Hyperfine structure, Zeeman effect and isotope shift in the resonance lines of potassium.

Abstract In an earlier work (Jackson and Kuhn 1935, 1936) the hyperfine structure in the resonance lines of the abundant isotope 39 of potassium was observed by the method of absorption in an atomic beam; but no intensity measurements were made. Qualitatively, the short wave-length component appeared to be the stronger, which led to the assumption of a negative magnetic moment. Magnetic deflexion experiments (Millman 1935; Fox and Rabi 1935), though in accurate agreement as regards the width of splitting, gave a positive magnetic moment. The absorption experiments were therefore repeated under conditions which excluded overlapping of neighbouring orders of the interferometer spectrum and thus permitted a quantitative determination of the intensities. This was achieved by using an etalon of 5 cm. length only (instead of 10 cm. in the old experiment) and by running the light source at low pressure of potassium. The measurements, the main results of which were published in a preliminary note (Jackson and Kuhn 1937a), gave an intensity ratio 1·45 of the hyperfine structure components, the long wave-length one being the stronger. The nuclear spin of 3/2 which follows from the magnetic deflexion experiments would require an intensity ratio of 1·66, which is definitely outside the limits of error of the intensity measurements. It was also observed that at high density of the atomic beam the component of shorter wave-length appeared wider than the other component. Both these facts were explained quantitatively by assuming that the lines of the 14 times rarer isotope K 41 overlapped the weaker component of K 39. This assumption is in agreement with the comparatively small splitting of the ground-level of K 41 found by the magnetic deflexion method (Manley 1936), and also explains why in the earlier photograms the component of shorter wave-length appeared stronger.