The theoretical formulas for $d^{5}s$ are compared with the experimental data of Fe III. The mean deviation between theory and experiment is found to be ±852 ${\mathrm{cm}}^{-1\mathrm{}}$.

We compared the values of the parameters $B$ and $G_2$ which are required to give correctly the observed separations between certain pairs of terms. Adding to the theory a correction term proportional to $L(L+1\mathrm{})\mathrm{}$ leads to consistent values of all the radial integral parameters. The mean deviation between the theory corrected in this manner and experiment is reduced to 105 ${\mathrm{cm}}^{-1\mathrm{}}$.

• Received 8 August 1951

DOI:https://doi.org/10.1103/PhysRev.84.1089