An apparently inevitable consequence of the homogeneity of conceptual space and time is the _relativity_ of spatial and temporal position. As we have seen, positions in conceptual space and time are not distinguishable until you take them in pairs. In other words, to fix one position in space or one date you have to give its relation to another position or date, and similarly to fix this you must specify a third, and so on indefinitely. To say _where_ A is means to say how you get to it from B, and B again is only known by the way it is reached from C, and so on without end. Logically, this is a simple consequence of the nature of space and time as conceptually analysed into endless series. To specify any term in the series you must give the unique relation it bears to some other term, its logical _distance_. And, in a series which has neither first nor last term, this second term cannot be defined except by its logical distance from a third. In actual perception this difficulty is avoided, owing to the fact that immediate feeling gives us the _here_ and _now_ from which all our directions are measured. But in conceptual space or time there is nothing to distinguish any one here which we may take as our “origin of co-ordinates,” or any one now which we take as our present from any other, and hence the endless regress seems inevitable. It follows, of course, that in conceptual space and time there is no principle by which to distinguish different directions. In perception they can be distinguished as right and left, up and down, and so forth. But since what is right to one percipient is left to another, in conceptual space, where complete abstraction is made from the presence of an individual percipient, there is neither right nor left, up nor down, nor any other qualitative difference between one direction and another, all such differences being _relative_ to the individual percipient. When we wish to introduce into conceptual space distinctions between directions, we always have to begin by arbitrarily assigning some standard direction as our point of departure. Thus we take, _e.g._, an arbitrarily selected line A—————B as such a standard for a given plane, and proceed to distinguish all other directions by the angle they make with A B and the _sense_ in which they are estimated (whether as from B to A or from A to B). But both the line A B and the difference of sense between A B and B A can only be defined by similar reference to some other standard direction, and so on through the endless regress.