SINCE 1925 I have endeavoured to apply, for spectrographic work in the domain of the X-rays, the ruled glass grating, used with very high glancing angle, approaching 90°, which is particularly suited to give a high resolving power. Successively, I obtained direct diffraction and absolute measurement of wave-length for an X-ray beam of Kα. and Kβ of copper and iron (C. R. Acad. Sci., Dec. 21, 1925), and showed that the same arrangement could be utilised for extreme ultra-violet spectra (copper condensed spark spectrum between 140 A. and 3500 A.) by adjusting the ‘tangential grating’ in vacuum (Revue d'Optique, 5, 97; 1926: C. R., 182, 1141; 1926: Journal de Physique, Jan. 1927). Afterwards, using as the source of rays for the vacuum spectrograph a water-cooled metal X-ray tube and a glass grating with 30,000 lines to the inch from Prof. R. W. Wood, I have been able to obtain line spectra of soft X-rays (10 A.–100 A.) upon an ordinary photographic plate. The grating method seems, therefore, the most suitable to bridge the gap between the ultra-violet and X-rays (C. R. Acad. Sci., 185, 62; 1927: Journal de Physique, Nov. 1927) (Fig. 1). Soltan and I have registered and measured the K lines of boron (68.0 A.), carbon (44.9 A.), nitrogen (31.8 A.), oxygen (23.8 A.), the L spectrum of iron, the M lines of molybdenum (65.0 A., 54.9 A.), and a regular doublet (δλ = 3 A.) from N series of heavier elements (for tantalum, tungsten, platinum, gold ; lying between 46.8 A. and 61.4 A.) (C. R. Acad. Sci., 185, 642; 1927: Journal de Physique, Dec. 1927, in which there are numerous reproductions of spectra).