XXXV.  But those who affirm that it has only three divisions, do not think that the argumentation ought to be conducted in any other way, but they find fault with this arrangement of the divisions.  For they say that neither the proposition nor the assumption ought to be separated from their proofs; and that a proposition does not appear to be complete, nor an assumption perfect, which is not corroborated by proof.  Therefore, they say that what those other men divide into two parts, proposition and proof, appears to them one part only, namely proposition.  For if it be not proved, the proposition has no business to make part of the argumentation.  In the same way they say that that which those other men call the assumption, and the proof of the assumption, appears to them to be assumption only.  And the result is, that the whole argumentation being treated in the same way, appears to some susceptible of five divisions, and to others of only three; so that the difference does not so much affect the practice of speaking, as the principles on which the rules are to be laid down.

But to us that arrangement appears to be more convenient which divides it under five heads; and that is the one which all those who come from the school of Aristotle, or of Theophrastus, have chiefly followed.  For as it is chiefly Socrates and the disciples of Socrates who have employed that former sort of argumentation which goes on induction, so this which is wrought up by ratiocination has been exceedingly practised by Aristotle, and the Peripatetics, and Theophrastus; and after them by those rhetoricians who are accounted the most elegant and the most skilful.  And it seems desirable to explain why that arrangement is more approved of by us, that we may not appear to have adopted it capriciously; at the same time we must be brief in the explanation, that we may not appear to dwell on such subjects longer than the general manner of laying down rules requires.

XXXVI.  If in any sort of argumentation it is sufficient to use a proposition by itself, and if it is not requisite to add proof to the proposition; but if in any sort of argumentation a proposition is of no power unless proof be added to it; then proof is something distinct from the proposition.  For that which can be joined to a thing or separated from it, cannot possibly be the same thing with that to which it is joined or from which it is separated.  But there is a certain kind of argumentation in which the proposition does not require confirmatory proof, and also another kind in which it is of no use at all without such proof, as we shall show.  Proof, then, is a thing different from a proposition.  And we will demonstrate that point which we have promised to show in this way:—­The proposition which contains in itself something manifest, because it is unavoidable that that should be admitted by all men, has no necessity for our desiring to prove and corroborate it.