On Generation and corruption   

parts change their places. They change their places, but not in the

same way as the parts of a revolving globe. For the parts of the globe

change their places while the whole continues to occupy an equal

place: but the parts of the rowing thing expand over an

ever-increasing place and the parts of the diminishing thing

contract within an ever-diminishing area.

It is clear, then, that these changes-the changes of that which is

coming-to-be, of that which is 'altering', and of that which is

growing-differ in manner as well as in sphere. But how are we to

conceive the 'sphere' of the change which is growth and diminution?

The sphere' of growing and diminishing is believed to be magnitude.

Are we to suppose that body and magnitude come-to-be out of

something which, though potentially magnitude and body, is actually

incorporeal and devoid of magnitude? And since this description may be

understood in two different ways, in which of these two ways are we to

apply it to the process of growth? Is the matter, out of which

growth takes place, (i) 'separate' and existing alone by itself, or

(ii) 'separate' but contained in another body?

Perhaps it is impossible for growth to take place in either of these

ways. For since the matter is 'separate', either (a) it will occupy no

place (as if it were a point), or (b) it will be a 'void', i.e. a

non-perceptible body. But the first of these alternatives is

impossible. For since what comes-to-be out of this incorporeal and

sizeless something will always be 'somewhere', it too must be

'somewhere'-either intrinsically or indirectly. And the second

alternative necessarily implies that the matter is contained in some

other body. But if it is to be 'in' another body and yet remains

'separate' in such a way that it is in no sense a part of that body

(neither a part of its substantial being nor an 'accident' of it),

many impossibilities will result. It is as if we were to suppose

that when, e.g. air comes-to-be out of water the process were due

not to a change of the but to the matter of the air being 'contained