edition of Pindar; a very fair knowledge of the language is assumed beforehand. It might be thought that the process of reduction, as it were, could be continued, until finally an account was prepared where no technical terms were used. But such an account would be, at best, like a translation of Greek poetry; the essential quality would be gone. Such translations have, of course, their uses, but the attraction of science for the scientific man, like the attraction of a poem for the poet, is not to be communicated in this way. In art the separation of matter and form is not really possible, and the same is true of the sciences. III In their apologias, which have now become so common, men of science never weary of pointing out that it is the method of science which is really worthy of adoption by philosophers and that the results of science are merely provisional. The philosopher who bases his system upon the results reached at any given time by any given science has ensured the ultimate downfall of his system. He is sometimes told that the adoption of scientific methods, on the other hand, will enable him to make sure progress. At first sight there seems to be a contradiction here, for if the scientific method is infallible why are the results reached by it provisional? To judge from the history of science, the scientific method is excellent as a means of obtaining plausible conclusions which are always wrong, but hardly as a means of reaching the truth. The contradiction is only apparent, however, for it will be found that there is a part of every discarded hypothesis which is incorporated in the new theory. The discarded hypothesis proves to have been too general; the scientific man made a mistake of the same kind as the philosopher who uses the hypothesis as the basis of a general system. It is now known, for instance, that Newton’s theory of gravitation is very probably not exactly true; in most cases, however, it remains very nearly true, and there are large regions of dynamical astronomy which are unaffected by the alteration. The Newtonian laws of motion, again, are not sufficient to describe the motion of bodies moving with very large velocities, but they are very nearly true for all ordinary velocities. That the theories which have taken the place of those abandoned are exactly true is very improbable; they are, however, nearer the truth. We may say, therefore, that while the scientific method may, quite possibly, never enable us to reach the exact truth, successive applications of it enable us to approximate nearer and nearer to the exact truth. In this lies its chief difference from the methods usually adopted in philosophy, which aim at obtaining, at one blow, theories which shall never need revision. It is for this reason that philosophy does not progress.