The First Spark Spectrum of Palladium (Pd II)

An analysis of Pd II is given which differs completely from the previous one made by McLennan and Smith. The terms found include $4d^{2}D\left(4\mathrm{}{d}^9\right)$; $5s{}_{}{}^{42}{}_{}{}^{}F$, ${}_{}{}^{42}{}_{}{}^{}P$, ${}_{}{}^2{}_{}{}^{}D$, ${}_{}{}^2{}_{}{}^{}G\left(4\mathrm{}{d}^{8}5s\right)$; all the related triads from the structure $4d^{8}5p$; $6s{}_{}{}^{42}{}_{}{}^{}F$, ${}_{}{}^4{}_{}{}^{}P_3^{}{}_{}{}^{}$, ${}_{}{}^2{}_{}{}^{}G_5^{}{}_{}{}^{}\left(4\mathrm{}{d}^{8}6s\right)$; and fragments of terms due to $4d^{8}5d$. Zeeman effects of the majority of the lines have been measured and approximate $g$-values found for the low and middle sets of terms. The $g$-values depart largely from Landé's values, as do also the term intervals. Evidence is given that the term predicted by theory, $5s^{2}S\left(4\mathrm{}{d}^{8}5s\right)$, is present but it has not been fixed with respect to the other terms. The lowest term $4d^{2}D$ gives lines in the vacuum region and was found by Mr. H. E. White, of Cornell University, from plates taken there. The ionization potential is calculated from the $s{}_{}{}^4{}_{}{}^{}F$ terms as 19.8 volts from $4d^9$ to $4d^8$. As in other spectra of this type there is some indication in the limits of the component series that the Hund theory of limits is incorrect.