This may be made more clear by examples. If a child, who has had little experience of the laws of nature, and has learned nothing from books, is gravely assured by his instructor, that in a distant region of the ocean there is an island where stones fly upward instead of downward, and men walk on their heads instead of their feet, the young philosopher, however acute and ingenious we may suppose him to be, certainly could not offer one valid argument against the alleged fact. He could only stare, and wonder, and say that it might be so _for all that he knew to the contrary_. Just so, when the atheist tells us, that far off in infinite space is a region, of which we can see nothing, even with our best telescopes, except a faint glimmer of light, floating like a cloudlet in the heavens, where the primitive atoms of matter, directed by gravity alone, are slowly congregating together, and forming suns, and planets, and secondary satellites, and giving birth to such intricate harmonies of mutually dependent and revolving worlds as those which have prevailed for ages in our own system; or that, thousands of years ago, the same unassisted laws of matter, which we now see producing only such comparatively meagre and insufficient results, actually caused animalcules to be produced from pure sand, and fishes to be created out of oysters, and birds to be generated by slimy and grovelling reptiles, and men to be born from monkeys;--if he should tell us all this, certainly we could offer no direct confutation of the wonderful tale. In regard to alleged facts of this character, the wisest of men are, and always must be, mere children. But it would be monstrous to say, that this wild assertion derived any support from their admitted bewilderment and incapacity. This would be to attempt to found knowledge upon ignorance. The dim analogies resting on questionable facts, the bold assumptions and slippery arguments on which such daring hypotheses must be based, can be refuted, for the most part, only by reasoning in kind,--by arguments nearly as uncertain, it may be, as those which they are brought to answer. We cannot _prove_ a negative; we can only show the insufficiency of the ground on which the opposite assumption is made to rest; and enough is done for this end, when it is made to appear, that the whole scheme is a _mere_ hypothesis.