Hyperfine Structure in the Band Spectrum of the Mercury Hydride HgH

The structure of the spectrum of the mercury hydride has been reinvestigated with a more powerful experimental arrangement. In 1931 a complex structure due to different isotopes of mercury was discovered (Part I) and a closer study with improved apparatus led in 1935 to the establishment of a new band spectroscopic effect of the so-called nuclear isotope shift (Part II), the correlation of which with the corresponding atomic effect was explained by Bohr. Certain differences in structure between lines belonging to different branches seemed to be detectable, and these differences were subjected in this paper (Part III) to a careful study. In the bands of the system ${}_{}{}^2{}_{}{}^{}\Pi _{\frac{1}{2}}^{}{}_{}{}^{}\to {}_{}{}^2{}_{}{}^{}\Sigma$ a new (fifth) component was found on the short wave-length side for $P$ and $R$ band lines (Fig. 1). This component is gradually merging with the fourth for higher rotational energies. In the $Q$ lines a corresponding component is overlapped with the fourth for lower rotational energies and causes for higher rotational energies a deviation of the measured position of the fourth component toward too small separations. Careful measurements of separation reveal that the fifth component in $P$ and $R$ lines should be correlated to the mercury isotope 199 (Fig. 2). Since the intensity of this component is equal only to 9 percent of the total intensity of the group of components, the fifth component does not represent the whole contribution of 16.45 percent of the isotope 199. It is concluded therefore that the lines emitted by the molecules containing the isotope 199 must be split at least into two components. Thus a first proof of the existence of a hyperfine structure in band spectra is obtained. The approximate structures of groups of components corresponding to the odd isotopes 199 and 201 are obtained by subtracting the intensities corresponding to even isotopes from the intensity curves found by measuring the intensity distribution for the whole group of components (Fig. 3, a, b and c). No satisfactory explanation of the structures obtained could be found. The results of the measurements of the separation of components corresponding to even isotopes deviate from the expected ones (normal rotational isotope effect) and the deviation (difference of 27 percent in slope) cannot be explained by any known cause. An additional strong anomaly of a different kind is found for higher rotational energies in the bands of the system ${}_{}{}^2{}_{}{}^{}\Sigma _{}^I{}_{}{}^{}\to {}_{}{}^2{}_{}{}^{}\Sigma$ (Fig. 4).