In 1984, Raup and Sepkoski found a 26 million year periodicity in the timing of terrestrial extinctions over the last 250 million years. Nowadays, the data is much better.
We re-examined the data, using two different fossil data sets to look for
simultaneous, in-phase signals. We found one, exactly where it should be, at 27 My. (The time series for geological stages has expanded 3% in the time period RS examined.) Furthermore, it is highly regular--too regular for the orbit of Nemesis, a dark companion of the Sun hypothesized to induce the extinctions.
Calculations showed Nemesis should have an orbit which varied substantially due to orbital perturbations--and the time series analysis shows highly stable results. More detail can be found in a