To the mind of the vast majority of earlier observers the phrase "foundations of the universe" would have suggested something enormously massive and solid. From what we have already seen we are prepared, on the contrary, to pass from the inconceivably large to the inconceivably small. Our sun is, as far as our present knowledge goes, one of modest dimensions. Arcturus and Canopus must be thousands of times larger than it. Yet our sun is 320,000 times heavier than the earth, and the earth weighs some 6,000,000,000,000,000,000,000 tons. But it is only in resolving these stupendous masses into their tiniest elements that we can reach the ultimate realities, or foundations, of the whole.
Modern science rediscovered the atoms of Democritus, analysed the universe into innumerable swarms of these tiny particles, and then showed how the infinite variety of things could be built up by their combinations. For this it was necessary to suppose that the atoms were not all alike, but belonged to a large number of different classes. From twenty-six letters of the alphabet we could make millions of different words. From forty or fifty different "elements" the chemist could construct the most varied objects in nature, from the frame of a man to a landscape. But improved methods of research led to the discovery of new elements, and at last the chemist found that he had seventy or eighty of these "ultimate realities," each having its own very definite and very different characters. As it is the experience of science to find unity underlying variety, this was profoundly unsatisfactory, and the search began for the great unity which underlay the atoms of matter. The difficulty of the search may be illustrated by a few figures. Very delicate methods were invented for calculating the size of the atoms. Laymen are apt to smile—it is a very foolish smile—at these figures, but it is enough to say that the independent and even more delicate methods suggested by recent progress in physics have quite confirmed them.
Take a cubic millimetre of hydrogen. As a millimetre is less than 1/25th of an inch, the reader must imagine a tiny bubble of gas that would fit comfortably inside the letter "o" as it is printed here. The various refined methods of the modern physicist show that there are 40,000 billion molecules (each consisting of two atoms of the gas) in this tiny bubble. It is a little universe, repeating on an infinitesimal scale the numbers and energies of the stellar universe. These molecules are not packed together, moreover, but are separated from each other by spaces which are enormous in proportion to the size of the atoms. Through these empty spaces the atoms dash at an average speed of more than a thousand miles an hour, each passing something like 6,000,000,000 of its neighbours in the course of every second. Yet this particle of gas is a thinly populated world in comparison with a particle of metal. Take a cubic centimetre of copper. In that very small square of solid matter (each side of the cube measuring a little more than a third of an inch) there are about a quadrillion atoms. It is these minute and elusive particles that modern physics sets out to master.
At first it was noticed that the atom of hydrogen was the smallest or lightest of all, and the other atoms seemed to be multiples of it. A Russian chemist, Mendeleeff, drew up a table of the elements in illustration of this, grouping them in families, which seemed to point to hydrogen as the common parent, or ultimate constituent, of each. When newly discovered elements fell fairly into place in this scheme the idea was somewhat confidently advanced that the evolution of the elements was discovered. Thus an atom of carbon seemed to be a group of 12 atoms of hydrogen, an atom of oxygen 16, an atom of sulphur 32, an atom of copper 64, an atom of silver 108, an atom of gold 197, and so on. But more correct measurements showed that these figures were not quite exact, and the fraction of inexactness killed the theory.