The general fact of most importance consists of changes in the blood circulation. Increase of intellectual activity means an increase of work in the cortical cells, dependent on a congested, sometimes a temporarily anaemic state. Hyperaemia seems rather the rule, but we also know that slight anaemia increases cortical excitability. "Weak, contracted pulse; pale, chilly skin; overheated head; brilliant, sunken, roving eyes," such is the classic, frequently quoted description of the physiological state during creative labor. There are numerous inventors who, of their own accord, have noted these changes--irregular pulse, in the case of Lagrange; congestion of the head, in Beethoven, who made use of cold douches to relieve it, etc. This elevation of the vital tone, this nervous tension, translates itself also into motor form through movements analogous to reflexes, without special end, mechanically repeated and always the same in the same man--e.g., movement of the feet, hands, fingers; whittling the table or the arms of a chair (as in the case of Napoleon when he was elaborating a plan of campaign), etc. It is a safety-valve for the excessive flow of nervous impulse, and it is admitted that this method of expenditure is not useless for preserving the understanding in all its clearness. In a word, increase of the cerebral circulation is the formula covering the majority of observations on this subject. Does experimentation, strictly so called, teach us anything on this point? Numerous and well-known physiological researches, especially those of Mosso, show that all intellectual, and, most of all, emotional, work, produces cerebral congestion; that the brain-volume increases, and the volume of the peripheral organs diminishes. But that tells us nothing particularly about the imagination, which is but a special case under the rule. Latterly, indeed, it has been proposed to study inventors by an objective method through the examination of their several circulatory, respiratory, digestive apparatus; their general and special sensibility; the modes of their memory and forms of association, their intellectual processes, etc. But up to this time no conclusion has been drawn from these individual descriptions that would allow any generalization. Besides, has an experiment, in the strict sense of the word, ever been made at the "psychological moment"? I know of none. Would it be possible? Let us admit that by some happy chance the experimenter, using all his means of investigation, can have the subject under his hand at the exact moment of inspiration--of the sudden, fertile, brief creative impulse--would not the experiment itself be a disturbing cause, so that the result would be _ipso facto_ vitiated, or at least unconvincing?