Soon after the weights of the atoms had been found a medical student named Prout in an anonymous essay called attention to the fact that there were curious numerical relationships between them. Speculation on the subject went on for many years, until in 1865 the great Russian chemist Mendeléeff published his conclusions. He had arranged the elements in the form of a table _in the order of their atomic weights_. The table consisted of twelve rows of names forming eight vertical columns, and the remarkable thing was that all those elements which fell into any particular column, although their atomic weights were very widely different, had similar properties. This enabled him to _predict_ the discovery of certain new elements, for the table contained a number of blank spaces. Three elements _have been found_ since, and their atomic weights and properties are just such as to fill three of the blank spaces. One blank space, it is thought, may be filled some day by the gas coronium, which like helium has been discovered in the sun, but unlike it has not yet been detected here. When it is, there is the place in the table which it may fill. The table then commenced with what is still called Group 1, but for reasons too complicated to explain here it appeared as if there must be a group before that, a group the chief characteristic of which would be the inactivity of the elements included in it. These were expected to be of various atomic weights, but these weights, it was anticipated, would so occur in the intervals between the others that they would all fall into a new column to the left of "Group 1." In the year 1892 Lord Rayleigh was investigating the question of the density of a number of different gases, including, so it happened, nitrogen. Now there are several ways of procuring nitrogen. One is to get it from the atmosphere by ridding it of the oxygen with which it is normally mixed. Another way is to split up some compound, such as ammonia, of which it forms a part, in such a way as to catch the nitrogen and leave the other elements with which it was combined elsewhere. Lord Rayleigh tried both ways, and he found that the nitrogen from the atmosphere was denser than that derived from ammonia. Sir William Ramsey then carried the matter a step further. He heated atmospheric nitrogen in the presence of magnesium, under which conditions some of the nitrogen combines with the latter element to form nitride of magnesium. That, it was found, made the remaining nitrogen denser still. The explanation then seemed obvious. Suppose we imagine a mixture of sawdust and iron filings: it will be heavier than an equal quantity of pure sawdust. And if we contrive to take away some of the sawdust from the mixture we shall find that what is left is heavier still, when compared with an equal bulk of pure sawdust. For it is clear that as we take away sawdust we thereby increase the proportion of the heavier iron filings and so we make the mixture heavier.