In the process by which we are supposed to arrive at complex ideas of substances the beginning is the same as the end. 39. The same paralogism appears under a slightly altered form when it is stated (in the first letter to Stillingfleet) that the idea of substance as the ‘general indetermined idea of _something_ is by the abstraction of the mind derived from the simple ideas of sensation and reflection.’ Now ‘abstraction’ with Locke means the ‘separation of an idea from all other ideas that accompany it in its real existence.’ (Book II. chap. xii. sec. 1.) It is clear then that it is impossible to abstract an idea which is not _there_, in real existence, to be abstracted. Accordingly, if the ‘general idea of something’ is derived by abstraction from simple ideas of sensation and reflection, it must be originally given with these ideas, or it would not afterwards be separated from them. Conversely they must carry this idea with them, and cannot be simple ideas at all, but compound ones, each made up of ‘the general idea of something or being,’ and of an accident which this something supports. How then does the general idea of substance or ‘something,’ _as derived_, differ from the idea of ‘something,’ as given in the original ideas of sensation and reflection from which the supposed process of abstraction starts? What can be said of the one that cannot be said of the other? If the derived general idea is of something related to qualities, what, according to Locke, are the original ideas but those of qualities related to something? It is true that the general idea is of something, of which nothing further is known, related to qualities in general, not to any particular qualities. But the ‘simple idea’ in like manner can only be of an indeterminate quality, for in order to any determination of it, the idea must be put together with another idea, and so cease to be simple; and the ‘something,’ to which it is referred, must for the same reason be a purely indeterminate something. If, in order to avoid concluding that Locke thus unwittingly identified the abstract general idea of substance with any simple idea, we say that the simple idea, because not abstract, is not indeterminate but of a real quality, defined by manifold relations, we fall upon the new difficulty that, if so, not only does the simple idea become manifoldly complex, but just such an ‘idea of a particular sort of substance’ as, according to Locke, is derived from the derived idea of substance in general. As an idea of a quality, it is also necessarily an idea of a correlative ‘something;’ and if it is an idea of a quality in its reality, _i.e._ as determined by various relations, it must be an idea of a variously qualified something, _i.e._ of a particular substance. Then not merely the middle of the twofold process by which we are supposed to get at ‘complex ideas of substances’--_i.e._ the _abstract_ something; but its end--_i.e._ the _particular_ something--turns out