In his new book Meselson, Stahl, and the Replication of DNA, science historian Frederic Holmes recounts the story of what one researcher called “the most beautiful experiment in biology”. The so-called Meselson-Stahl experiment, which was carried out in 1957, confirmed that DNA replicates in the way predicted by the then recently discovered double-helix structure. When Holmes asked five researchers why this particular experiment was so beautiful, their answers included simplicity, precision, cleanness and strategic importance.

Holmes’s book naturally made me wonder about the most beautiful experiment in physics, and what criteria one would use to make the judgment. I asked Holmes which physics experiments he thought beautiful. He replied that he was reluctant to appraise physics, but did mention Helmholtz’s 1850 measurement of the velocity of the nerve impulse, though admitted that this was more in the field of physiology.

Beautiful science

So what makes for beauty in science? In A Mathematician’s Apology, G H Hardy proposes that the essential criteria for beauty in his field are unexpectedness, inevitability and economy – although he also mentions that depth, or how fundamental a proof is, is relevant. I’m also fond of those passages in Michael Faraday’s The Chemical History of a Candle, in which he says that a candle’s beauty is not prettiness of colour or shape, but rather something else: “Not the best-looking thing, but the best-acting thing.” In Faraday’s eyes, a candle taps all of the known laws of the universe.

The heat of the flame melts the wax and draws up currents of air to cool the wax at the periphery, thus creating a cup for the molten wax, which remains horizontal thanks to gravity – “the same force of gravity which holds worlds together”. Capillary action draws the melted wax up the wick from cup to flame, while the flame’s heat triggers a chemical reaction in the wax that sustains the flame.

But what about beautiful experiments? I asked physicist Samuel Devons of Columbia University – an old hand at re-enacting historic physics experiments – for his thoughts. A beautiful experiment has to make an important discovery, he said. “It can’t be a demonstration of what’s in the textbooks, or a check on the theory – it has to change what people knew and believed.” A beautiful experiment must be “not too complicated, not too expensive and not more accurate than it needs to be”. Finally, he told me, it must be within reach of students – a criterion that fails both the otherwise beautiful 1922 Stern-Gerlach experiment demonstrating electron spin, and the 1887 Michelson-Morley experiment on the propagation of light.

Devons’ candidates for beautiful experiments include Cavendish’s work during the 1770s to measure how electric charge distributes itself on a hollow sphere and to measure the force between electric charges. Cavendish found that the force varies with the inverse square of the distance between the charges. Devons’ other choices include Weber and Kohlrausch’s 1856 experiment establishing the relationship between electrostatic and electrodynamic charge. But he also mentioned several series of experiments, by scientists such as Franklin, Faraday and Volta, that he would call beautiful.

My candidates for beautiful experiments include the 1956-7 parity-violation experiment led by Chien-Shiung Wu at Columbia University and colleagues at the National Bureau of Standards in Washington. I would also choose Maurice Goldhaber’s 1957 experiment in which he established that neutrinos have a “negative helicity” – in other words, that their intrinsic angular momentum or “spin” is in the opposite direction to their momentum (see The origin of neutrino mass (summary), pp35-39 print version).

Neither one is a student experiment. But Wu and colleagues overturned at one convincing stroke one of the most fundamental and firmly held assumptions in physics, while Goldhaber’s experiment was so fiendishly ingenious that most physicists at the time did not even think it was possible in principle. In most scientific discoveries, one feels that even if the actual discoverers had missed the boat, the discoveries themselves would still have been made eventually. But this one is different. One physicist later wrote that, had Maurice Goldhaber not existed, “I am not sure that the helicity of the neutrino would ever have been measured”.

I also think of the 1919 British expeditions demonstrating the gravitational bending of starlight, confirming Einstein’s 1915 prediction. But neither the eclipse (a familiar natural event) that made it possible, nor the determination of stellar positions (a familiar astronomical technique), was extraordinary. So can beauty lie solely in an experiment’s dramatic consequences?

Finally, I think of Archimedes, pondering a problem while sitting in a bath. Though today’s science historians are dubious, the ancient source tells us that he noticed that “the amount of water which flowed over by the tub was equal to the amount by which his body was immersed [which] indicated to him a method of solving the problem” – causing him to run through the town yelling with joy. Can an inadvertent discovery transform a routine event into a beautiful experiment?

The critical point

Most discussions of beauty in physics – including Graham Farmelo’s new book It Must Be Beautiful: Great Equations of Modern Science (Top equations add up to beauty, Physics World March p47) – focus almost entirely on beauty’s role in theory and explanation. I find this amazing, even perverse. The term beauty is usually applied to things made by human beings that are material (rather than intellectual and abstract) and that reveal something about nature with clarity, simplicity and depth in a way that transforms our perspective of it. But isn’t this what happens when scientists stage a great experiment?

What, then, is the most beautiful experiment in physics, and what is the connection between its beauty and its scientific value? Is beauty merely a subjective experience, or an objective property of good science – icing on the cake, or essential ingredient? I invite you to submit your candidates and suggestions at the Web site indicated below. In a future column, I shall list the candidates and talk about the connection between beauty and science.

* What is the most beautiful experiment in physics, and what makes it beautiful? Send your answers to Robert P Crease at the address or e-mail given below, or by fax to +1 631 632 7522. You can also enter your response via the Brookhaven National Laboratory Web site at www.bnl.gov/bnlweb/physq/.