Chasing light: measuring, mirroring and bending it on purpose

Discover gave you the facts. Understand gave you the rules. This layer asks you to act like the people who worked those rules out in the first place: predict, test, and see whether the prediction survives.

You will chase a genuinely hard historical question — how do you clock something that seems to arrive everywhere at once? — and then turn detective on curved mirrors, lenses, the bent-straw trick taken to its limit, and the rainbow. Every chapter starts with a prediction. Make it before you read on.

For most of history, nobody could tell the difference between "light arrives instantly" and "light arrives so fast you cannot measure the delay." Galileo tried in 1638: two people with covered lanterns stood on hilltops a few kilometres apart, and each uncovered their lantern the instant they saw the other's light. He found no measurable delay at all — not because light is instant, but because even at hilltop distances the delay is a few millionths of a second, hopelessly too small for a human with a lantern and a pulse to catch.

The first real evidence came not from a lantern but from a moon of Jupiter, watched over years.

Rømer's method needed the whole Solar System as a stopwatch. In 1849, the French physicist Hippolyte Fizeau found a way to measure light's speed using equipment that fit on a hillside near Paris.

He shone a beam through a gap in the teeth of a fast-spinning toothed wheel, sent it 8,633 metres to a mirror on a distant hill, and back through the same wheel. If the wheel had turned by exactly one tooth's width while the light made its round trip, the returning beam would be blocked by the next tooth instead of passing back through the gap it left by.

Spin the wheel just fast enough for the light to vanish behind the very next tooth, and you can calculate the round-trip time from the wheel's speed alone.

Every mirror you have met so far has been flat. Curve the reflecting surface and the law of reflection still applies at every single point — but because the surface tilts differently from point to point, the reflected rays no longer stay parallel. They can be gathered together or spread apart, and that changes everything about the image.

There are two shapes. A concave mirror curves inward, like the inside of a spoon, and can gather rays to a real point called the focus. A convex mirror curves outward, like the back of the spoon, and always spreads rays apart.

A lens bends light by refraction instead of reflection, at both of its curved surfaces. A convex (converging) lens is thicker in the middle and bends parallel rays inward to a focus, just like a concave mirror bends them by reflection. A concave (diverging) lens is thinner in the middle and always spreads rays apart, exactly mirroring what a convex mirror does.

The formula is identical in form to the mirror formula you just used: 1/v = 1/f − 1/u. Only the geometry — refraction through glass instead of reflection off a coated surface — is different.

Your own eye is a living example. Light entering it is bent mostly by the curved front surface, the cornea, with the flexible internal lens doing the fine adjustment. Together they act as a single converging lens system with a power of about 59 dioptres — roughly a focal length of 17 mm — squeezing a real, upside-down image of the world onto the light-sensing retina at the back of the eyeball. Your brain, entirely used to this, simply learns to treat that upside-down signal the right way up; nothing in the eye itself ever flips it back.

One lens can magnify. Put a second lens in the path and you can multiply the effect — which is exactly how both a microscope and a refracting telescope work, using nothing more exotic than two convex lenses lined up on the same axis.

The first lens (the objective) forms a real image of the object. The second lens (the eyepiece) is then used as a simple magnifying glass to examine that image, rather than the original object. The two magnifications multiply together.

You already know a ray bends away from the normal when it leaves a slow material (like water or glass) for a faster one (like air). Now push that idea harder: what if the ray hits the boundary at a very steep slant?

As the angle inside the water or glass increases, the refracted ray outside bends further and further from the normal — until, at one particular angle, the refracted ray would have to bend a full 90°, skimming exactly along the surface. Push past that angle and refraction simply stops working: no ray can escape at all, and instead all of the light reflects back inside, as if the boundary had become a perfect mirror. This is total internal reflection, and the angle where it begins is the critical angle.

You have already dispersed white light with a prism into a spectrum, always violet-to-red across a flat band. A rainbow does something stranger: it forms a curved arc, always at the same angle from the point directly opposite the Sun, and it needs no glass at all — only raindrops.

White sunlight is a mix of every visible wavelength, from violet (about 400 nanometres) to red (about 700 nanometres). Air is almost — but not quite — perfectly transparent: its molecules, far too small to see, still nudge passing light a little off course, an effect called scattering. The amount of scattering is ferociously sensitive to wavelength: it goes as 1 ÷ wavelength⁴, so a wavelength that is only slightly shorter scatters far more than one slightly longer.