Newton’s principle of inertia is based upon a number of these experiments. It is by no means in the nature of a self-evident mathematical truth. This is so true that ancient thinkers believed, contrary to classical mechanics, that the movement ceases as soon as the cause of it is removed. Certain of the Greek philosophers even thought that all bodies travel in a circle, if nothing interferes with them, because the circular is the noblest of all movements. We shall see later how the principle of inertia of Einstein’s generalised mechanics has a strange affinity to this idea, and at the same time to the curious declination, the _clinamen_, which the great and profound Lucretius attributed to the free path of the atoms. But we must not anticipate. * * * * * This belief, that an object left freely to itself and not acted upon by any force preserves its velocity and direction, cannot pretend to be more than an experimental truth. But the observations on which it is based, especially those of Galileo, but any that may be imagined by physicists, could not possibly be conclusive, because in practice it is impossible to protect a moving body from every external force, such as atmospheric resistance, friction, or other. I am aware that Newton grounded his principle on astronomical as well as terrestrial observations. He noticed that, _apart from any attraction by other celestial bodies_, and as far as we can see, the planets seem to maintain their direction and velocity relatively to the vault of heaven. But Relativists think that the words I have italicised in the preceding sentence, which reflect Newton’s idea, really beg the question. His argument assumes that the planets do not circulate _freely_; that they are governed in their motions by a force which he called universal attraction. We shall see how Einstein came to think that this is not a force, and in that case the issue of the argument is very different. However that may be, the classical principle of inertia is a truth based upon (imperfect) experience, and it is therefore subject to the constant control of facts. All that we can say about it is that practically—that is to say, approximately—it harmonises with what we find. Newton did not regard it as such, not as a more or less precise approximation, but as a strict truth. That is why, when he saw that the planets do not travel in straight lines but in curved orbits, he concluded—which is a _petitio principii_—that they were subject to a central force, gravitation. That is why heavy bodies did not seem to him amenable to the mechanical laws which he had formulated for bodies left freely to themselves. That is why, in a word, Newton’s law of gravitation and his laws of dynamics are two distinct and separate things.