According to his own expression then he endeavours “to transform science into philosophy.” But on what conditions will the transformation be effected? If science were to lose in it its characteristics of positiveness, of reality, and of relativity, to assume those of a metaphysical doctrine, this change would be neither desirable nor possible. The transformation will simply consist in giving to science the philosophical character which it does not yet possess, namely universality. While thus acquiring a new property, positive science should lose none of those which it already possesses, and which constitute its value. Thus, in the “transformation of science into philosophy,” what is transformed at bottom is not science which remains itself while becoming general from being special: it is philosophy rather which is transformed. The latter will henceforth undoubtedly be conceived as the highest and most comprehensive form of positive knowledge, but as constituting a part of that knowledge. It has been said that Comte does away with philosophy, by reducing it to being merely the “generalisation of the highest results of the sciences.” This is not a proper interpretation of his thought. Up to the present time the duties performed by the philosophical doctrines have been indispensable. Comte intends that his system shall fulfil them in future. Beside science properly so-called, which is always special, philosophy which represents the “point of view of the whole” must arise. On this condition alone can the government of minds and the “perfect logical coherence” become possible. Philosophy will then not merely be a “generalisation of the highest results of the sciences.” The synthesis of the sciences must be brought about according to a principle to which they will be all related. It must really be a “summing up of experience.” But if this philosophy thus coalesces with science it must also be _real_ like it, and all real knowledge is necessarily positive and relative. In short, the distinction between science and philosophy implies no specific difference between these two kinds of speculation. On the contrary, there exists between them homogeneity of doctrine and unity of method.