raindrop in the sky at 42° from that point, around a full circle, is having its own private pile-up pointed straight at your eye. Water bends red light slightly less than violet light, so the two colours pile up at slightly different radii — about 42.5° for red, 40.6° for violet — which is the entire reason the rainbow has a spectrum at all, spread thinly from red on the outside to violet on the inside.

A second bow, backwards

Some light reflects twice inside a drop instead of once before it escapes. That costs it extra brightness — each reflection leaks some light straight through instead — which is why the secondary rainbow, outside the first at about 51°, is always fainter. The extra bounce also reverses the order the colours exit in, so the secondary bow shows violet on the outside and red on the inside, the mirror of the primary. Because light can't pile up anywhere in the gap between 42° and 51° — no entry point for either path produces a deviation in that range — the sky there is measurably darker than the sky on either side of it, a band known since the 3rd century as Alexander's dark band, named for Alexander of Aphrodisias, who described it without ever explaining it.

Theodoric of Freiberg worked out the one-reflection, two-refraction path with a water-filled glass sphere around 1304; Descartes repeated the experiment and the trigonometry in 1637 and found the 42° minimum by brute calculation, angle by angle, without calculus. It took Isaac Newton's prism work shortly after to explain why the angle differs by colour at all — that white sunlight is already a mixture of colours bent by very slightly different amounts, not a single thing that somehow turns into several.

The scene traces real geometric optics for a spherical drop with red modelled at refractive index 1.331 and violet at 1.344 (true values vary continuously with wavelength; these bracket the visible range). It assumes drops large enough — much bigger than the wavelength of light, true for anything above a fine mist — for this ray picture to hold at all; very small droplets produce a colourless fogbow instead, because diffraction takes over and blurs every colour's angle into the next. The sky panel places the rainbow at a fixed 42° (primary) and 51° (secondary) from the antisolar point and ignores the very slight broadening a real bow gets from the sun's own disc not being a true point source.