The spherical horse in simple harmonic motion is a canonical reference to a basic lack of real world practicality commonly encountered in physics, particularly as taught in the classroom. To illustrate:
A gangster assembled an engineer, a chemist, and a physicist, and ordered them to find a way to guarantee that his horse would win in the next week's races. The day before the race, they reconvened.
Gangster: Engineer, what have you got?
Engineer: I've invented a way to weave metallic threads into the saddle blanket so that they will act as the plates of a battery and provide electrical shock to the horse.
Gangster: That's very good! But let's hear from the chemist.
Chemist: I've synthesized a powerful stimulant that disolves into simple blood sugars after ten minutes and therefore cannot be detected in post-race tests.
Gangster: Excellent, excellent! Physicist?
Physicist: Well, let us consider a spherical horse in simple harmonic motion...
Doing "fortune -m spherical" in a unix environment is likely to give you an example or two. Sometimes animals other than horses are used. (A spherical bear in simple harmonic motion is another commonly encountered creature. Edit: It has been pointed out to me that cows are common, though I've never personally witnessed them used.)

While this is somewhat of an exageration, it is unlikely that you will ever sit in a physics class wherein anything approximating reality is discussed. The reason for this, of course, is that physics classes teach theory, for which it is unnecessary to know such details as what actually may happen in the real world. Everything from levers to Newton, to energy, to Quantum Physics is easiest explained using abstract, clean examples. It is just too complicated to try to figure out the topography of a horse and how it affects the animal's coefficient of friction, when a spherical approximation will do just as well for learning the concepts.

The most common (and reasonable) defense of this methodology is that physicists will first solve the simplified problem, then extend the solution to deal with more complex situations. The reality is, however, that there is no time in the classroom to do this, and practicing physists don't gain anything by solving dirty problems, their careers are better served by abstracting away as much as possible. The problem of horse topography as it affects its trajectory is left to the mechanical engineer.

The only solution to this dichotomy is, of course, to breed a spherical horse. We are getting there.


Thanks to hilding for letting me know that cows are commonly used, and cjeris for pointing out the difference between topology and topography (The topology of a horse is a torus - maybe physicists could meet genetecists half way at that point.)