Advances in X-Ray Spectrometry

Manne Siegbahn. Precision-measurements in the X-Ray Spectra. Phil. Mag. 37, p. 601 (1919).

The extensive material on X-ray spectra that has been collected up to now is of great interest in itself and, in particular, for the theory of atomic structure. However, precisely in this respect the theoreticians have somewhat outstripped the accuracy possessed by the empirical material. The author set himself the aim, by improving the details of the experimental technique and, in part, the methods of measurement, of increasing this accuracy. And he succeeded in obtaining figures approximately a hundred times more accurate than the former ones.

The apparatus which he used does not represent anything fundamentally new (a vacuum spectrograph)²). Only a few partial improvements have been introduced into the construction of the spectrograph, increasing the accuracy of the readings.

Further, especially for X-ray spectrographic work, the author has developed a type of powerful tube. It is entirely metallic, and only the glass holder into which the anticathode is inserted insulates it from the cathode. The anticathode, cathode, and the whole tube (double walls) are cooled by running water. The cathode is a calcined tungsten spiral, and the cap surrounding it is provided with a screw thread, by means of which this small cap can be raised or lowered and thereby, within wide limits, the size of the focal spot can be changed. Such a tube in some cases could operate continuously for 10–15 hours at 40–50 milliamperes (10–15 k. v.)³).

First of all, the author determined the wavelength of Cu Kα, which in many measurements plays the role of a standard. The result:

\[ 1537{,}358 \pm 0{,}033 \cdot 10^{-11}\ \text{cm}, \]

where the rock-salt constant was taken as \(2{,}81400 \cdot 10^{-8}\) cm. Dale.

¹) Phys. Zeitschr. 19, 208, 1918; 20, 127, 1919; 20, 729, 1919. Abstract—“Advances in Physical Sciences,” vol. II, issue 2, p. 287.

²) For a description see M. Siegbahn, Jahrbuch der Radioaktivität und Elektronik 13, p. 296 (1916). See also the excellent review by E. Wagner, Phys. ZS. 18, p. 495 (1917).

³) G. W. C. Kaye reports interesting results of technical successes in the manufacture of X-ray tubes of the Coolidge type in America (G. W. C. Kaye, Present-day Radiography. The Electrician No. 2177, p. 142 (1920)). Some of the experimental tubes withstood a current of \(20\ M.A.\) at 7000 V (absorbed power 18 HP) continuously for 1–3 days.

using the \(K\alpha\) lines of \(Cu\) and \(Fe\) and the \(L\alpha\) line of \(Sn\), the cadmium constant was determined:

\[ \log 2d = 0{,}7823347 \]

With the aid of \(Cu\ K\alpha_1\) and \(Pt\ L\alpha_1\), the constant of potassium ferrocyanide was determined:

\[ d = 8{,}408 \cdot 10^{-8}, \]

whereas Moseley used the value \(d = 8{,}454 \cdot 10^{-8}\), which gave a systematic error of \(0{,}54\%\).

For completeness we give the table of new values for waves.

Finally, the author studied the fine structure of the \(K\alpha\) line of copper (the doublet). For \(\Delta \lambda\) the value obtained was

\[ \Delta \lambda' = [0{,}00372 \pm 0{,}00004]\cdot 10^{-8}\ \text{cm}. \]

Comparison of this result with the figures obtained from Sommerfeld’s theory \(^{1}\) gave complete agreement.

In conclusion, a summary of formulas for the \(K\alpha\) series is given. Moseley’s simple relation no longer gives sufficient agreement with the empirical results. Sommerfeld \(^{2}\), Debye \(^{3}\), and Kroo \(^{4}\), proceeding from definite theoretical conceptions, obtained much more complicated formulas. The measurement results, as the examples given below show, speak in favor of Kroo’s formula.

E. Shpolsky.

1) A. Sommerfeld, Ann. d. Phys., 51 p. (1916).
2) A. Sommerfeld, loc. cit.
3) P. Debye, Phys. ZS 18, p. 276 (1917).
4) J. Kroo, Phys. ZS 19, p. 307 (1918).