Helium isotope shift measurements previously reported have been extended and the ${\mathrm{He}}^3$ hyperfine structure partially resolved. The hfs of the ${}_{}{}^3{}_{}{}^{}S$ levels is inverted, indicating a negative nuclear magnetic moment as expected. The splitting of the $2{}_{}{}^3{}_{}{}^{}S$ level is 0.221±0.005 ${\mathrm{cm}}^{-1\mathrm{}}$, while higher ${}_{}{}^3{}_{}{}^{}S$ levels have a splitting of about 0.216 ${\mathrm{cm}}^{-1\mathrm{}}$, agreeing exactly with the expected values. The observed hfs agrees very well with that calculated by taking into account the perturbations of neighboring fine structure levels. The stronger hfs components of transitions to the ${}_{}{}^3{}_{}{}^{}P$ and ${}_{}{}^3{}_{}{}^{}D$ levels are not resolved, but the structure must be considered in evaluating the isotope shifts. By including contributions from reduced mass and coupling effects, it is found that the predicted and observed shifts differ by ∼0.1 ${\mathrm{cm}}^{-1\mathrm{}}$ in the sense that the ${\mathrm{He}}^3$ ${}_{}{}^3{}_{}{}^{}S$ and ${}_{}{}^3{}_{}{}^{}D$ terms are higher and the ${}_{}{}^1{}_{}{}^{}S$ and ${}_{}{}^1{}_{}{}^{}D$ terms lower than expected, with respect to the ${}_{}{}^3{}_{}{}^{}P$ and ${}_{}{}^1{}_{}{}^{}P$ levels. The differences for $S$ levels show an approximately linear decrease with term value $T$ of $\sim -2.2\mathrm{}\times {10}^{-6\mathrm{}}T$. In addition, the trend of the differences for successive $D$ series members indicates an increase in the repulsion of the ${}_{}{}^3{}_{}{}^{}D$ and ${}_{}{}^1{}_{}{}^{}D$ terms near the crossing at $n=8\mathrm{}$ owing to the fact that the ${\mathrm{He}}^3$ ${}_{}{}^3{}_{}{}^{}D$ his is larger than the ${\mathrm{He}}^4$ ${}_{}{}^3{}_{}{}^{}D$ fine structure. The results suggest that the Hughes-Eckart coupling terms for the $2{}_{}{}^3{}_{}{}^{}P$ and $2{}_{}{}^1{}_{}{}^{}P$ levels are too low, by about 12 and 20 percent, respectively, when calculated with one-electron wave functions.

• Received 7 August 1950

DOI:https://doi.org/10.1103/PhysRev.82.406