This project sets out to quantify the degassing behaviour of semi-volatile metals from magma. Metals are naturally present in trace quantities in magma and devolatilisation of magma is the main process of concentrating them into ore deposits and releasing them to the atmosphere and hydrosphere. Thus it presents the starting point of dispersion of sometimes highly toxic, but also economically no longer dispensable, metals into our environment. Magmatic volatiles (H2O, CO2, S, Cl, F), including metals will largely be lost from the magma upon eruption. However due to the relative timing of degassing and quenching by eruption trace amounts of volatiles remain frozen in time within the melt. For this project I suggest an inverse method to study magma degassing, whereby the variations in volatile concentrations within the melt will be studied. Such a method has the advantage that it can also be applied to study the output of unobserved or prehistoric eruptions, whereby the volcanic gas has long dissipated. A combined modeling, experimental and observational approach will be used. Metal emissions, particularly in dynamic systems, depend not only on the relative equilibrium partition coefficient of a metal between gas and melt phase, but also on the availability of metal and complexing agents. Thus diffusion of metal towards a gas bubble as it grows and ascends may become an important, limiting factor in the effective emission. In this project we will model the relative contributions of diffusion and equilibrium partitioning to volcanic emissions. Model results will be tested by experimentally decompressing a rhyolitic melt. The different diffusivities of various metals and their different behaviour during degassing provide an as yet little explored toolbox to study volcanic eruptions, in past, present and future.