Brownian motion

• 1. Introduction

  ›

  Brownian motion is the ceaseless, random jostling of tiny particles suspended in a liquid or a gas. Watch a grain of pollen floating in water under a microscope, and you will see it twitch and dart in no predictable direction, driven by forces invisible to the eye. That observation, made by the Scottish botanist Robert Brown in 1827, gave the phenomenon its name, but it raised a question that would occupy scientists for nearly a century: what, exactly, is doing the pushing? The answer turned out to involve one of the deepest arguments in the history of science, namely whether atoms and molecules are real objects or merely convenient fictions. The path from Brown's microscope to a settled answer runs through a poet writing in ancient Rome, a mathematician who modeled the stock market, a young patent clerk named Albert Einstein, and a Nobel Prize awarded in 1926.

• 2. An Ancient Clue

  ›

  Around 60 BC, the Roman philosopher-poet Lucretius composed a scientific poem called On the Nature of Things. In verses 113 through 140 of Book II, he described what happens when a shaft of sunlight enters a darkened room: a multitude of tiny particles mingle and dance in the beam. Lucretius read that dance as evidence for atoms, writing that the motion "originates with the atoms which move of themselves" and then passes upward through compound bodies until it reaches particles visible to the naked eye. Scholars later noted that the glittering, microscopic jiggling of small dust particles in such a beam is caused chiefly by true Brownian dynamics, even though the broader tumbling of larger dust is driven mostly by air currents. One commentary on Lucretius captured the irony perfectly: he "perfectly describes and explains the Brownian movement by a wrong example." The correct example would have to wait nearly two thousand years for a botanist studying plant reproduction.

• 3. Brown's Observation

  ›

  Robert Brown was not trying to discover a law of physics. He was studying plant reproduction in 1827 when he placed pollen grains from the plant Clarkia pulchella in water beneath a simple microscope. The minute particles inside those grains, on the order of one four-thousandth of an inch in size, exhibited a continuous, jittery motion that he could not explain. To test whether the motion was some form of life force unique to biological material, Brown repeated the experiment with particles of inorganic matter, including glass and rock dust. Those particles moved just as erratically. Life was not the cause. The physical origin of the motion remained unexplained. Brown had identified a real phenomenon, but he did not have the tools, either mathematical or physical, to say what was behind it. That gap would attract two very different minds over the following decades.

• 4. Bachelier and Einstein

  ›

  Louis Bachelier, a French mathematician, was the first to build a rigorous probabilistic model of the random walk that characterizes Brownian motion. He did so in 1900, in a doctoral thesis titled "The Theory of Speculation", prepared under the supervision of Henri Poincare. Bachelier's goal was to analyze the stock and option markets, not physics, but the mathematics he developed for price fluctuations described the same underlying process that Brown had watched under his microscope. His work was largely unknown until the 1950s. Five years after Bachelier's thesis, in 1905, Albert Einstein published a paper explaining the motion of pollen particles as the result of individual water molecules striking them. A Brownian particle undergoes roughly 100 trillion collisions per second, according to Einstein's analysis, far too many for classical mechanics to track individually. He instead derived a diffusion equation showing that a particle's displacement is proportional not to the elapsed time but to the square root of the elapsed time. That relationship gave experimenters something precise and testable, and Einstein used it to calculate the size of atoms and to estimate Avogadro's number, approximately 6.02 mol, the count of atoms in a mole of any gas.

• 5. Perrin and the Atom Debate

  ›

  Einstein's predictions were not immediately confirmed. Early attempts by Theodor Svedberg in 1906 and 1907 were criticized by both Einstein and Jean Perrin as not measuring the right quantities. Victor Henri took cinematographic shots through a microscope in 1908 and found disagreement with the formula, though the analysis remained uncertain. Confirmation finally came in experiments by Chaudesaigues in 1908 and by Perrin in 1909. Jean Baptiste Perrin's meticulous experimental work settled a debate that had split the scientific community for decades: atoms and molecules are physically real. The importance of the result stretched beyond particle physics. Einstein had shown that the kinetic theory's account of the second law of thermodynamics was an essentially statistical law, and Perrin's experiments confirmed that picture. In 1926, the Nobel Committee awarded Perrin the Nobel Prize in Physics for what they described as "his work on the discontinuous structure of matter." The argument that Brown's jittery pollen had once only implied was now settled.

• 6. Smoluchowski's Parallel Theory

  ›

  Marian Smoluchowski developed a theory of Brownian motion starting from the same premises as Einstein and arriving at the same probability distribution for particle displacement. Where the two theories diverged was in their numerical coefficient: Smoluchowski's mean squared displacement came out 64/27 times larger than Einstein's. Arnold Sommerfeld later remarked in a necrology on Smoluchowski that the numerical coefficient differing by 27/64 "can only be put in doubt." The discrepancy traced to different theoretical approaches: Einstein applied Stokes drag to macroscopic drift velocity, while Smoluchowski performed a more detailed kinematic collision analysis but introduced a variance when averaging over the Maxwellian velocity distribution. Smoluchowski also addressed a puzzle that his own model raised: why should a particle be pushed in any net direction when collisions arrive with equal probability from all sides? His answer relied on statistical fluctuations. In a gas, more than 10 to the 16th collisions occur per second; in a liquid, roughly 10 to the 20th per second. At those scales, statistical imbalances are not just possible but inevitable. A mean excess of between 100 million and 10 billion collisions per second in one direction, he argued, could produce an instantaneous velocity for the Brownian particle of anywhere between 10 and 1,000 centimeters per second.

• 7. The Mathematics of Randomness

  ›

  Norbert Wiener gave the first complete and rigorous mathematical treatment of Brownian motion in 1923, and the underlying mathematical object has been called the Wiener process in his honor ever since. The Wiener process is characterized by four properties: it is almost surely continuous; it has independent increments; it starts at zero; and its increments follow a normal distribution. It is one of the best-known Levy processes, a class of stochastic processes with stationary independent increments. The French mathematician Paul Levy later proved a theorem giving necessary and sufficient conditions for a continuous process to count as Brownian motion, a result that serves as an alternative definition of the phenomenon. The Wiener process is recurrent in one or two dimensions, meaning that a particle will return to any neighborhood of its starting point infinitely often, but it is not recurrent in three dimensions or higher. Donsker's theorem, sometimes called the invariance principle, shows that the Wiener process can be constructed as the scaling limit of a random walk, connecting the discrete and continuous pictures of randomness.

• 8. Measuring the Unmeasurable

  ›

  One quantity that Einstein's theory could not easily address was the instantaneous velocity of a single Brownian particle, because his diffusion framework described behavior on timescales much larger than the timescale of individual atomic collisions. That measurement was finally achieved in 2010. Researchers trapped a glass microsphere in air using optical tweezers and measured its instantaneous velocity directly for the first time. The data confirmed the Maxwell-Boltzmann velocity distribution and verified the equipartition theorem at microscopic timescales. The same framework that Lucretius used to argue for atoms in a beam of dusty sunlight now extends to the motion of the supermassive black hole Sgr A* at the center of the Milky Way galaxy: gravitational interactions with surrounding stars give it a Brownian-like velocity predicted to be less than one kilometer per second.