Observation - One sees an event or phenomenon and desires to know more about it.
Hypothesis - One forms a guess as to what might explain the event.It might be an "educated guess" or a more sophisticated explanation of the phenomenon.
Experimentation - One then tests the hypothesis by devising a multitude of experiments with controls to eliminate all but one variable at a time. This process of testing continues until the hypothesis is either confirmed or denied. Modification of the hypothesis is to be expected during this process.
Formation of a Theory - When testing has yielded a fairly definite understanding of the event it can be stated as a theory. Further testing will tend to confirm, modify, or deny the theory. The rigor with which one subjects the theory to testing is a matter of one's (subconscious?) desire to see the theory confirmed or denied. The skeptic does more testing than the gullible.
Statement of a Law - After long-term extensive testing, with completely consistent results, the theory can possibly be stated as a law. But even then, it must be understood, the "law" is still subject to further modification as new tests are devised. It is conceivable, for example, that as we learn more about the physical phenomenon we call gravity, that "law" could be modified.
The one essential requirement for the scientific method of investigation to be applicable is the reproducibility of the event in question. One must be able to repeat it or to continually observe it in order to experiment (and thus test a hypothesis) with it.
Herein lies the limit of science. It is, at once, its Achilles' heel and its greatest strength. Science derives great authority from the experimental method.