The Aether — The Experiment That Found Nothing

In the first article phlogiston was a theory held the wrong way round. This one is different. The aether was not reversed; it simply was not there. And the way its absence was established became the most famous scene in the history of experiment.

Wave or particle

That question comes first. If light is a stream of tiny particles, no medium is needed — a bullet does not need air to fly. If it is a wave, everything changes, because a wave is something oscillating, and there has to be a something.

Huygens took light to be a wave; Newton took it to be particles. Newton's authority was such that the particle account held the field for over a century.

In 1801 Thomas Young passed light through two narrow slits. What appeared on the wall was not two bright spots but a set of light and dark bands — a pattern particles cannot make.

What the aether had to be

Here is the heart of the story. The aether was not a word waved at a problem. Physicists worked out, with great care, what such a substance would have to be like.

1. It must fill all space without gaps, since starlight reaches us from every direction.
2. It must not impede the planets at all: the Earth has orbited for billions of years without slowing.
3. Light is a transverse wave, oscillating across its direction of travel — and transverse waves pass only through rigid solids.
4. So the aether must be thinner than air and stiffer than steel.

Maxwell sharpens the question

In 1865 Maxwell bound electricity and magnetism into one set of equations. Solving them gave a wave, and the equations delivered its speed: about three hundred thousand kilometres a second — the already measured speed of light.

So light is an electromagnetic wave. But the same equations left an awkward question, because that speed came out as a single number.

That yields a testable prediction. The Earth orbits at thirty kilometres a second. If the aether is at rest, we are standing in an aether wind — and light running with the wind should differ in speed from light running across it.

Swimming across the river, and up it

In 1887 Michelson and Morley set out to measure exactly that. The difficulty was that the Earth's speed is a ten-thousandth of light's, so the effect appears at the level of one part in a hundred million — beyond any clock then existing.

So they used a method that measures no time at all. They set light against light.

1. Split one beam of light in two with a half-silvered mirror.
2. Send one along the motion and one across it — equal distances at right angles — and bounce both back.
3. Recombine them. Any difference in travel time shows up as interference fringes.
4. Now rotate the whole apparatus by ninety degrees. The two arms swap roles, so the fringes must shift.

They floated the apparatus on a stone slab in a bath of mercury, because a footstep in the building disturbed the reading. They turned the slab slowly and measured, then measured again in the season when the Earth's motion had reversed.

The fringes did not move.

A little further in

Rescuing the theory

The aether was not dropped at once. George FitzGerald and Hendrik Lorentz arrived independently at the same idea: a body moving through the aether contracts along its direction of travel.

The amount is exquisite. The path shortens by precisely what the light loses, so the two effects cancel and nothing shows.

Same equations, different world

In 1905 a patent examiner in Bern made no attempt to rescue the aether. He inverted the question.

What if light is measured at the same speed by everyone — not as an oddity to be explained, but as a starting point?

Starting there, Lorentz's contraction came out unchanged. The reason was different. Nothing was being squeezed by an aether; time and length simply behave that way for anything in motion.

A widespread story

Textbooks often put it like this: Michelson and Morley failed, and so Einstein produced relativity.

Tidy, but not accurate. The 1905 paper does not cite that experiment by name. Its opening paragraph is gripped by a quite different problem — that moving the magnet and moving the conductor give the same current, while the theory of the day explained the two cases in entirely different ways. Einstein's later recollections of how much the experiment mattered to him were not consistent.

What it left

There is no aether. The century spent hunting it left a good deal behind.

The third is the best of them. LIGO, which first detected gravitational waves in 2015, is a Michelson interferometer: a beam split in two, sent down two arms at right angles, returned and recombined — the design of 1887, with the arms grown to four kilometres.

1. 1690Huygens — treats light as a wave
2. 1801Young — settles the wave account with interference fringes
3. 1865Maxwell — shows light is an electromagnetic wave
4. 1887Michelson and Morley — fail to find the aether wind
5. 1892Lorentz and FitzGerald — try to save the theory with length contraction
6. 1905Einstein — makes the aether unnecessary
7. 2015LIGO — detects gravitational waves with the same design of interferometer

The question that remainsMichelson and Morley counted their experiment a failure, because nothing came of it. Experiments that find nothing still mostly go into a drawer rather than into print. The science we read is a collection of the ones that found something — so what is in the drawers?