law and usages affecting the matter which is the subject of inquiry.  For all these circumstance are matters requiring proof, and we will explain them more carefully, when we come to speak about conjectural statements separately.  Therefore, each of these points in a reprehension of the statement of the adversary must be laboured, and it must be shown either that such and such a thing is no proof, or that it is an unimportant proof, or that it is favourable to oneself rather than to the adversary, or that it is altogether erroneously alleged, or that it may be diverted so as to give grounds to an entirely different suspicion.

XLIV.  But when anything is alleged as a proper object of comparison, since that is a class of argument which turns principally on resemblance, in reprehending the adversity it will be advisable to deny that there is any resemblance at all to the case with which it is attempted to institute the comparison.  And that may be done if it be proved to be different in genus or in nature, or in power, or in magnitude, or in time or place, or with reference to the person affected, or to the opinions generally entertained of it.  And if it be shown also in what classification that which is brought forward on account of the alleged resemblance and in what place too the whole genus with reference to which it is brought forward, ought to be placed.  After that it will be pointed out how the one thing differs from the other, from which we shall proceed to show that a different opinion ought to be entertained of that which is brought forward by way of comparison, and of that to which it is sought to be compared.  And this sort of argument we especially require when that particular argumentation which is carried on by means of induction is to be reprehended.  If any previous decision be alleged, since these are the topics by which it is principally established, the praise of those who have delivered such decision, the resemblance of the matter which is at present under discussion to that which has already been the subject of the decision referred to, that not only the decision is not found fault with because it is mentioned, but that it is approved of by every one, and by showing too, that the case which has been already decided is a more difficult and a more important one than that which is under consideration now.  It will be desirable also to invalidate it by arguments drawn from the contrary topics, if either truth or probability will allow us to do so.  And it will be necessary to take care and notice whether the matter which has been decided has any real connexion with that which is the present subject of discussion, and we must also take care that no case is adduced in which any error has been committed, so that it should seem that we are passing judgment on the man himself who has delivered the decision referred to.

It is desirable further to take care that they do not bring forward some solitary or unusual decision when there have been many decisions given the other way.  For by such means as this the authority of the decision alleged can be best invalidated.  And it is desirable that those arguments which are assumed as probable should be handled in this way.