“Prevision” thus becomes the essential characteristic of scientific knowledge, and that independently of any utilitarian mental reservation. For the eventual applications of science do not determine its theoretical advance. The prevision with which we are here concerned consists solely in the possibility of knowing with certainty without observing. It is knowledge _a priori_ in the Aristotelian sense of the word, of which mathematics present the most perfect model. A rectilinear triangle being given, I do not need experience to know with certainty that the sum of the angles in it is equal to two right angles. Thus understood prevision applies to the present, and even to the past, as well as to the future. When Comte writes “All science has prevision for its aim,”[38] we must understand: “All science tends to substitute deduction to experience, rational to empirical knowledge.” This prevision, a necessary consequence of the constant relations discovered between phenomena, will allow men never to mistake real science for fruitless erudition, which accumulates facts without deducing them one from another. Thus the formula cited above enlarges itself: “Science is composed of laws and not of facts.” The more deduction is substituted to experience, the better is the extension and connection of our knowledge realised. Consequently, the more also does science draw near to that unity which is imperatively claimed by our understanding, and which is for it the criterion of truth. “Real science,” says Comte, “regarded from the highest point of view, has no other general object but to establish or to fortify unceasingly the intellectual order, which is the basis of all other order.”[39] The mind which applies itself to the contemplation of the world requires, before everything, to find it intelligible. “Real” science satisfies it, not in imagining wills and causes, as did theology and metaphysics, but in discovering order in the constant relations between phenomena. When this order is harmonious, that is to say, when the several classes of phenomena are conceived as homogenous, and as similarly governed by laws, “the spontaneous unity of our understanding is consolidated.” It matters little that the various orders of phenomena are given to us as irreducible to one another. The highest object of science is to determine the point of view from which all phenomena appear intelligible, and this point of view is one as the understanding itself is one. III. Perhaps it would have been easy to pass from this conception of positive science to a theory of knowledge, and to a metaphysical view of nature, both idealistic. But Comte neither could nor would push his theory in this direction. In this respect nothing is more significant than his way of understanding the relativity of science.