A superseded, or obsolete, scientific theory is a scientific theory that was once commonly accepted but (for whatever reason) is no longer considered the most complete description of reality by mainstream science; or a falsifiable theory which has been shown to be false. This label does not cover theories that are yet to gain wide support in the scientific community (protoscience or fringe science). This also does not cover theories that were never widely accepted. Some theories which were only supported under specific political authorities, like Lysenkoism, may also be included.
In some cases, the theory has been completely discarded. In other cases, the theory is still useful because it provides a description that is "good enough" for a particular situation, and is more easily used than the complete theory (often because the complete theory is too mathematically complex to be easily usable). An example of this is the use of Newtonian physics in many mechanical engineering applications, and even in calculating the orbits of satellites, because the deviation from such physics is smaller than other sources of error. Karl Popper suggested that all scientific theories should be falsifiable otherwise they could not be tested by experiment. Anything that cannot be shown by experiment to be false would therefore be an axiom and have an absolute status, beyond any confirmation or refutation.
Electron cloud model following the invention of Quantum Mechanics in 1925 and the eventual atomic orbital models derived from the quantum mechanical solution to the hydrogen atom.
The Open Polar Sea, an ice-free sea once supposed to surround the North Pole
Rain follows the plow - the theory that human settlement increases rainfall in arid regions (only true to the extent that crop fields evapotranspirate more than barren wilderness)
Superseded geological theories
Continental drift was superseded by plate tectonics
Astrology, which led to the development of astronomy - astrology was discredited by al-Farabi, Alhacen, al-Biruni, Avicenna and Averroes
Phrenology, was once widely studied but now considered a pseudoscience
Numerology, as distinct from number theory, now considered a pseudoscience
Approximate theories
Here are theories that are no longer considered the most complete representation of reality, but are still useful in particular domains. For many theories a more complete model is known, but in practical use the coarser approximation provides good results with much less calculation.
Steady State - replaced by an expanding universe model starting with the Big Bang. See also: dark matter.
Atomic theory - Atoms are no longer thought to be indivisible, but are now seen to be composites.
Heliocentric universe theory - This is still used in the coordinate system of celestial mechanics.
Newtonian mechanics - extended by Theory of Relativity and quantum mechanics. Still useful in engineering and physics at either middling (human) scales or where appreciable fractions of the speed of light need not be considered.
Bohr model of the atom - extended by the quantum mechanical model of the atom.
Newton's sine-square law for the force of a fluid on a body - no longer considered useful at low speeds, though it has found application in hypersonic flow