Measurement of Absorption of X-rays.\char22{}The use of a filtered beam of rays can at best give only approximate values of the coefficients, for such a beam is not homogeneous. It is pointed out that logarithmic absorption is no proof of homogeneity. In these experiments homogeneous radiation was secured by using a Bragg spectrometer and the effects of secondary radiation were avoided by placing the absorbing material between the two slits which limited the beam. By taking these and other precautions a precision of 1-2 per cent. was obtained. For wave-lengths below 0.15 \AA{}. it was found better to place the thin analyzing crystal of NaCl nearly at right angles to the beam and to use the radiation reflected from atom planes perpendicular to the ones usually used.