The assistance of so competent a scientist was of the greatest value to us, and we are deeply grateful to him for having consented to take up this work. The results of the spectrum analysis brought conviction to us when we were still in doubt as to the interpretation of the results of our research. The first specimens of fairly active barium chloride containing radium, examined by M. Demarçay, exhibited together with the barium lines a new line of considerable intensity and of wave-length λ = 381·47 µµ in the ultra-violet. With the more active products prepared subsequently, Demarçay saw the line 381·47 µµ more distinctly; at the same time other new lines appeared, and the intensity of the new lines was comparable with that of the barium lines. A further concentration furnished a product for which the new spectrum predominated, and the three strongest barium lines, alone visible, merely indicated the presence of this metal as an impurity. This product may be looked upon as nearly pure radium chloride. Finally, by further purification, I obtained an exceedingly pure chloride, in the spectrum of which the two chief barium lines were scarcely visible. The following is a list, according to Demarçay, of the principal radium lines for the portion of the spectrum included between λ = 500·0 and λ = 350·0 µµ. The intensity of each line is represented by a figure, the strongest being marked 16:— λ. Intensity. 482·63 10 472·69 5 469·98 3 469·21 7 468·30 14 464·19 4 460·03 3 453·35 9 443·61 8 434·06 12 381·47 16 364·96 12 All the lines are clear and narrow, the three lines 381·47, 468·30, 434·06 are strong, and equal the most intense of those actually known. Two well-marked misty bands are also visible in the spectrum. The first, which is symmetrical, extends from 463·10 to 462·19, with a maximum at 462·75. The second, which is stronger, fades towards the ultra-violet; it begins, sharply defined, at 446·37, and passes through a maximum at 445·52; the region of the maximum extends as far as 445·34, then a nebulous band, gradually fading, extends about as far as 439. In the least refrangible part, not photographed in the spark spectrum, the only significant line is 566·5 (approx.), much more feeble, however, than 482·63. The general aspect of the spectrum is that of the metals of the alkaline earths; these metals are known to have well-marked line spectra with certain nebulous bands.