The satellite structure accompanying the x-ray diagram lines $L\alpha$, $L{\beta }_1$ and $L{\beta }_2$ has been investigated for the elements Rb(37) to Sn(50). This structure is more complex than has heretofore been assumed. Not only is the number of satellites greater than previously reported, but a continuous spectrum seems to extend for some distance toward shorter wave-lengths beyond the satellites. Tables are given showing wave-lengths, values of $\frac{\nu }{R}$ and of $\frac{\Delta \nu }{R}$ ($\Delta \nu =\mathrm{difference}\mathrm{in}\mathrm{frequency}\mathrm{between}\mathrm{satellite}\mathrm{and}\mathrm{parent}\mathrm{line}$) for five (seven for some elements) satellites of $L\alpha$; four of $L{\beta }_1$; and five of $L{\beta }_2$. The empirical relation previously reported for the $K\alpha$ satellites is found to hold for these $L$ satellites: namely, the square root of the difference in frequency between a satellite and its parent line is a linear function of atomic number—similar to Moseley's law for diagram lines. The difficulties of obtaining values of the wave-lengths of satellites as accurate as are the measured wave-lengths of the corresponding diagram lines, are discussed. It is shown that some of these difficulties may be eliminated by obtaining a densitometer curve for the energy distribution through the parent line at a voltage just below the excitation voltage of the satellites. From measurements on a plate taken at high dispersion it appears that satellites may be resolved into fine-structure components.

Close

PrevNext

View figure in article