There is, therefore, not a science _of_ history but an ancillary science _for_ history, which ascertains that which has been. For the historical outlook itself the data are always symbols. Scientific research, on the contrary, is science and only science. In virtue of its technical origin and purpose it sets out to find data and laws of the causal sort and nothing else, and from the moment that it turns its glance upon something else it becomes _Metaphysics_, something trans-scientific. And just because this is so, historical and natural-science data are different. The latter consistently repeat themselves, the former never. The latter are _truths_, the former _facts_. However closely related incidentals and causals may appear to be in the everyday picture, fundamentally they belong to different worlds. As it is beyond question that the shallowness of a man’s history-picture (the man himself, therefore) is in proportion to the dominance in it of frank incidentals, so it is beyond question that the emptiness of written history is in proportion to the degree in which it makes the establishment of purely factual relations its object. The more deeply a man lives History, the more rarely will he receive “causal” impressions and the more surely will he be sensible of their utter insignificance. If the reader examines Goethe’s writings in natural science, he will be astounded to find how “living nature” can be set forth without formulas, without laws, almost without a trace of the causal. For him, Time is not a distance but a feeling. But the experience of last and deepest things is practically denied to the ordinary savant who dissects and arranges purely critically and allows himself neither to contemplate nor to feel. In the case of History, on the contrary, this power of experience is _the_ requisite. And thus is justified the paradox that the less a historical researcher has to do with real science, the better it is for his history. To elucidate once more by a diagram: