"The eruption of volcanoes appears one of the most unpredictable phenomena on Earth. Yet the situation is rapidly changing. Quantification of the eruptive record constrains what is possible in a given volcanic system. Timing is the hardest part to quantify.The main process triggering an eruption is the refilling of a sub-volcanic magma chamber by a new magma coming from depth. This process results in magma mixing and provokes a time-dependent diffusion of chemical elements. Understanding the time elapsed from mixing to eruption is fundamental to discerning pre-eruptive behaviour of volcanoes to mitigate the huge impact of volcanic eruptions on society and the environment.The CHRONOS project proposes a new method that will cut the Gordian knot of the presently intractable problem of volcanic eruption timing using a surgical approach integrating textural, geochemical and experimental data on magma mixing. I will use the compositional heterogeneity frozen in time in the rocks the same way a broken clock at a crime scene is used to determine the time of the incident. CHRONOS will aim to:1) be the first study to reproduce magma mixing, by performing unique experiments constrained by natural data and using natural melts, under controlled rheological and fluid-dynamics conditions;2) obtain unprecedented high-quality data on the time dependence of chemical exchanges during magma mixing;3) derive empirical relationships linking the extent of chemical exchanges and the mixing timescales;4) determine timescales of volcanic eruptions combining natural and experimental data.CHRONOS will open a new window on the physico-chemical processes occurring in the days preceding volcanic eruptions providing unprecedented information to build the first inventory of eruption timescales for planet Earth. If these timescales can be linked with geophysical signals occurring prior to eruptions, this inventory will have an immense value, enabling precise prediction of volcanic eruptions."