It is with these examples before them that investigators take the trouble to weigh a mass in a vacuum, to watch the index of a balance through a telescope at a distance of twelve feet, or use an interferometer to space out an inch into a million parts. Their one desire is to arrive at truth as nearly as they can, to bring grounds of disagreement to the vanishing point, and ensure exactness in all the computations based on their work. As art advances from plane to plane it demands new niceties of measurement, discovers sources of error unsuspected before, and avoids these errors by ingenious precautions. To-day observers earnestly wish for means of measurement surpassing those at hand. Take the astronomer for example. One would suppose that the two points of the earth’s orbit which are farthest apart, divided as they are by about 185,000,000 miles, would afford sufficient room between them for a base-line wherewith to measure celestial spaces. But the fact is otherwise. So remote are the fixed stars that nearly all of them seem unchanged in place whether we observe them on January 3 or July 3, although meanwhile we have changed our point of view by the whole length of the ellipse described by the earth in its motion. Then, too, the chemist is now concerned with analyses of a delicacy out of the question a century ago. His reward is in discovering the great influence wrought by admixtures so slight in amount as almost to defy quantitative recognition. In the experiments by M. Guillaume, elsewhere recited, his unit throughout every research was one-thousandth of a millimetre, or 1/25,400 inch. Argon, a gas about one-fourth heavier than oxygen, forms nearly one-hundredth part of the atmosphere, and yet its discovery by Lord Rayleigh dates only from 1894. His feat depended not only upon refined modes of measurement, but also upon his challenging the traditional analyses of common air. The utmost resources of refrigeration, of spectroscopy, and of measurement were required to detect four elements associated in minute quantities with argon, and of like chemical inertness. These are helium, having a density of 1.98 as compared with 16 for oxygen; neon, of 9.96 density; krypton, of 40.78; and xenon, of 64. Argon itself has a density of 19.96. “Air contains,” says Sir William Ramsay, “one or two parts of neon per 100,000, one or two parts of helium per 1,000,000, about one part of krypton per 1,000,000, and about one part of xenon per 20,000,000; these together with argon form no less than 0.937 per cent. of the atmosphere. As a group these elements occupy a place between the strongly electro-negative elements of the fluorine group, and the very positive electro-positive elements of the lithium group. By virtue of their lack of electric polarity and their inactivity they form, in a certain sense, a connecting link between the two.”[25]