precipitation--but that it, _i.e._ its substance, cannot have then started into existence; for this implies a total impossibility and is utterly inconceivable. The certainty with which we assume this beforehand (_à priori_), proceeds from the fact, that our Understanding possesses absolutely no form under which to conceive the beginning and end of Matter. For, as before said, the law of causality--the only form in which we are able to conceive changes at all--is solely applicable to _states_ of bodies, and never under any circumstances to the existence of _that which undergoes_ all changes: _Matter_. This is why I place the principle of the permanence of Matter among the corollaries of the causal law. Moreover, we cannot have acquired _à posteriori_ the conviction that substance is permanent, partly because it cannot, in most instances, be empirically established; partly also, because every empirical knowledge obtained exclusively by means of induction, has only approximate, consequently precarious, never unconditioned, certainty. The firmness of our persuasion as to this principle is therefore of a different kind and nature from our security of conviction with regard to the accuracy of any _empirically_ discovered law of Nature, since it has an entirely different, perfectly unshakable, never vacillating firmness. The reason of this is, that the principle expresses a _transcendental_ knowledge, _i.e._ one which determines and fixes, _prior_ to all experience, what is in any way possible within the whole range of experience; but, precisely by this, it reduces the world of experience to a mere cerebral phenomenon. Even the most universal among the non-transcendental laws of Nature and the one least liable to exception--the law of gravitation--is of empirical origin, consequently without guarantee as to its absolute universality; wherefore it is still from time to time called in question, and doubts occasionally arise as to its validity beyond our solar system; and astronomers carefully call attention to any indications corroborative of its doubtfulness with which they may happen to meet, thereby showing that they regard it as merely empirical. The question may of course be raised, whether gravitation takes effect between bodies which are separated by an _absolute_ vacuum, or whether its action within a solar system may not be mediated by some sort of ether, and may not cease altogether between fixed stars; but these questions only admit of an empirical solution, and this proves that here we have not to do with a knowledge _à priori_. If, on the other hand, we admit with Kant and Laplace the hypothesis, as the most probable one, that each solar system has developed out of an original _nebula_ by a gradual process of condensation, we still cannot for a moment conceive the possibility of that original substance having sprung into being out of _nothing_: we are forced to assume the anterior existence of its particles