The thrill and adventure of science are exemplified by relativity theory.
Development of any science proceeds from less general to more general, i.e. in an inductive direction. For example, close to the beginning of science the relativity theory could not stand. The same is even true for the Newtonian theory. However, this inductive direction of development is an optical illusion.
The whole pattern is a series of leaps out and then a return to the observational data. The deductive method consists of leaps into the unknown, and this is not a rationally justifiable step in science. While it is true of every type of thought, in sci-ence there is testing which eliminates those leaps inconsistent with observational data. The language of empirical science is characterised by the readiness to be falsified. A non-empirical science is a system of tautologies.
The further spread of data is covered by additional leaps. That is to say, sweeps into the deductive direction with wider and wider spreads comes from the very nature of science. Merely to leap out to a hypothesis covering observations at hand is just to develop an ad hoc hypothesis. The truly scientific hypothesis covers more than the available observations; it is a leap in the dark, and so gives scope for falsification.
The principle of falsifiability is an attitude, not a logical position.