ON THE ACCURACY OF HIGH-VOLTAGE MEASUREMENTS BY X-RAY SPECTROGRAPHIC MEANS

D. N. Nasledov. Kiev.

The question of measuring high voltages is especially important for the physics of X-rays, for in many works relating to it it is extremely important to be able to measure, with great accuracy, the voltage applied to the tube. Unfortunately, this question had been developed experimentally very poorly, and only quite recently, it may be said with confidence, was it finally resolved. The present note gives a brief survey of those works that contributed to the resolution of this question.

As is known, the spectrographic measurement of voltages reduces to determining the limiting wavelength of the “white” radiation of an X-ray tube. As for the voltage applied to the tube and producing the X-ray radiation, it can be calculated from Einstein’s formula, which reads:

\[ V = \frac{12.3}{\lambda_{min}}. \]

Here \(V\) is the voltage on the tube in kilovolts, and \(\lambda_{min}\) is the limiting wavelength in Ångström units.

The investigations of Duane and Hunt [1], Hull and Rice [2], Müller [3], Wagner [4], Webster [5], and others showed that this law must be counted among the exceptionally exact laws. But the technique of these measurements is so complicated and cumbersome that in practice this method found no application. It was important, having a technical spectrograph of the Seemann or Friman, March, and Staunig type, to obtain the possibility of measuring, with great accuracy, the limiting wavelength \(\lambda_{min}\), i.e. the voltage \(V\).

In the literature we encounter a very heated controversy, conducted by Kustner and Seemann, and devoted to a discussion of the possibilities of rapid and accurate measurement of \(\lambda_{min}\). Kustner [6], on the basis of his investigations with a Seemann spectrograph, asserted that even when determining \(\lambda_{min}\) with the aid of Koch’s microphotometer, there was no possibility of measuring \(\lambda_{min}\) with an accuracy that could satisfy even a not very demanding experimenter. It turned out in particular that the fluctuations of microphotometrically measured \(\lambda_{min}\) reached 20%. Seemann [7] raised a whole series of objections to Kustner, but owing to the absence of experimental material his objections were of no great significance.

Only quite recently Glocker and Kaupp [8] carried out a whole series of exceptionally careful measurements, by which, it may be considered, the controversy comes to an end. In order to exclude the possibility of voltage fluctuations, the authors of the above-mentioned work supplied the tube from a “Stabilivolt”-type capacitor installation, the primary circuit of which they fed with current from an accumulator battery, converted

preliminarily into alternating current. The spectrograms were obtained with a Seemann spectrograph. Hartmann’s microphotometer made it possible to measure \(\lambda_{\min}\). The distances were reckoned from the \(K\beta^{IV}\) line. Spectral photographs were obtained on various kinds of photographic plates, with and without an intensifying screen. The latter is especially important, since up to now the supposition has been expressed that an intensifying screen distorts the results. In addition, the most varied exposures were used. The following table gives the results of Glocker and Kaupp’s measurements:

Table.

From this table it is evident that the fluctuations of the quantities measured are extremely small—they do not exceed \(\pm 1\%\). A further, exceptionally important conclusion is that the intensifying screen and the exposure time exert no influence whatever on the results. With different sorts of plates, the results obtained were always the same.

A very curious circumstance is that the authors of this work tried to measure \(\lambda_{\min}\) with the aid of a microscope and of an ordinary eye scale. In the first case an ordinary microscope with 20-fold magnification was used. It turned out that in this case too it is possible to measure \(V\) with an accuracy of up to \(\pm 2\%\). Again, the exposure time plays no role. In the second case (the scale) they also achieved a very high accuracy of measurement; namely, up to \(\pm 4\%\).

All this undoubtedly has very great significance for high-voltage technique, since here we have a very simple and at the same time very accurate method of measuring voltages.

LITERATURE

1) Duane and Hunt. Phys. Rev. Vol. 6, p. 166 (1915). 2) Hull and Rice. Phys. Rev. (2). Vol. 8, p. 326; Journ. of Franklin Inst. Vol. 182, p. 403. 3) Müller. Arch. sc. phys. et nat. Vol. 1, p. 127. 4) Wagner. Ann. d. Phys. 57, 401 (1918); Phys. Zeitschrift, 21, 621 (1920). 5) Webster. Proc. Nat. Acad. 2, 90 (1916); Phys. Rev. 7, 599, 1916; Proc. Nat. Acad. 3, 181 (1917). 6) Küstner. Strahlentherapie 1924, Vol. 17, H. 1; Fortschritte auf dem Geb. d. Röntgenstr. 1924, 31, S. 483. 7) Seemann. Verh. d. Röntgen-Ges. 1924, 15, S. 189. 8) Glocker und Kaupp. Strahlentherapie 1926, 22, H. 1, S. 160.