Strange as it may seem at a first glance, it is a very similar trouble that we find with the reasoning of Aristotle. It seems strange, I say, because we are accustomed to associate with Aristotle just those things which would seem to indicate a scientific temper, and to give promise of great results: 1. Extensive accumulation of facts. Many of those works of Aristotle which remain to us are vast treasuries of facts collected from almost every field of Nature, and we have reason for thinking that he made other wonderful collections of facts which have not come down to us. His work has been a standing marvel to all time. 2. Extraordinary powers of reasoning. 3. The fact that he asserted in the strongest terms the need of building up the whole superstructure of knowledge on _experience_. And yet throughout his works, side by side with the evidences of profound knowledge and profound speculation, there are repeated instances of reasonings which are not only unsound, but altogether puerile--e. g., in the beginning of his treatise on the heavens he proves the world to be perfect by reasoning of the following kind: "The bodies of which the world is composed are solids, and therefore have three dimensions. Now, three is the most perfect number; it is the first of numbers, for of one we do not speak as a number; of two we say both; but three is the first number of which we say all; moreover, it has a beginning, a middle, and an end." That is a fair instance of his scientific incompetency. He has the facts, he is able to reason, but he does not reason _according to_ the facts; he loses sight of them and builds up great arguments on words and names. To give one more example: "He is endeavoring to explain the fact that when the sun's light passes through a hole, whatever be the form of the hole, the bright image, if formed at any considerable distance from the hole, is circular. This, of course, is easily seen to be a necessary consequence of the circular figure of the sun, if we conceive light to be diffused from the luminary by means of straight rays proceeding from every point. But Aristotle attempts to explain the fact by saying that the sun's light has a circular nature which it always tends to manifest. He employs the vague and loose conception of a circular _quality_ instead of the distinct conception of rays" (Whewell). It is a kind of reasoning which may be applied with great show of success to everything, but which really proves nothing. And so, as a matter of fact, Aristotle did not leave one single scientific generalization of value to succeeding ages.