The 26–27-Million-Year Cycle: Extinctions, Volcanism, and a Contested Cause

Overview

Where the 2.4 Myr and 405 kyr eccentricity cycles are grounded in exact celestial mechanics, the 26–27 Myr cycle sits at the opposite end of the evidentiary spectrum: a periodicity detected purely statistically, in the timing of mass extinctions and other major geological events, whose cause remains genuinely unresolved after more than four decades of research. It is also one of the best-tested — repeatedly re-run against revised geological timescales, richer datasets, and independent event categories, and it keeps surviving.

The original discovery

The signal was first identified by David Raup and John Sepkoski in 1984. Analyzing the extinction intensity of marine animal families over the preceding 250 million years, they found 12 extinction events showing statistically significant periodicity (P < 0.01), with a mean interval of 26 million years between them — a finding immediately controversial because two of the events lined up with the terminal-Cretaceous and Late Eocene extinctions already tied to meteorite impacts. This raised the "Nemesis" hypothesis: that the Sun might have an undetected companion star in a wide, eccentric orbit, periodically disturbing the Oort Cloud and showering the inner solar system with comets.

The finding was attacked on statistical grounds almost immediately. Stigler and Wagner argued that the coarse resolution of the 1980s geological timescale — with stage boundaries themselves spaced fairly regularly — could manufacture an apparent periodicity out of essentially random extinction timing. Raup and Sepkoski countered with genus-level data (finer-resolution than families) and continued to find the same 26 Myr signal, though they acknowledged the debate needed more data to settle.

Re-confirmation with better data: 26 → 27 Myr

That better data arrived in two major waves.

Melott & Bambach (2010) reanalyzed both the Sepkoski genus compendium and the newer Paleobiology Database, using the revised 2004 geological timescale, and extended the record back to 500 million years — twice as far as the original study. They found a periodicity of 27 million years, confirmed at 99% confidence, with the shift from 26 to 27 Myr essentially just a bookkeeping consequence of the geological timescale itself having grown about 3% longer since 1984. Notably, they used the increased regularity of the signal to argue against the Nemesis hypothesis: a real companion star's orbit would have been perturbed by the many stellar encounters the Sun has had over 500 million years, which should smear the periodicity out — yet the interval stayed metronomically regular, which better fits a mechanism more stable than a distant, gravitationally jostled star.

Melott & Bambach (2013), following the 2012 revision of the international geological timescale, repeated the analysis again with even finer taxonomic resolution (genera rather than families). The 27 Myr periodicity not only survived but came through with improved statistical significance, and the excess of extinction events lining up with the periodicity's predicted maxima was confirmed at the p ≈ 0.01–0.02 level across the full Phanerozoic.

Beyond extinctions: a broader geological pulse

Independently, Michael Rampino and collaborators have spent over three decades documenting a closely related periodicity — usually reported as 26–30 Myr — not just in extinctions but across a much wider range of geological phenomena: continental flood-basalt eruptions, major plate-tectonic reorganizations, ocean-anoxic events, sea-level fluctuations, intraplate magmatism, and marine strontium-isotope excursions. A 2026 synthesis by Rampino compiling 89 major geological events over the last 260 million years found them clustering into 10 peaks spaced roughly 26 million years apart — extinctions, flood basalts, and anoxic events tending to co-occur within the same peaks rather than scattering independently, which is itself a striking result: it suggests a single underlying pacer rather than several coincidentally similar but separate cycles. The most recent estimate from this broader dataset, released in 2026, puts the period at 27.5 million years, with a much weaker secondary signal near 8.9 Myr.

The unresolved question: what causes it?

Unlike the 2.4 Myr and 405 kyr cycles, no candidate mechanism for the 26–27 Myr cycle has anything like a settled physical derivation. Three broad classes of explanation remain in competition:

1. Internal Earth processes. Deep-mantle plume cycles or convective overturns could plausibly drive periodic flood-basalt volcanism and associated extinctions on multi-million-year timescales, without needing any astronomical trigger at all.

2. Galactic-plane crossings. As the solar system orbits the galactic center, it periodically passes through the crowded mid-plane of the Milky Way's disc, which could gravitationally perturb the Oort Cloud and increase cometary bombardment — a revival, in modified form, of the old Nemesis idea, but driven by galactic structure (including, in some versions, a hypothesized thin disc of dark matter) rather than a companion star. Rampino and others have shown correlations between impact-crater ages, extinction pulses, flood basalts, and estimated galactic-plane-crossing times, with roughly 6 of 13 proposed impact pulses over 260 Myr lining up with plane crossings in one widely used model (Randall & Reece 2014).

3. Very long-period orbital/climatic forcing on the crust. A more speculative proposal is that long-term redistribution of water, ice, and sediment via slow orbital or climatic cycles subtly alters crustal and mantle stresses, encouraging episodic tectonic and volcanic activity — an indirect, climate-mediated route rather than a direct astronomical trigger.

None of these has been confirmed, and Rampino's own recent framing treats the question as explicitly open, with the resolution likely to depend on tightening the radiometric dates of the largest flood basalts and impact craters — some of which already carry uncertainties at or below a million years, small enough that a few more precise dates could start to discriminate between the competing mechanisms.

Cross-check against your own analysis

This is one of the few long cycles where independent, modern statistical methods keep landing on essentially the same number from essentially the same class of data: Raup & Sepkoski's original 26 Myr, Melott & Bambach's 27 Myr (2010, reconfirmed 2013), and Rampino's 26–27.5 Myr figure from a much broader, non-extinction-only event set. Your own CATS run on the Puetz marine-genera dataset returned 27.15 Myr, and an independent Lomb-Scargle periodogram on the same raw data reproduced it almost exactly (27.15 Myr) — sitting right in the middle of this now four-decade-long, repeatedly-revised literature consensus.

Key sources

• Raup, D.M. & Sepkoski, J.J. (1984). Periodicity of extinctions in the geologic past. PNAS 81, 801–805.

• Melott, A.L. & Bambach, R.K. (2010). Nemesis Reconsidered. MNRAS Letters 407, L99–L102.

• Melott, A.L. & Bambach, R.K. (2013). Do periodicities in extinction — with possible astronomical connections — survive a revision of the geological timescale? ApJ 773, 6.

• Rampino, M.R. & Caldeira, K.; Rampino, M.R. et al. (2019, 2021a, 2021b) — series of papers on ~26–30 Myr periodicity across flood basalts, anoxic events, sea-level change, and intraplate magmatism.

• Rampino, M.R. (2015). Disc dark matter in the Galaxy and potential cycles of extraterrestrial impacts, mass extinctions and geological events. MNRAS 448, 1816.

• Randall, L. & Reece, M. (2014). Galactic-plane-crossing model referenced in Rampino's impact/extinction correlation work.

• Rampino, M.R. (2026). The temporal characteristics of the geologic record: Global correlations and similar multi-million-year cycles of major geologic events, with potential internal-Earth versus astronomical causes — 89-event, 260 Myr synthesis giving 27.5 Myr with a secondary 8.9 Myr signal.