The Rutherford Gold Foil Experiment Explained

In 1909, at the University of Manchester, physicist Hans Geiger and undergraduate student Ernest Marsden fired a beam of alpha particles at a sheet of gold foil only a few hundred atoms thick. Almost every particle passed straight through. But roughly 1 in 8,000 bounced backward, and that small, stubborn anomaly forced physicists to throw out their entire model of the atom and replace it with the nuclear structure taught in classrooms today.

The atom everyone believed in before 1909

The plum pudding model — J.J. Thomson's 1904 description of the atom as a diffuse ball of positive charge with electrons scattered through it like fruit in a pudding — was the accepted picture of atomic structure at the start of the twentieth century. Thomson had discovered the electron in 1897 at Cambridge's Cavendish Laboratory, proving that atoms were not the indivisible, featureless spheres that chemists had assumed since John Dalton's atomic theory a century earlier (Thomson, Philosophical Magazine, 1897).

Thomson's follow-up model tried to explain how a negatively charged particle could exist inside an atom that was, overall, electrically neutral. His answer was that positive charge filled the entire volume of the atom uniformly, like a cloud, with electrons embedded inside it at rest or in simple orbits. Crucially, nothing in this picture was concentrated. Positive charge, in Thomson's atom, was everywhere and nowhere in particular, spread as thin as fog across the whole diameter of the atom.

This mattered enormously for what came next, because a diffuse, evenly spread charge can only nudge a passing charged particle gently. It has no way to deliver a sharp, forceful deflection, since there is no location inside the atom where the electric field is especially strong. Any experiment that measured how much objects bounced off atoms was, in effect, a direct test of whether the plum pudding model's core assumption, uniform charge distribution, was actually true.

Ernest Rutherford was not testing this by accident. Rutherford had already won the 1908 Nobel Prize in Chemistry for identifying and classifying alpha and beta radiation and for his work on the transformation of elements during radioactive decay, and by 1909 he headed the physics laboratory at the University of Manchester, where he had access to a radioactive source strong enough to produce a usable beam of alpha particles (Nobel Prize in Chemistry 1908, Nobel Foundation). He knew alpha particles were positively charged and comparatively heavy, roughly 7,300 times the mass of an electron, which made them a useful probe: fire enough of them through matter and their scattering pattern would reveal how charge is actually arranged inside the atoms they pass.

Who actually ran the gold foil experiment

Rutherford designed the research question and interpreted the results, but he did not personally sit at the apparatus counting scintillation flashes for months on end. That work was carried out by Hans Geiger, a German physicist then working as Rutherford's research assistant, and Ernest Marsden, a 20-year-old undergraduate physics student whom Rutherford had assigned to the project as a training exercise. Geiger would go on to invent the Geiger counter for detecting radiation; Marsden would later have a lunar crater and an asteroid named after him, in recognition of a discovery he made as a student barely out of his teens.

Geiger and Marsden's apparatus was simple in concept but demanding in practice. A sealed glass tube of radon gas, a radioactive source, emitted alpha particles that were collimated into a narrow beam and aimed at a thin sheet of gold foil, chosen because gold can be hammered into sheets only a few hundred atoms thick without tearing. Behind the foil sat a rotating microscope attached to a screen coated in zinc sulfide, a compound that emits a tiny flash of light, called a scintillation, whenever a charged particle strikes it. To count deflected particles, one of them had to sit in a darkened room and personally watch for faint flashes through the eyepiece, tallying them by hand and by eye, adapted to darkness for twenty minutes or more before a session began.

Geiger and Marsden's original counting method had no electronics at all: a researcher sat in total darkness, eyes adapted for roughly 20 minutes, and counted individual light flashes on a zinc sulfide screen through a microscope, sometimes for hours at a stretch, to build the statistical picture that ultimately overturned the plum pudding model.

What the experiment predicted versus what it found

If Thomson's plum pudding model were correct, the expected outcome was straightforward: alpha particles should pass through gold foil with only small, statistically distributed deflections, since no region inside the atom carried a concentrated enough charge to produce anything more dramatic. Geiger and Marsden set out in 1908 to measure exactly this, publishing their first results the same year in the Proceedings of the Royal Society, confirming that the overwhelming majority of particles did pass through with only slight scattering, matching expectations (Geiger, Proceedings of the Royal Society A, 1908).

