But how are they to be harmonised if the causal series is fundamentally non-rational? Suppose yourself able to observe the development of beliefs in some alien being (say an inhabitant of Mars) as a bacteriologist observes a growing colony of microbes: suppose, further, that your observation showed how these beliefs arose from causes which had in them no tincture of reason, and that, so far as you could see, they were quite unsupported by any independent evidence which—_for you_—had weight or even meaning. Would you rate their value high? Surely not. Now it is quite true that when we examine our own system of beliefs we cannot imitate this attitude of complete detachment, since in the very act of examination some of these beliefs are assumed. But we _can_ examine the beliefs of other people, and we _do_, as a matter of common-sense practice, rate low the value of the beliefs whose sources we perceive to be non-rational. How, then, can we refuse to apply to ourselves a principle of judgment which we thus apply without scruple to our neighbours? Whenever we do so apply it, we shall, I think, be forced to admit that all creeds which refuse to see an intelligent purpose behind the unthinking powers of material nature are intrinsically incoherent. In the order of causation they base reason upon unreason. In the order of logic they involve conclusions which discredit their own premises. Nor is there, as far as I can see, any mitigation of this condemnation to be looked for except by appealing to the principle of Selection. And how far will this help us out of the difficulty? Just so far as an imitation of intelligent purpose can be a substitute for its reality, but no further. And how far is this? At first sight we might suppose that, at the worst, the cognitive series and the causal series might be harmonised on the basis of natural selection if knowledge never aspired to rise above the level which promoted race survival, if no faculties of knowing were trusted beyond the point where they ceased effectively to foster the multiplication of the species. Up to this point it would seem that, if selection be true, there is congruity between beliefs and their origin. The sequence of events which brought them into being suggests no doubt about their value. This scheme of thought, therefore, though narrowly restricted, is apparently coherent.