It is much harder to attack this position, but its weakness is best exposed by considering change as we know it directly, and comparing this with change as represented in terms of qualities. Change, when represented in terms of qualities, forms a series in which different qualities are strung together one after the other by the aid of temporal relations of before and after. The change perceived when we look at the spectrum would thus have to be described in terms of a series of colours, red before orange, orange before yellow, yellow before green, and so on. We might certainly go into greater detail than this, distinguishing any number of shades in each of the colours mentioned, but the description would still have to be given in the same form, that of a series of different colours, or shades of colour, strung together by relations of before and after. Now the fact which we know directly does not change so: it forms a continuous becoming which is not made up of any number, however great, of fixed stages. When we want to represent this changing fact in terms of qualities we have to put together a series of qualities, such as red, orange, etc., and then say that “the colour” changes from one of these to another. We pretend that there is “a colour” which is not itself either red or green or orange or blue, which changes into all these different colours one after another. It is not very difficult to see that this abstract colour which is neither red nor orange nor green nor blue is not a fact but only an abstraction which is convenient for purposes of description: it is not quite so easy to see that this criticism applies equally to each of the separate colours, red, orange, etc., and yet a little attention shows that these also are really nothing but abstractions. With reference to the whole changing fact which is known directly through any period the change in respect of colour is clearly an abstraction. But just as there is no “colour” over and above the red, the orange, the green, etc., which we say we see, so there is really no “red,” “orange,” “green,” over and above the changing process with which we are directly acquainted. Each of these, the red, the orange, and so on, just like the abstract “colour,” is simply a fictitious stage in the process of changing which it is convenient to abstract when we want to describe the process but which does not itself occur as a distinct part in the actual fact.