It follows from this that science contains within itself neither its starting-point nor its terminus. Both are given it by “common sense” whence it springs. The starting-point is the spontaneous observation of constant relations between the most simple phenomena. The terminus is the knowledge of these same relations among all given phenomena, as complete and as precise as our requirements demand. Indeed the common sense, or the popular wisdom, is soon baffled by the complexity of phenomena. If we had no other guide we should know very little, and in nearly all cases we should be reduced to a kind of empirical divination. The function of science is to substitute a real knowledge of laws to this divination. This function would never have been fulfilled if the human mind had not possessed the property of being able to separate theory from practice. Undoubtedly the former proceeds from the latter. As has been said, every science is born from a corresponding art, and from the desire to perfect it. But this perfecting would not have gone very far, if the human mind had never lost sight of it. Happily, man is capable of temporarily forgetting his immediate interests in the pursuit of knowledge. By degrees, from the complexity of concrete cases, he has learnt to disengage the elements common to a whole class of phenomena. He has thus formed the idea of law, or the invariable relation between given phenomena. Beyond the intellectual satisfaction which this knowledge gave him, he found in time applications of it which he would never have imagined beforehand. To quote an example from a civilisation already very advanced, when the Greek geometers patiently applied themselves to the study of conic sections, did they suspect that their labours would one day serve in calculating certain astronomical determinations upon which the safety of mariners would depend? In this way, science, utilitarian in its origin, since it sprang from the practical needs of man, utilitarian in its end, since it aims at providing for those needs, has nevertheless been unable to develop itself and will be unable still to do so in the future, except by neglecting this very utility. Better to fulfil its destiny, it must provisionally forget it; and it will be the more useful, in the long run, in proportion as it will have been the more disinterested. We never know, _a priori_, if a discovery which finds no application to-day, combined later with another one, will not be of capital interest for mankind. Therefore it is of the highest importance that theoretical order should remain clearly distinct from the practical order.