Mineral waters, as a rule, contain small quantities of argon mixed with oxygen, nitrogen, carbon dioxide, and in some cases sulphuretted hydrogen and helium, a gas of which more hereafter. The waters actually examined were the Bath waters, which contain much nitrogen, a little argon, and a trace of helium; the Buxton waters, containing nitrogen and a little argon; the water from “Allhusen’s Well,” Middlesborough, which evolved gas of an inflammable nature consisting mainly of nitrogen, but also containing marsh-gas, and argon to the extent of O·4 per cent; water from boiling springs in Iceland evolved gas containing somewhat more argon than air does, viz. 1·14 per cent; and lastly, water from the Harrogate sulphur springs yielded a gas largely consisting of a mixture of sulphuretted hydrogen, carbon dioxide and nitrogen, but giving also an appreciable amount of argon. Such determinations show that argon is not merely confined to the atmosphere above the earth, but that it penetrates the earth and is contained in subterraneous water. These results have been obtained by Lord Rayleigh, Professor Ramsay, Mr. Travers, and Mr. Kellas.[26] Similar experiments have been made by Dr. Bouchard in Paris[27] on effervescing waters from Cauterets in the Pyrenees. One of those springs yielded a mixture of nitrogen with a small amount of argon and helium; another yielded only nitrogen and argon; while a third gave nitrogen and helium. Such are, up to the present, the sources of argon. It has been several times stated that the element helium, which is closely allied to argon in its physical properties, and in its inertness, is a normal constituent of our atmosphere, although in very small amount. This, however, is not the case. For the argon of the atmosphere has been very carefully examined for helium in two ways. First, Lord Rayleigh dissolved a large quantity of argon in water, leaving only a minute bubble undissolved. Now while 1000 parts of water dissolve 40 parts of argon, they dissolve only 7 parts of helium; and if helium were present, it should be found in the residue. But careful spectroscopic examination failed to reveal the characteristic lines of helium. Next, Professor Ramsay and Dr. Collie diffused argon fractionally; and as the densities of argon and helium are very different (2 and 20), the helium should have been visible in the first portion which diffused. There was no trace to be detected. But, further, they submitted this portion of argon to a discharge of electricity for several hours in a vacuum-tube provided with platinum electrodes. This process, as they proved, carries out the helium with the platinum which is splashed on to the sides of the tube. The argon was then removed by pumping it off in the cold. On heating the splashed-off platinum, no helium spectrum could be observed. In a similar experiment, in which a very minute trace of helium had been added to the argon, there was no difficulty in detecting the helium on