other regions of space, the law of the inverse square ceases to appear as even an approximation to the truth--as, for instance, the law of the inverse _fifth_ power has been thought to be nearer the mark for small molecular distances. And indeed the state of the great theories of Science in the present day--the confusion in which the Atomic theory of physics finds itself, the dismal insufficiency of the Darwin theory of the survival of the fittest; the collapse in late times of one of the fundamental theories of Astronomy, namely that of the stability of the lunar and planetary orbits; the cataclysms and convulsions which Geology seems just now to be undergoing; the appalling and indeed insurmountable difficulties which attach to the Undulatory theory of Light; the final wreck and abandonment of the Value-theory, the foundation-theory of Political Economy--all these things do not seem to point to very near limits of rigorous exactness! An impregnable theory, or one nearing the limit of impregnability, is in fact as great an absurdity as an impregnable armour-plate. Certainly, given the cannon-balls, you can generally find an armour-plate which will be proof against them; but given the armour-plate, you can always find cannon-balls which will smash it up. The method of Science, as being a method of artificial limitation or actual ignorance, is curiously illustrated by a consideration of its various branches. I have taken some examples from Astronomy, which is considered the most exact of the physical sciences. Now does it not seem curious that _Astronomy_--the study of the heavenly bodies, which are the most distant from us of all bodies, and most difficult to observe--should yet be the most perfect of the sciences? Yet the reason is obvious. Astronomy is the most perfect science _because we know least about it_--because our ignorance of the actual phenomena is most profound. Situated in fact as we are, on a speck in space, with our observations limited to periods of time which, compared with the stupendous flights of the stars, are merely momentary and evanescent, we are in somewhat the position of a mole surveying a railway track and the flight of locomotives. And as a man seeing a very small arc of a very vast circle easily mistakes it for a straight line, so we are easily satisfied with cheap deductions and solutions in Astronomy which a more extensive experience would cause us to reject. The man may have a long way to go along his "straight line" before he discovers that it is a curve; he may have much farther to go along his curve before he discovers that it is not a circle; and much farther still to go before he finds out whether it is an ellipse or a spiral or a parabola, or none of these; yet _what_ curve it is will make an enormous difference in his ultimate destination. So with the astronomer; and yet Astronomy is allowed to pass as an exact science![21]