But a simple conclusion is reprehended if that which follows does not appear of necessity to cohere with that which has gone before.  For this very proposition, “If he breathes, he is alive,” “If it is day, it is light,” is a proposition of such a nature that the latter statement appears of necessity to cohere with the preceding one.  But this inference, “If she is his mother, she loves him,” “If he has ever done wrong, he will never be chastised,” ought to be reprehended in such a manner as to show that the latter proposition does not of necessity cohere with the former.

Inferences of this kind, and all other unavoidable conclusions, and indeed all argumentation whatever, and its reprehension too, contains some greater power and has a more extensive operation than is here explained.  But the knowledge of this system is such that it cannot be added to any portion of this art, not that it does of itself separately stand in need of a long time, and of deep and arduous consideration.  Wherefore those things shall be explained by us at another time, and when we are dealing with another subject, if opportunity be afforded us.  At present we ought to be contented with these precepts of the rhetoricians given for the use of orators.  When, therefore, any one of these points which are assumed is not granted, the whole statement is invalidated by these means.

XLVII.  But when, though these things are admitted, a conclusion is not derived from them, we must consider these points too, whether any other conclusion is obtained, or whether anything else is meant, in this way,—­If, when any one says that he is gone to the army, and any one chooses to use this mode of arguing against him, “If you had come to the army you would have been seen by the military tribunes, but you were not seen by them, therefore you did not go to the army.”  On this case, when you have admitted the proposition, and the assumption, you have got to invalidate the conclusion, for some other inference has been drawn, and not the one which was inevitable.

And at present, indeed, in order that the case might be more easily understood, we have brought forward an example pregnant with a manifest and an enormous error; but it often happens that an error when stated obscurely is taken for a truth; when either you do not recollect exactly what admissions you have made, or perhaps you have granted something as certain which is extremely doubtful.  If you have granted something which is doubtful on that side of the question which you yourself understand, then if the adversary should wish to adapt that part to the other part by means of inference, it will be desirable to show, not from the admission which you have made, but from what he has assumed, that an inference is really established; in this manner:—­“If you are in need of money, you have not got money.  If you have not got money, you are poor.  But you are in need of money, for if it were not so you would not pay