Rutherford then asked Marsden, as something close to a routine follow-up rather than a targeted search for anomalies, to check whether any alpha particles scattered at large angles, angles well beyond what a diffuse charge distribution should be able to produce. This was, by Rutherford's own later account, a low-expectation assignment. In 1909, Marsden reported back with a result nobody had anticipated: a small but consistent fraction of alpha particles, later measured at roughly 1 in 8,000, deflected by more than 90 degrees, and a handful bounced back in almost the exact direction they had come from (Geiger and Marsden, Proceedings of the Royal Society A, 1909).

The table below sets the two competing predictions side by side against what the detector actually recorded.

Model	Predicted charge distribution	Predicted alpha scattering	What Geiger and Marsden measured
Thomson's plum pudding (1904)	Positive charge spread evenly through the whole atom	Small deflections only, no large-angle bounces	Contradicted: about 1 in 8,000 particles deflected over 90 degrees
Rutherford's nuclear model (1911)	Positive charge concentrated in a tiny central nucleus	Most particles pass through empty space; rare direct hits bounce back sharply	Matched: large majority passed through, rare fraction bounced back at large angles

The gold foil data ruled out uniform charge distribution and matched a model where nearly all of an atom's mass and charge sits in a minuscule central point.

Why couldn't the plum pudding model explain the bounce-back?

A uniformly distributed positive charge simply cannot generate an electric field strong enough, at any single point, to reverse the direction of a fast, heavy alpha particle. The repulsive force between two charges grows sharply as the distance between them shrinks, so a large deflection requires the alpha particle to pass extremely close to a small, concentrated source of charge, something Thomson's spread-out model did not contain anywhere.

Rutherford's famous reaction and the two-year gap before publication

Rutherford did not immediately publish a new theory. He spent roughly two years working through the mathematics needed to describe how a charged particle scatters off a concentrated point charge, calculations that eventually became the foundation of what is now called Rutherford scattering. He described his own initial reaction to Marsden's finding in a quote that remains one of the most repeated lines in the history of physics: "It was almost as incredible as if you fired a 15-inch shell at a piece of tissue paper and it came back and hit you" (as recorded in E. N. da C. Andrade, Rutherford and the Nature of the Atom, 1964, p. 111).

The image works because a 15-inch naval shell, a wartime-scale artillery round roughly 380 millimeters (15 inches) in diameter, is heavy and moving fast enough that nothing in a sheet of tissue paper should be able to stop it, let alone send it back the way it came. Alpha particles carry only a fraction of that momentum, but relative to the electric fields Thomson's model predicted inside an atom, the analogy captures the scale of the surprise precisely: something in the gold foil was acting less like tissue paper and more like a wall.

By 1911, Rutherford had worked out that the data fit a model in which almost all of an atom's mass and its entire positive charge are packed into an extremely small volume at the center, a region he named the nucleus, while the electrons occupy the comparatively enormous, mostly empty space around it (Rutherford, Philosophical Magazine, 1911). To put the scale in perspective: if an atom were expanded to the size of a football stadium, the nucleus would be roughly the size of a housefly sitting at the center, yet it would still contain essentially the atom's entire mass.

Why the scattering pattern only makes sense with a nucleus

The nuclear model explains the Geiger–Marsden data because it predicts exactly two populations of outcomes, and no others, which is precisely what the detector recorded. Most alpha particles travel through the largely empty volume of an atom without ever passing close to the nucleus, so they continue on with little to no deflection. A very small number happen to travel almost directly at a nucleus, and because the nucleus concentrates all of the atom's positive charge into a volume roughly 100,000 times smaller than the atom itself, the repulsive electric force at close range is powerful enough to reverse the particle's path entirely.

