It is clear that intelligent acting involves deliberation. The almost inevitable motor response to a stimulus, which is characteristic of the reflex or instinct, does not occur in the intelligent action: instead of this we find that we choose between two or more responses to the same stimulus. We reply to the latter by doing _this_ now, and _that_ another time; and we see at once what results flow from acting differently upon the same part of our environment, or acting in the same way upon different parts. Perception, that is, knowledge of the world, arises from acting; and as our actions, when carried out intelligently, become almost infinitely varied, the environment appears to us in very many aspects. In every action we modify that part of our surroundings on which we operate. We can produce many modifications that are of no use to us: these we do not attend to. We produce others that are useful, and then we note the sequences of events involved in our actions. Thus we discover or invent natural law--an environment which is an orderly one. We can calculate and predict what will happen: we produce, for instance, a _Nautical Almanac_, at once the type of useful knowledge and of knowledge of sequences of events rigidly determined--knowledge in short that is mechanistic; and which has been engendered by the necessity for acting on our environment in our own interests. All this, the reader may note, is Bergson’s theory of intellectual knowledge, a theory which, new and paradoxical at first, becomes more and more convincing the longer we think about it, until at last it seems so obvious that we wonder that it ever seemed new. Our modes of thinking become constrained into certain grooves, just because these modes of thinking have been those that were generated by our modes of acting. So long as our thinking relates only to our acting, its exercise is legitimate. But if its object is pure speculation its results may be illusory, for a method has been applied to objects other than those for which it was evolved. Let us now extend our intellectual methods to the investigation of the organism. Necessarily we must reason about the latter as a mechanism if we reason about it at all. If it is a mechanism it must conform to the laws of energetics, for science, so far as it is quantitative, whether its results are expressed in the form of equations or inequalities, is based on these principles.