Analogy as an epistemological guiding principle

The origin of analogy as an epistemological principle goes back to Electromagnetism. The two theories of electricity and magnetism were found to behave so similarly that various concepts in one was inspired by an analogous concept in the other. The analogy worked so well that it resulted in the remarkable success of Maxwell's Electromagnetism. The belief in the analogy was further strengthened when Dirac found that assuming the existence of magnetic monopoles it is possible to --partly-- explain the quantisation of electric charge.

Many advances in physics are result of comparison and analogy. The naturally arising question thus is whether this `principle of analogy' is a useful --if not true-- principle for epistemology.

The reason this question interests me is that I find in various reflections, analogy is a quite powerful tool leading to fruitful results; it would not be wrong if I say I have been able to construct the whole of already-existing physics for myself following this principle. Is it the old mysterious unity of nature which empowers this principle? I do not know yet.