This is a useful place for an analogy. Imagine standing in an otherwise empty football stadium and throwing tennis balls at where you're told a brick sits somewhere on the field. If the brick is spread out as fine dust across the entire field, as Thomson's model implied, every ball lands with a light dusting of resistance and none bounce back. But if the brick is actually sitting, intact, in one small spot, almost every ball misses it completely and sails through untouched, while the rare ball that hits it dead-on ricochets straight back at you. The gold foil data is exactly that second pattern: overwhelming pass-through, rare, sharp ricochet, nothing in between at meaningful rates.

Rutherford's nuclear model was refined within a few years by his student Niels Bohr, who joined Rutherford's Manchester group in 1912 and went on to add quantized electron orbits to explain why atoms are stable and why they emit light at specific, discrete wavelengths rather than collapsing or radiating continuously, work covered in more depth in Niels Bohr's biography. Bohr's contribution addressed a gap Rutherford's original model left open: under classical physics, an orbiting electron should lose energy and spiral into the nucleus in a fraction of a second, which plainly does not happen. The relationship between those quantized electron orbits and the light atoms emit is explored further in what is a photon, explained.

How does the nuclear model compare to the models that came before and after it?

Thomson's 1904 model spread charge evenly through the atom; Rutherford's 1911 model concentrated positive charge and mass into a tiny central nucleus with mostly empty space around it; Bohr's 1913 model kept Rutherford's nucleus but restricted electrons to fixed, quantized orbits, explaining atomic stability and spectral emission lines that Rutherford's version alone could not.

From Manchester to modern particle physics

The gold foil experiment is frequently cited as the first modern scattering experiment, meaning the first case where physicists inferred the internal structure of something too small to see directly by firing particles at it and analyzing how those particles deflected statistically. This method, sometimes called deep inelastic scattering in its later, higher-energy forms, remains the working principle behind particle accelerators today. When physicists at CERN's Large Hadron Collider fire protons at each other and analyze the debris to hunt for particles like the Higgs boson, they are running a vastly scaled-up version of the same basic logic Geiger and Marsden used in a darkened Manchester laboratory: send something in, measure how it comes back out, and infer what's inside from the pattern.

The precision required for the original experiment is easy to underestimate today. Geiger and Marsden had no photomultiplier tubes, no digital counters, and no computer-assisted statistics. Every scintillation flash was tallied by a human eye trained to detect faint light after prolonged dark adaptation, and the statistical significance of a roughly 1-in-8,000 event rate had to be established purely through repeated, exhausting manual counting sessions. That the resulting data was precise enough to support a specific mathematical model of nuclear scattering, one that still holds up under modern quantum mechanical treatment, says as much about the rigor of Geiger and Marsden's methodology as it does about Rutherford's theoretical insight.

It is also worth being precise about what the gold foil experiment did not establish. It did not reveal what the nucleus is made of, since protons were not identified as nuclear constituents until Rutherford's own later experiments in 1917 to 1919, and neutrons were not discovered until James Chadwick's work in 1932 (Chadwick, Proceedings of the Royal Society A, 1932). Rutherford's 1911 nucleus was, at that stage, simply a concentrated point of positive charge and mass, without an internal structure of its own. Filling in that structure took another two decades of separate experiments.

Why this counts as one of the cleanest anomalies in physics history

What makes the gold foil experiment such a durable teaching example is how little drama surrounded its discovery. Rutherford was not hunting for a way to disprove Thomson, his own former mentor at the Cavendish Laboratory. He assigned Marsden a routine measurement, expected a negative result, and got one that could not be explained away, dismissed as instrument error, or ignored, because the effect was small but statistically unmistakable and reproducible on repeated runs.

This is a pattern that recurs across the history of experimental physics: a widely accepted model survives for years not because every possible test has confirmed it, but because nobody has yet run the specific test that exposes its blind spot. The Michelson-Morley experiment, conducted more than two decades earlier, followed a similar arc, expecting to detect Earth's motion through a hypothetical "ether" and instead finding nothing, an anomaly that eventually helped clear the way for Einstein's special relativity. Robert Millikan's oil drop experiment, run in the same general period as Rutherford's work, followed the opposite pattern, confirming rather than overturning existing theory, by directly measuring the charge of a single electron and showing that electric charge comes only in fixed, indivisible units.

