Peralkaline igneous rocks mostly occur in rift-related and intraplate settings. They form up to km-thick volcanic sequences or composite plutonic complexes of variable size. Studies of ultrabasic to alkaline rocks provide valuable information that allows to understand the origin of primary melts, their crystallization history and finally, the processes which lead to the accumulation of halogens and economically important elements, such as HFSE being typical of evolved peralkaline (agpaitic) rocks. While a lot of modern studies dealt with plutonic-subvolcanic agpaites, most volcanic agpaites are insufficiently studied. However, especially volcanic agpaites raise questions on the details of why and how degassing of halogens (and other volatiles) at such shallow emplacement levels is sufficiently prevented in order to precipitate halogen-rich agpaitic mineral assemblages. The geological, mineralogical and geochemical similarities between the volcanic rocks of the Gregory rift and Crater Highlands (Tanzania) and the deeper eroded Devonian Kola alkaline province (Russia) and the Proterozoic Gardar Alkaline Province (Greenland) have been stressed, making the Gregory rift an especially useful area to study the sub-surface processes acting in magmatic systems, leading to the formation of evolved peralkaline rocks including agpaitic rocks. The three central questions to be addressed within this project are: 1. What are the detailed genetic relationships between spatially and temporally associated primitive lavas of the Endulen area and the more evolved rocks of the Tarosero volcano? 2. What is the detailed role of volatiles (H2O, CO2, sulphur and halogens) during the differentiation history and how do they influence the evolution towards agpaitic phonolites? 3. What are the governing factors enabling a long lasting differentiation from basaltic towards phonolitic compositions, culminating in the formation of some of the most evolved igneous rock types (agpaitic phonolites) known on Earth?