Five rare earth-bearing minerals found in rocks of the Chilwa Alkaline Province, Malawi, are described. Zirconolite, occurring in nepheline syenite, is unusual in being optically zoned, and microprobe analyses indicate a correlation of this zoning with variations in Si, Ca, Sr, Th, U, Fe, Nb and probably water; it is argued that this zoning is a hydration effect. A second compositional zoning pattern, neither detectable optically nor affected by the hydration, is indicated by variations in Th, Ce and Y such that, although total REE abundances are similar throughout, there appears to have been REE fractionation during zirconolite growth from relatively heavy-REE and Th-enrichment in crystal cores to light-REE enrichment in crystal rims.

Chevkinite is an abundant mineral in the large granite quartz syenite complexes of Zomba and Mulanje, and analyses are given of chevkinites from these localities. There is little variation in composition within each complex, and only slight differences between them; they are all typically light-REE-enriched. The Mulanje material was shown by X-ray diffraction to be chevkinite and not the dimorph perrierite, but chemical arguments are used in considering the Zomba material to be the same species. Other rare earth minerals identified are monazite, fluocerite and bastnäsite. These are briefly described and microprobe analyses presented.

Europium LIII-edge X-ray absorption near-edge structure (XANES) was employed to determine the Eu(II)/Eu(III) ratios in minerals. This ratio can be determined based on the peak-area ratio of white lines, the resonance peak, in normalized XANES spectra for Eu(II) and Eu(III) species. For precise determination of the Eu(II)/Eu(III) ratios, however, it was revealed that the transition probabilities for each individual Eu(II) and Eu(III) species in the system must be quantified, because we found that the peak area in normalized XANES spectra is different in each Eu(II) and Eu(III) species. Despite this ambiguity, the method was applied to Eu in natural hydrothermal apatites (Eu = 39 and 64 ppm) and fluocerite (Eu = 282 ppm). The relationship between the Eu(II)/Eu(III) ratio in these hydrothermal minerals, and the distribution coefficients of Eu(II) and Eu(III) were discussed, taking into account Eu anomalies in their REE patterns. It is considered that by combining the Eu(II)/Eu(III) ratios determined by XANES and the degree of Eu anomaly in REE patterns, we can provide new information on the distribution of Eu(II) and Eu(III) in various geochemical studies.