Reflection involves running over various ideas, sorting them out, comparing one with another, trying to get one which will unite in itself the strength of two, searching for new points of view, developing new suggestions; guessing, suggesting, selecting, and rejecting. The greater the problem, and the greater the shock of doubt and resultant confusion and uncertainty, the more prolonged and more necessary is the process of "mere thinking." It is a more obvious phase of biology than of physics, of sociology than of chemistry; but it persists in established sciences. If we take even a mathematical proposition, not _after_ it has been demonstrated--and is thus capable of statement in adequate logical form--but while in process of discovery and proof, the operation of this subjective phase is manifest, so much so, indeed, that a distinguished modern mathematician has said that the paths which the mathematical inquirer traverses in any new field are more akin to those of the experimentalist, and even to those of the poet and artist, than to those of the Euclidean geometer. What makes the essential difference between modern research and the reflection of, say, the Greeks, is not the absence of "mere thinking," but the presence of conditions for testing its results; the elaborate system of checks and balances found in the technique of modern experimentation. The thinking process does not now go on endlessly in terms of itself, but seeks outlet through reference to particular experiences. It is tested by this reference; not, however, as if a theory could be tested by directly comparing it with facts--an obvious impossibility--but through use in facilitating commerce with facts. It is tested as glasses are tested; things are looked at through the medium of specific meanings to see if thereby they assume a more orderly and clearer aspect, if they are less blurred and obscure.