it breaks down we correct its conclusions from the more concrete standpoint. Taking into account the relation of the cinnabar to the self we observe that only when the vibrations caused by it affect the nerve-endings, and thereby the brain, do we perceive the coloured cinnabar. Though all observers apprehend the same cinnabar, each apprehends it in relation to a different self, and therefore, it may be, differently. There is, indeed, no fundamental difference between the contradictory perceptions of colour and the varying apprehensions of shape and size at differing distances. In both cases there is difference in the spatial relations and therefore in the causal processes involved. That this difference lies in the one case within the body and in the other case partly outside it is no fundamental difference.
But, it may be urged, this pretended solution is really an admission of the truth of the objection. Throughout the argument spatial arrangement has been assumed to possess a reality that can be accurately defined and which is known as conditioning the apprehension of colour or shape. Since the nature of the vibrations in ether is recognised as constant while the colour is individual, the former alone supplies the means of determining the nature of the cinnabar as it is in itself. In a similar manner our knowledge of the actual size of an object enables us to neutralise differences of subjective appearance.
Now there are here involved two distinct questions. First, the more general problem, which of the many qualities of bodies afford the most economical and effective means of scientifically describing them and of determining their causal relations to one another. Science has answered that question by showing how all qualitative differences are best explained by reference to the spatial and quantitative. And that has been achieved through preservation, with the least possible change, of the absolute standpoint of pure experience. The attitude carries with it no assertion as to the reality or unreality of the secondary qualities. The problem of natural science consists in following out to ideal completion (a completion possible only in thought though always in terms of actual experience) of those quantitative relations which are given us as holding between objects in space.
The second problem, that which alone concerns us in this inquiry, is as to the significance of the relative standpoint which in certain cases requires to be adopted even by the scientist, and which constitutes that modification in the absolute standpoint which I have referred to above as being the least possible. Does the relative standpoint imply that
