Rutherford's nuclear atom model was proposed on the basis of the results of the alpha-particle scattering experiment carried out in 1911 by his students Hans Geiger and Ernest Marsden, in which a narrow beam of alpha particles from a radioactive source was directed at a thin gold foil, and the pattern of scattered particles measured across scattering angles from 0 to 180 degrees using a movable fluorescent screen. The results were startling: most alpha particles passed straight through the foil essentially undeflected; a smaller fraction were deflected through small angles; a very small fraction (about one in a thousand) were deflected by more than 90 degrees; and a very few were scattered straight back, deflected by a full 180 degrees.
Because most of the beam passed through almost unaffected, Rutherford concluded that an atom consists mostly of empty space. But the rare large-angle and back-scattering events could only be explained if the atom's positive charge and almost all of its mass were concentrated into an extremely small, dense region - a tiny nucleus, of size of order 10−14 m - at the very centre of the atom, with the much lighter, negatively-charged electrons surrounding it. Since static charges cannot form a stable equilibrium, Rutherford proposed that the electrons must instead be continuously revolving around this central nucleus in circular orbits, analogous to planets orbiting the Sun.
Despite correctly identifying the existence of the atomic nucleus - arguably Rutherford's most important contribution to physics - this model has two serious drawbacks it cannot resolve on its own. First, it cannot explain how electrons are actually distributed around the nucleus, nor why the atom is stable at all: classical electrodynamics predicts that any continuously accelerating (orbiting) charge must radiate away energy, causing the electron's orbit to shrink steadily in an inward spiral until it collapses into the nucleus - something real atoms are never observed to do. Second, this predicted continuous energy loss should produce a continuous emission spectrum, whereas real atoms are observed to emit light only at sharp, discrete wavelengths (line spectra). Resolving both of these unexplained gaps is precisely what motivated Niels Bohr's subsequent, quantised model of the atom.