This experiment was repeated with 30 milligrams of radium bromide about four months old, lent for the purpose by the writer. The emanation and CO_{2} were removed by passing them through a U tube immersed in liquid air. A practically complete spectrum of helium was observed, including the lines of wave-lengths 6677, 5876, 5016, 4972, 4713 and 4472. There were also present three other lines of wave-lengths about 6180, 5695, 5455 which have not yet been identified.
In later experiments, the emanation from 50 milligrams of the radium bromide was conveyed with oxygen into a small U tube, cooled in liquid air, in which the emanation was condensed. Fresh oxygen was added, and the U tube again pumped out. The small vacuum tube, connected with the U tube, showed at first no helium lines when the liquid air was removed. The spectrum obtained was a new one, and Ramsay and Soddy considered it to be probably that of the emanation itself. After allowing the emanation tube to stand for four days, the helium spectrum appeared with all the characteristic lines, and in addition, three new lines present in the helium obtained by solution of the radium. These results have since been confirmed. The experiments, which have led to such striking and important results, were by no means easy of performance, for the quantity of helium and of emanation released from 50 mgrs. of radium bromide is extremely small. It was necessary, in all cases, to remove almost completely the other gases, which were present in sufficient quantity to mask the spectrum of the substance under examination. The success of the experiments has been largely due to the application, to this investigation, of the refined methods of gas analysis, previously employed by Sir William Ramsay with so much skill in the separation of the rare gases xenon and krypton, which exist in minute proportions in the atmosphere. The fact that the helium spectrum was not present at first, but appeared after the emanation had remained in the tube for some days, shows that the helium must have been produced from the emanation. The emanation cannot be helium itself, for, in the first place, helium is not radio-active, and in the second place, the helium spectrum was not present at first, when the quantity of emanation in the tube was at its maximum. Moreover, the diffusion experiments, already dis-
