Begin with two numbers and the different routes by which they arrived.
One came from electrical measurements. The other came from experiments on light: send it over a known distance and devise a way to determine how fast it travels.
In the early 1860s, James Clerk Maxwell used the electrical result in his theory and obtained a propagation speed. It was close to an independently measured speed of light, both around 310,000 kilometers per second.
Neither was today's precise value. Nor did the records agree to their final digit. Measurements were still improving. What invites attention is not numerical perfection but the meeting of two lines of inquiry: electricity and magnetism on one side, light on the other.
Why should they meet here?
More than a familiar number
Coincidences are easy to find. A birthday matches a house number; two unrelated bills add up to a pleasing total. We notice, perhaps smile, and learn nothing further about their connection.
Maxwell's result did more. The theory described an electromagnetic disturbance that could propagate, and under the relevant conditions it specified the speed of propagation. The number was not selected afterwards from a collection of possible answers because it resembled the speed of light.
Imagine following the arrangement of things in one room and discovering that it leads to a door into another. You have not decided to visit the second room and then drawn a convenient doorway on the wall. Something in the first room's relations has given you a route beyond it.
The image must stop there. An imaginary door is easy to supply. The theoretical route must actually be worked through, and the experimental result must have an accountable history.
The agreement supported more than a resemblance between numbers. It supported the proposal that light itself was an electromagnetic phenomenon. What had seemed separate subjects were beginning to disclose a common relation.
Today we can reduce the episode to a fact to be memorized: light is an electromagnetic wave. Before moving on, however, consider what was not guaranteed. Light had no obligation to travel at that speed merely because the relations between electricity and magnetism had been elegantly expressed. The world could have failed to cooperate.
What happened to the little wheels?
The path did not initially look like a modern textbook.
Maxwell had used a mechanical picture to help him think: rotating vortices, with smaller wheels transmitting motion between them. It can look peculiar now, an attempt to make an unfamiliar domain accessible through familiar kinds of mechanism.
In his electromagnetic-field paper published in 1865, the argument no longer required that particular machinery. The connection between electromagnetic propagation and the speed of light remained. This should not be retold as though he had already abandoned the ether. He still thought in terms of a medium and motion. Moving beyond one specific model was not the instantaneous arrival of every later physical idea.
The qualification makes the story more interesting, not less.
A discovery does not lose its force because it needed scaffolding. Neither does useful scaffolding acquire a right to remain forever. Some parts of an explanation help us imagine what we cannot yet handle; some relations can survive substantial changes in that explanation.
The beauty may not belong to the first picture. It may lie in being able to change the picture without losing a relation that has earned its place.
The world is allowed to disagree
The easiest way to spoil this story is to turn it into one about a genius who recognized beauty and was therefore right.
Beauty offers no such warranty.
However rigorous the relations on paper, the paper has not signed on the world's behalf. When experiment does not agree, we must examine the derivation, the measurement and the conditions under which they were compared. We must also permit the theory to change. We cannot always blame a disagreement on the world having been measured badly.
Conversely, a mismatch does not instantly overthrow everything. Instruments malfunction. Calculations contain mistakes. A result can be used outside its proper range. Listening carefully means distinguishing these possibilities, not choosing the most satisfying slogan about falsification.
Waiting for an answer is therefore not merely hoping for a yes. It means exposing a claim to inspection in a way other people can retrace. What did it specify? How was the measurement made? How large is the discrepancy? Under what conditions is the comparison meaningful?
This is also why beauty here is not simply the pleasure of a successful guess. Guessing the number of sweets in a jar can be gratifying. It need not reveal anything about their relations. In Maxwell's case, something the theoretical relations required met a response along an independent experimental route.
Not “I can always explain you,” but “I made a definite claim here, and here you answered.”
The answer need not always be yes
What if the results had failed to agree?
There might still be something worth attending to, but that is no reason to call every failure beautiful. Losing the data, neglecting calibration, or guessing repeatedly without investigating the errors may leave very little to look at.
A significant disagreement can have a definite contour. It keeps appearing in one place. Relations that remain well supported elsewhere help reveal the shape of what is missing. The question is no longer the vague admission that we do not know everything. It is why this particular part will not fit.
A new answer has not arrived, yet the old one has not become useless everywhere. Work continues between those facts. What is worth seeing may be the increasingly precise boundary, not the beauty of a claim already shown to be false.
Nor must Maxwell's successful encounter be justified by all the unanswered questions that followed it. Two independent routes did meet. That achieved relation remains worth revisiting. Being established is not the same as being dead.
Some people learn the equations for an examination. Others spend their working lives calculating with them. Another reader may follow only this account of how the two routes met. These are different degrees of access. Reading a story is not equivalent to mastering a derivation, and we should not pretend otherwise. But the difference need not exclude a nonspecialist from every part of what happened.
We can begin by recognizing why “the theory is beautiful” cannot be the last sentence.
The speed written on one page still had to be set beside the reading from somewhere else. However well the page was written, it could not provide the instrument's answer.
And the two could agree. That is the encounter worth pausing over.