If we put a grain of indigo in a glass of water, or a grain of musk in a perfectly still room, we soon realise that molecules travel. Similarly, the fact that gases spread until they fill every "empty" available space shows definitely that they consist of small particles travelling at great speed. The physicist brings his refined methods to bear on these things, and he measures the energy and velocity of these infinitely minute molecules. He tells us that molecules of oxygen, at the temperature of melting ice, travel at the rate of about 500 yards a second--more than a quarter of a mile a second. Molecules of hydrogen travel at four times that speed, or three times the speed with which a bullet leaves a rifle. Each molecule of the air, which seems so still in the house on a summer's day, is really travelling faster than a rifle bullet does at the beginning of its journey. It collides with another molecule every twenty-thousandth of an inch of its journey. It is turned from its course 5,000,000,000 times in every second by collisions. If we could stop the molecules of hydrogen gas, and utilise their energy, as we utilise the energy of steam or the energy of the water at Niagara, we should find enough in every gramme of gas (about two-thousandths of a pound) to raise a third of a ton to a height of forty inches. I have used for comparison the speed of a rifle bullet, and in an earlier generation people would have thought it impossible even to estimate this. It is, of course, easy. We put two screens in the path of the bullet, one near the rifle and the other some distance away. We connect them electrically and use a fine time-recording machine, and the bullet itself registers the time it takes to travel from the first to the second screen. Now this is very simple and superficial work in comparison with the system of exact and minute measurements which the physicist and chemist use. In one of his interesting works Mr. Charles R. Gibson gives a photograph of two exactly equal pieces of paper in the opposite pans of a fine balance. A single word has been written in pencil on one of these papers, and that little scraping of lead has been enough to bring down the scale! The spectroscope will detect a quantity of matter four million times smaller even than this; and the electroscope is a million times still more sensitive than the spectroscope. We have a heat-measuring instrument, the bolometer, which makes the best thermometer seem Early Victorian. It records the millionth of a degree of temperature. It is such instruments, multiplied by the score, which enable us to do the fine work recorded in these pages. [Illustration: _Reproduced from "The Forces of Nature" (Messrs. Macmillan)._ A SOAP BUBBLE