In this study we present a comprehensive noble gas study of mantle xenoliths from various European Cenozoic volcanic provinces. The main body of samples is from the Massif Central, France, and the Eifel, Germany. Smaller subsets of samples are from Spitsbergen and the Graz Basin, Austria. In all the helium; neon, and argon isotopic abundances of a total of forty-five mantle xenoliths, phenocrysts, and xenocrysts were determined. The He-3/He-4-ratios within each volcanic province are very uniform, irrespective of the diverse lithologies and P-T conditions which are represented by our sample suite. Mean He-3/He-4 ratios of the Massif Central, Eifel, Spitsbergen, and Kapfenstein are 6.53 +/- .25, 6.03 +/- .14, 6.65 +/- .25, and 6.1 +/- .7 (1 sigma) times the atmospheric ratio (R(a)), respectively. The strontium and neodymium isotopic composition of some of the cpx-separates are highly variable and therefore in contrast to the uniform He signature. We thus conclude that He and probably also the other noble gases in the xenoliths are effectively decoupled from the non-volatile elements. Therefore, the He signature that is preserved in the xenoliths is actually that of their host magmas. Published strontium, neodymium, and lead isotope data of the unevolved host magmas of the xenoliths correlate well with our xenolith He data. The position of the fields of the investigated volcanic provinces in He-Sr, He-Nd, and He-Pb variation diagrams depict ternary mixtures between DMM-EM-HIMU endmembers as the source the host magmas and their volatiles. The neon isotopic composition of the gases released from the xenoliths is in most cases atmospheric and probably reflects atmospheric contamination; only a few samples reveal indications for MORE-type Ne or evidence of mass-fractionation. The Ar-40/Ar-36-ratios of the xenoliths are mostly radiogenic, with the highest ratio being 17,000. However, all samples have suffered a certain degree of atmospheric contamination. We calculate that the uncontaminated Ar-40/Ar-36-ratio Of the European subcontinental mantle is greater than or equal to 30,000. The He-4/Ar-40 ratios have been fractionated, since with values around 0.25 they are below both the current production ratio and the MORE ratio. They cannot be explained by atmospheric contamination, as the samples with high Ar-40/Ar-36 ratios also have low He-4/Ar-40 ratios. Helium and Ar were probably fractionated during initial closed system degassing of the host magma. We calculate the original unfractionated He-4/Ar-40-ratios of the Massif Central and Eifel magma source which are 2.56 +/- .18 and 2.71 +/- .17, respectively. The noble gas systematics are compatible with the notion that addition of crustal material has modified a former MORE-source signature. We use our mean He-3/He-4 ratios and reconstructed He-4/Ar-40-ratios to model the time and degree of contamination of a MORE-like source with crustal material. Combining our model with evidence from published oxygen, strontium, neodymium, and lead isotope data we judge that the source of the Massif Central and Eifel magmas became enriched in K, U, and Th shortly prior or during the Variscan orogeny (pre-collisional/collisional). In this case the enrichment in these elements is therefore most probably related to subduction of crustal material in a destructive continental margin setting.
