Having seen how these obtain at the same time: ‘some quantity is not a thing really distinct from substance,’ and ‘quantity is an accident’; it remains to show that the opinion of Augustine does not contradict the previous statements; for Augustine does not intend that magnitudes are accidents really distinct from substances and inhering in them. For it seems wonderful that one accident might be able to make one part to be separate from another and yet God might not be able to do this, unless He would join one other thing to the parts of a substance. For when two distinct things are brought into being, who could prevent God from making one to be outside another, although He could join no other thing to them? And so that whole thing composed of them might be able to be great, although no other thing would be produced in it. For the blessed Augustine posits that magnitude is an accident; for a substance itself can be larger or smaller through a proper change of substance, according to that which can have parts more or less separate, which can touch without every thing inhering de novo in the substance itself. For when a substance having parts separate in act and in situation is produced, those parts can be still more separate without any added thing inherent in them, and then that substance is considered larger; because for a substance to be larger is nothing other than to have parts more separate in situation. Moreover, those parts can be less distant by virtue of an active cause, and then the substance is also smaller, because for a substance to be smaller is nothing other than to have parts less separate; and that this is not surprising, can be shown through a contrary opinion. For if quantity be an absolute thing really distinct from substance and quality, air may be rarefied; if this is done, either the total preceding quantity is corrupted or the total preceding quantity is not corrupted.
It is not possible to grant; first, for in any rarefaction infinite absolute things distinct as regards themselves as wholes, not making one numerically, would then be destroyed. For since there may be an infinite substance in each continuous time, such infinite quantities would be destroyed in that time in which something without interruption becomes rarefied; because, from this, that rarefaction is continuous in each instant, there would be quantity distinct according to itself as a whole and distinct from a prior and posterior quantity. But it is obvious that they do not make one numerically; for one does not remain with another.
Nor is it sufficient to say that those quantities have not been actually, but potentially only. For every accident really distinct and denominating its subject, which is not part of another, is actually; but these continuous quantities now corrupted were of this kind; therefore, it follows that things infinite actually, distinct according to themselves as a whole, not making one, would now be corrupted, which seems sufficiently absurd.
Nor can an agent even be granted that could destroy one quantity and generate another; for which reason God would not even be able to preserve one of those quantities while another is advening. Therefore, it does not appear reasonable to say, that a whole preceding quantity is corrupted and a whole new quantity is generated when air or another body becomes rarefied; for then it would follow that all the qualities would be corrupted no matter how much some body would become rarefied; for if quantity, which is posited as a subject of qualities, is corrupted and an accident itself is destroyed by the destruction of the immediate subject of some accident, it is necessary that all the immediate qualities inhering in that quantity be destroyed. But if you wish to say that not the whole preceding quantity is corrupted but something remains, now then that specially remains without any new part, or there is some new part which was not before. But that there may not be another new part of quantity which was not before, can not be said; since that part would not be retained in any subject, and it is manifest that it is not retained unless in some part of air or of fire; therefore, either it is then at the same time with that quantity which preceded in the same part, or that quantity recedes and no longer informs that part which it informed before; and yet it itself is not corrupted.
Since there may be no greater reason that one part of a preceding quantity is corrupted than another, it follows that that part of the quantity either remains without a subject or withdraws from that subject, in which it was before, to another; each one of these is unreasonable. Therefore, it remains that in rarefaction there is not some new part of a quantity but precisely that which was before; and yet notwithstanding this that a quantity the same numerically is at first smaller and afterwards larger; for parts of the same quantity were at first less separate in situation when the air was dense; but afterwards are more separate in situation when the air is rare. And so it is established that a thing numerically the same having parts distinct without any absolute new thing advening to it can be larger sometimes and sometimes smaller. And if this be possible of any quantity, there seems to be nothing inconsistent in conceding this of substance; for substance itself having parts sometimes becomes larger, sometimes becomes smaller, without any absolute thing advening to it.