The fine structure of lithium-like 1s(2)4d states in the literature be
haves irregularly as a function of Z. The fine structures of the B III
and Ne VIII fall well below the isoelectronic curve. The term energie
s of these two systems in the data tables also give worse agreement wi
th the theoretical prediction. In this work, we show that the reason f
or this unusual situation is caused by a misidentification in the orig
inal spectra. When the correct identifications are made, the fine stru
ctures of both systems fall on the isoelectronic curve and the agreeme
nt between theory and experiment is excellent.