has the character of clarifying on the basis of what is dear—of explanation. Explanation is always twofold. It accounts for an unknown by means of a known, and at the same time it verifies that known by means of that unknown. Explanation takes place in investigation. In the physical sciences investigation takes place by means of experiment, always according to the kind of field of investigation and according to the type of explanation aimed at. But physical science does not first become research through experiment; rather, on the contrary, experiment first becomes possible whre and only where the knowledge of nature has been transformed into research. Only because modern physics is a physics that is essentially mathematical can it be experimental. Because neither medieval doctrina nor Greek epistēmē is science in the sense of research, for these it is never a question of experiment. To be sure, it was Aristotle who first understood what empeiria (experientia) means: the observation of things themselves, their qualities and modifications under changing conditions, and consequently the knowledge of the way in which things as a rule behave. But an observation that aims at such knowledge, the experimentum, remains essentially different from the observation that belongs to science as research, from the research experiment; it remains essentially different even when ancient and medieval observation also works with number and measure, and even when that observation makes use of specific apparatus and instruments. For in all this, that which is decisive about the experiment is completely missing. Experiment begins with the laying down of a law as a basis. To set up an experiment means to represent or conceive [vorstellen] the conditions under which a specific series of motions can be made susceptible of being followed in its necessary progression, i.e., of being controlled in advance by calculation. But the establishing of a law is accomplished with reference to the ground plan of the objectsphere. That ground plan furnishes a criterion and constrains the anticipatory representing of the conditions. Such representing in and through which the experiment begins is no random imagining, That is why Newton said, hypothesis non fingo, "the bases that are laid down are not arbitrarily invented." They are developed out of the ground plan of nature and are sketched into it. Experiment is that methodology which, in its planning