The matter remained a mystery until April, 1895, when Dr. Ramsay, who was Lord Rayleigh's chemical collaborator in the discovery of argon, in examining the gas liberated by heating a specimen of Norwegian clèveite, found in its spectrum the D3 line, conspicuous and indubitable. The mineral was obtained from Dr. Hillebrand, one of our American chemists, who had previously studied it, and ascertained that it could be made to give off a gas which he identified with nitrogen. It really was nitrogen in part, but Ramsay suspected that he should also find argon, as he did and helium besides, which was unexpected.
Clèveite is a species of uraninite or pitchblende, and it soon appeared that helium could be obtained from nearly all the uranium minerals, and from many others; from many, mingled with argon; from others, nearly pure. In fact, it turns out to be very widely distributed, though only in extremely small quantities, and generally "occluded," or else in combination—seldom, if ever, free. It has been detected in meteoric iron, in the waters of certain mineral springs in the Black Forest and Pyrenees, and Kayser even reports traces of it in the atmosphere at Bonn.
It is generally obtained by heating the substance that contains it in a close vessel connected with an air pump of some kind by which the liberated gases are drawn off and collected. They are then laboriously treated to remove as far as possible all the foreign elements (nitrogen, etc.), since the presence of no more than five or ten per cent of other gases prevents the new elements from giving any spectroscopic evidence of their presence; they are too shy and modest to obtrude themselves. In many cases, as has been said, argon and helium come off together, and certain lines in their spectrum are nearly coincident, so that for a time there was supposed to be some close bond of connection between them. The latest observations, however, make it certain that this is not so: as Mr, Lockyer puts it, "argon is of the earth, earthy, but helium is distinctly celestial."
Its spectrum has been thoroughly studied by Crookes, Lockyer, and Runge, who agree as to all its leading characteristics.
Runge, whose work is the most complete and authoritative, finds that its lines have a remarkably regular arrangement, falling into two distinct "sets," each set consisting of a principal series and two subordinate ones, the lines in each series corresponding very accurately to a formula quite similar to that discovered by Balmer as governing the hydrogen spectrum.
In the whole spectrum he finds (by photography mainly, though two of the most important were detected by the bolometer) sixty-seven lines, twenty of which only are in the visible part of the spectrum. Of the sixty-seven, twenty-nine belong to
