20. A difference which we observe between the substance and its modifications is, that the substance is independent of the modifications, but the modifications are not independent of the substance. The substance, while remaining the same, changes its accidents, but an accident cannot change its substance and remain the same. The same block may receive different figures successively; but a figure, numerically the same, cannot pass from one block to another. Two blocks may have a similar or a different figure, whether cubic, spherical, or pyramidal, and one may take the figure of the other; but in that case, the figures are not identical, but similar, they are specifically but not numerically the same. 21. If I am asked how I know that there is only similarity and not numerical identity in the figures which bodies take successively, that there is no _permanence_ in the figures which change their subject, and consequently that the same figure cannot pass from one substance to another, in the same manner that the same substance passes from one figure to another; I shall not find it difficult to prove what I assert. There is no one who does not see what an extravagant thing it would be for a cubic figure to leave a body and pass to another. What is this figure separated from the body? How is it preserved during the transition? Why is it not exactly the same in both, but presented with slight modifications? Has it undergone a modification in its passage from one body to another? Then there would be a modification of a modification, and the figure in itself abstracted from all body, would be a kind of substance of a secondary order, permanent under modifications. These are but absurd dreams in which that is applied to the concrete which belongs to the idea only in the abstract. This transition of the forms would suppose their separate existence, and thus we might have all kinds of abstract figures, cubes, spheres, circles, triangles, etc., subsisting in themselves without application to any thing figured. 22. A still stricter demonstration of this truth is possible. If we suppose a figure, numerically the same, to pass from one body to another; the block A, which loses the cubic form, transmits it to the body B. Now, this individual form cannot be in both at the same time. Suppose that after the cubic form has left the block A, we turn it back before it has touched the body B, evidently it will not be the same in both: therefore the body B has not acquired the same, but only a similar form. It is also evident that in order to give the cubic form, we need not take it from another; therefore, the form of one is not _individually_ that of the other; otherwise we should have to say that it is and is not, that it is preserved and ceases to exist at the same time.