Zhi-Jie Xu If one really wants to bother using fossils, then one can use fossil calibrations for other taxa where there are fossils (I think they call this secondary sources). Problem with fossil calibration is that fossils only provide MINIUMUM ages. Of course molecular theorists invent all sorts of seemingly sophisticated 'probabilities' to impose a maximum age (usually not much older than the age of the oldest fossil). But its all guesswork based on nothing. All fossil (and island age) calibrations generate minimum ages only that can falsify later events, but not any earlier. One might use tectonic correlations as an alternative.