130. To conclude, there is a great deal in Lucretius (whether his own or derived from others does not matter to us) which is of considerable value, even from a modern scientific point of view, though, of course, of far greater value from the point of view of the student of development. But his attempted proofs are for the most part absurd, based, as they generally are, upon mere metaphysical speculations and altogether preposterous analogies. 131. (2.) Boscovich and others endeavoured to dispense with the atom altogether, substituting in its place the conception (which mathematicians often find useful) of a mere geometrical point, which is a _centre of force_, as it is called. Here we get rid of the idea of _substance_ entirely, but we preserve (all but inertia) those external relations by which alone the atom is capable of making known its presence. Even so great an experimental philosopher as Faraday may be quoted as, to some extent at least, agreeing with this notion. It seems to us, however, that this is the embodiment of an over-refinement of speculation, surrounded on almost all sides by the gravest difficulties. It may suffice merely to mention again the property of mass, or inertia, which Faraday himself seemed to look upon as the _one_ essential characteristic of matter, and which we can hardly bring ourselves to associate with the absence of what we understand by substance. 132. (3.) Another speculation leads us to imagine matter as not ultimately atomic—as, in fact, infinitely divisible. But, if it be so, it must (in order that various elementary physical facts may be capable of explanation) be practically continuous but intensely heterogeneous. That solid or liquid matter has a grained structure of not infinitely small dimensions is proved by many simple and generally known facts; among others by the separation of white light into its constituent colours when refracted through a prism, by the phenomena of capillarity, and by those of contact electricity. If such heterogeneity were only pronounced enough, it appears that the law of gravitation would be capable of accounting for at least the greater number of effects at present attributed to the so-called molecular forces and the force of chemical affinity. Here, however, we are met by the grand difficulty, that of accounting for gravitation. And the only attempt at explanation of gravitation-attraction, which can be called even plausible, can only, with very great straining, be made compatible with this idea of the nature of matter. 133. (4.) The fourth and most recent speculation revives the atom (in the literal sense of the word), but not ‘strong in solid singleness’ like those contemplated by Lucretius,—much rather yielding to the least external force, and thus escaping from the knife or wriggling round it, so that it cannot be cut,—not, however, on account of its hardness, but on account of its mobility, which makes it impossible for the knife to get at it.