The Absorption of theKαLine of Carbon in Various Gases and its Dependence upon Atomic Number

A description is given of the apparatus and methods used in a photographic investigation of absorption coefficients for the $K\alpha$ line of the carbon ($\lambda =44.6\mathrm{}\mathrm{A}$) in different gases and in gold leaf. The mass absorption coefficients, $\frac{\mu }{\rho }$, found were as follows: He, 3600: C${\mathrm{O}}_2$, 4780: ${\mathrm{N}}_2$, 3850: ${\mathrm{O}}_2$, 5765: Ne, 13100: A, 45700: Kr, 31800: Xe, 6740: Au, 12500. Some of the atomic absorption coefficients (${\mu }_a$) derived from these values on the assumption that the absorption is independent of chemical combination are as follows: C, 0.44×${10}^{-19\mathrm{}}$: N, 0.89×${10}^{-19\mathrm{}}$: O, 1.52×${10}^{-19\mathrm{}}$: Ne, 4.36×${10}^{-19\mathrm{}}$: The wave-length of the $K\alpha$ line of carbon lies between the $K$ and $L_I$ absorption limits of these elements. The values of $log{\mu }_a$ when plotted against $logZ$ fall upon a straight line whose slope is 4.4. Hence in this wave-length region this should be the value of $p$ in the equation ${\tau }_a=C{\lambda }^n{Z}^p$, provided that ${\lambda }_K<\lambda <{\lambda }_L$. The results of extrapolating current x-ray formulas to this wave-length are given and the values of ${\mu }_a$ so obtained are, in general, considerably too large. Since the exponent of $Z$ in the above equation is found experimentally to be as large or larger than in the case of ordinary x-rays it appears that the proper exponent of $\lambda$ may be considerably less than 3.