Rutherford's own summary of the discovery, given decades later in a lecture, captured how unplanned the result really was: he had not set out that year to disprove Thomson's atom, only to measure scattering angles as a matter of ordinary laboratory practice. The nucleus, in other words, was found because Rutherford's team took a small, easily dismissed anomaly seriously rather than writing it off as experimental noise.

What people get wrong about the gold foil experiment

The most common misconception is that Rutherford personally ran the apparatus and watched the scintillation screen himself. He directed the research program, secured the radioactive source, and worked out the mathematical interpretation, but the daily measurements were Geiger and Marsden's work, carried out over months of repetitive, physically taxing manual counting. Textbook shorthand that credits "Rutherford's experiment" without naming Geiger and Marsden erases the actual laboratory labor behind the discovery.

A second frequent error is describing the result as if most alpha particles bounced back. In reality, only about 1 in 8,000 particles deflected at large angles; the vast majority passed through the foil with little or no deflection, exactly as the plum pudding model predicted. The significance of the result lay entirely in that small fraction being nonzero and statistically robust, not in any large-scale reversal of the beam.

A third misconception treats the 1911 nuclear model as a finished, modern picture of the atom. Rutherford's nucleus had no internal structure, no protons or neutrons identified as distinct particles, and his model could not explain why electrons don't radiate away their energy and collapse into the nucleus, a gap in classical physics that troubled physicists until Bohr introduced quantized orbits in 1913 and quantum mechanics matured further over the following two decades.

The takeaway

The gold foil experiment worked because Geiger and Marsden measured something nobody expected to find and did not explain it away, and because Rutherford took two careful years to build a mathematical model that could account for it rather than rushing out a guess. The result replaced a picture of the atom as a uniform, pudding-like blob with the model still taught today: overwhelmingly empty space, organized around an almost impossibly dense, positively charged nucleus. The lesson embedded in the experiment's method, that structure too small to see can be inferred from how particles bounce off it, still runs every particle accelerator in operation.

---

Read next:

• Ernest Rutherford: the New Zealander who split the atom
• Niels Bohr's biography
• What is a photon, explained
• The Millikan oil drop experiment explained
• The Michelson-Morley experiment explained

FAQ

Q: What did the gold foil experiment prove about atomic structure?

It showed that atoms are mostly empty space with almost all of their mass and positive charge concentrated in a tiny, dense central nucleus, replacing J.J. Thomson's earlier plum pudding model of evenly distributed charge throughout the atom.

Q: Who actually performed the gold foil experiment?

Hans Geiger and Ernest Marsden carried out the alpha particle scattering measurements at the University of Manchester between 1908 and 1909, under Ernest Rutherford's direction. Rutherford developed the mathematical nuclear interpretation and published it in 1911.

Q: How many alpha particles actually bounced backward?

Roughly 1 in 8,000 particles deflected at large angles, some by more than 90 degrees. The large majority of alpha particles passed straight through the gold foil with little or no deflection, which is what made the rare large-angle bounces statistically significant.

Q: What model of the atom did the gold foil experiment replace?

It replaced J.J. Thomson's 1904 plum pudding model, which described the atom as a diffuse ball of positive charge with electrons embedded evenly throughout it, rather than concentrated in a central nucleus.

Q: What is the exact quote Rutherford gave about the result?

Rutherford said, "It was almost as incredible as if you fired a 15-inch shell at a piece of tissue paper and it came back and hit you," a comparison recorded in E. N. da C. Andrade's 1964 book Rutherford and the Nature of the Atom.

Q: Did Rutherford's nuclear model explain everything about atoms?

No. It did not explain why electrons don't spiral into the nucleus, what the nucleus itself was made of, or how atoms produce specific spectral lines of light. Niels Bohr addressed the orbital stability problem in 1913 with quantized electron orbits, while protons and neutrons were not identified as separate nuclear particles until 1919 and 1932 respectively.

Q: Why did gold foil work better than other metals for this experiment?

Gold is extremely malleable, meaning it can be hammered into sheets only a few hundred atoms thick without tearing or developing holes, which let alpha particles pass through a controlled, uniform thickness of material rather than an uneven or perforated sample.