Everyone knows what is physics, however the question "What do physi- cists do?" does not have simple answer. The first answer popping up might be that one studies Nature in all its possible appearances. Obviously, this answer is absolutely correct, however it is useless since it does not give any spirit of Physics, in the sense that one cannot understand the way how Physics develops. This development can be illustrated, for example, by the change of the concept of light.
Early scientific experiments have revealed the wave nature of light, leading to the concept of electromagnetic waves. However, later it was shown that under certain conditions light acts like a particle, which was named photon. This apparent contradiction has been nicely resolved in the framework of quantum theory, however both these concepts still remain broadly used in the modern physics. Why is it so? Is it because of some deep and complicated waveparticle duality? No, it is just because these concepts give a nice and easy (as compared to the description in the framework of the relativistic quantum theory) way to describe certain properties of light, i.e. these concepts are useful physical models. This example allows to formulate a very important conclusion: Physics aims to understand Nature in some simplified terms, i.e. physicists look for good models.
This approach is very deeply rooted and used not only for the description of different objects (point-like bodyE waveEetc) but it is also essential for the treatment of different physical processes. Namely, all the words, like idealE hotE coldE quantumE are in some sense physical models. These idealizations allow the mathematical description of the processes occurring around us and, consequently, a comparison of the theoretical predictions with observational data. Often, new experiments require further development of physical models, as the dual particle-wave concept of light led to the formulation of the quantum description. Thus, the development of physics occurs as a loop consisting of collection of observational data, suggestion of plausible models, verification of these models with other experiments. Ideally, this discards most of the models and allows a few others to be further developed. The evolution of the models is a very natural process, and it is very seldom that a successful model is suggested before any detailed experimental study of the corresponding object. However, there is an interesting exception from this rule an astrophysical object whose properties tend to fit very simplified models. This object is known as pulsar.
Pulsars (rapidly rotating neutron stars) are the remnants of supernovae explosions produced by stars, which separate the two standard branches of this process. According to the pioneering works of Chandrasekhar, if the core of the exploding star is low mass, a dwarf star is formed at the final stage of the stellar evolution; if the core is heavy, then a black hole is expected to be created. In late 1930s it was pointed out by Oppenheimer&Volkoff that theoretically there is a certain gap between these two solutions: when the mass of the core is not sufficiently large to collapse into a black hole, but the pressure inside the remnant is so large that electrons and protons can merge together. Thus, an extremely compact star, consisting of neutrons, should be formed. There had been no observational searches for this purely theoretical objects, until a peculiar radio source was detected during some routine observations in 1967. The source was characterized by very frequent radio signals with a period which appeared to be more precise than any clocks available at that time (actually, this statement remains true even now, since some pulsars are more stable even than atomic clocks). Shortly, it has been realized that the observed behavior fits the best the concept of neutron stars. Moreover, it appeared that all the key properties (like magnetic field strength, rotation period and typical energy release) of these objects are fully explained in the framework of the simple theoretical model suggested by Oppenheimer&Volkoff 30 years before the first pulsar has been detected.