3dly. Even amongst miracles _not_ internal, we have shown--that if one class (the merely evidential and polemic) are incommunicable, _i.e._ not capable of propagation to a remote age or place, they have sufficiently fulfilled their immediate purpose by their immediate effect. But such miracles are alien and accidental to Christianity. Christ himself reproved severely those who sought such signs, as a wicked, unbelieving generation; and afterwards he reproved, with a most pathetic reproach, that one of his own disciples who demanded such a sign. But besides these evidential miracles, we noticed also, 4thly. The constituent miracles of Christianity; upon which, as regarded Hume's argument, we reserved ourselves to the latter section: and to these we now address ourselves. But first we premise this _Lemma:_--That an _á priori_ (or, as we shall show, an _a posteriori_) reason for believing a miracle, or for expecting a miracle, will greatly disturb the valuation of _x_ (that is, the abstract resistance to credibility), as assumed in Hume's argument. This is the centre in which we are satisfied, lurks that πρωτον ψευδος which Hume himself suspected: and we add, that as a vast number of witnesses (according to a remark made in Sect. II.) will virtually operate as a reduction of the value allowed to _x_, until _x_ may be made to vanish altogether,--so in the reverse order, any material reduction of value in _x_ will virtually operate exactly as the multiplication of witnesses; and the case _Alpha_ will be raised to the case _Beta._ This _Lemma_ being stated as a point of appeal in what follows, we proceed to SECTION V. ON HUME'S ARGUMENT, AS AFFECTED BY THE PURPOSE.