Evidence review · Pole-shift claims
A popular claim says researchers drilling near the North Pole found alternating layers of plants and ice, proof that the poles shift in cycles, and that Einstein investigated it before he died. Here is what the cores, the fossils and the letters show.
Our analysis marks our own reasoning, and the verdicts are our assessment. Everything else comes from the numbered sources.
They found both, but near the Pole they don’t alternate: 250 km from it, fern layers from a warm Arctic lie below the first debris dropped by ice [1][29]. The alternation is in Greenland, at 77°N, where sediment beneath the ice sheet holds plants from at least two ice-free spells, each followed by glaciation [4]. See sections 1 and 2.
Rock magnetism places the forests near the poles when they grew [8][25], Earth’s orbit explains why ice sheets come and go [13][14], and Antarctic ice records run unbroken for at least 800,000 years [32]. See sections 2 to 5.
He corresponded with its author from 1952 to 1955 and wrote a foreword in 1954 [23][24], but did no experiments, fieldwork or data analysis of his own, and the author later conceded Einstein’s point that polar ice is too weak to move the crust [21][28]. See section 6.
1 · The layers
Drilling near the North Pole recovered fern layers from a warm Arctic, with debris dropped by ice only in younger layers above them [1][29]. Drilling through the Greenland ice sheet recovered plant-bearing sediment from at least two ice-free spells, each followed by glaciation [4].
ACEX drilling, 2004: cores from as deep as 428 m below the seafloor [1]
Younger layers with debris dropped by ice; persistent glacial conditions only in the last 14 million years [3]
A gap: about 25 million years of record missing [1]
First ice-rafted pebbles and sea-ice diatoms, about 46–47 million years ago, some 35 million years earlier than once thought [1][2]
Layers of a freshwater fern, Azolla, about 49 million years ago, swelling and shrinking with the 41,000-year tilt cycle [2][3]
About 55 million years ago, an Arctic far warmer than earlier estimates [2]
Ferns first, ice later: the fern layers lie about 67 m below the deepest ice-rafted pebble [1], and fresh surface water was present “before the onset of ice-rafted debris” [29].
Reached the bed in 1966; the core ended in frozen sediment with plant fossils [4][27]
The 3.4 m of frozen sediment [6] holds plants from at least two ice-free spells, each followed by glaciation [4].
Core sections are schematic and not to scale. The Arctic Ocean drilling recovered about two-thirds of the interval it cored [1]; the Greenland core is broken to fit.
2 · Where the plants grew
Every fossil site lies inside today’s polar circle. Rock magnetism places the Eocene forests of Arctic Canada at 74–80°N, close to where they are now [25].
Shaded: inside the polar circle, where the Sun stays down for at least one full day each winter
Arctic Ocean floor, 88°N. Fern layers, then ice-rafted debris [1][2]
Kap København, North Greenland, 82°N. Boreal forest with mastodons and reindeer about 2 million years ago, from ancient DNA [7]
Ellesmere and Axel Heiberg islands, 78–80°N. Eocene forests [10][11][12]
Camp Century, Greenland, 77°N. Plant fossils beneath the ice sheet [4][5]
NGRIP, Greenland, 75°N. Undisturbed ice back 123,000 years [16]
West Antarctica, then about 82°S. Rainforest 92–83 million years ago [8]
Little Dome C, East Antarctica, 75°S. Ice more than 1.2 million years old; preliminary analyses point to a continuous record that far back [15]
Sites are shown at today’s positions, except the Antarctic forest, shown at about where it was 90 million years ago [8]. The Arctic Canada forests grew a few degrees south of where they are now [12][25], and the Arctic Ocean site at about 75°N, still inside the polar circle [25]. Positions are approximate.
Where the forests grew
74–80°N
where rock magnetism places the Eocene forests of Ellesmere and Axel Heiberg islands when they grew [25]
The months of polar night are inferred from where the forests grew [25]. Their wood shows seasonal growth rings [12]. One idea was that these trees shed their leaves to save carbon through the dark winter; tested on living relatives under simulated polar light at 69°N, shedding gave no carbon advantage [25][26].
Had the poles shifted, the rocks would record it: forests that grew near the equator and were later carried to the poles would show a low latitude in their magnetism. These point to 74–80°N.
3 · Why the ice comes and goes
Slow, regular changes in Earth’s orbit alter how much summer sunlight reaches high latitudes. Ice sheets grow and shrink in step [13][14]; when they shrink, plants can return, as they did at Camp Century [4].
Each cycle: ice builds for about 90,000 years, then collapses within about 10,000 [14]
The orbit stretches and rounds out over about 100,000 years. Why that cycle dominates the recent ice ages is still debated [14]
None of these cycles moves the poles across Earth’s surface. Each changes how sunlight is shared out by season and latitude [13].
4 · The ice record
Antarctic ice cores run continuously for 800,000 years [32], and preliminary results from a newer core suggest 1.2 million [15]. The pole-shift claim needs a 90° flip every 12,000 years or so [17].
Antarctica, Beyond EPICA, Little Dome C (reached bedrock January 2025): preliminary analyses indicate a continuous record back about 1.2 million years [15]
Antarctica, EPICA Dome C: 800,000 years [32]
Greenland, NGRIP: undisturbed to 123,000 years [16]
Davidson’s claimed 90° flips, one every ~12,000 years [17]: about 100 in this span
A flip every 12,000 years would have broken the 800,000-year record about 67 times, and the 1.2-million-year record about 100 times. Ice sheets do shrink and regrow in place: Camp Century’s ice melted about 416,000 years ago and later returned [5][27]. What the cores never show is a break every 12,000 years.
5 · How fast poles really move
The solid Earth does slowly tip relative to its spin axis, a process called true polar wander [21]. The pole-shift claims need motion hundreds to thousands of times faster.
About 1° or less. Fastest long-lived true polar wander in the last 320 million years [18]
About 1.5°. Today’s spin-axis drift, about 17 cm a year since 2000, if it kept going [19]
About 3°. A proposed swing, since questioned: about 12° out and back, 86–78 million years ago [18][20]
1,000°. Pole-shift hypotheses: up to 1° per thousand years [21]
About 7,500°. Davidson’s 90° flip every ~12,000 years, averaged [17]
Degrees of movement per million years (log scale)
The speed the claims require
Our analysis300–2,500×
faster than the 3° per million years that paleomagnetic data allow, at most, over the last 500 million years [31]
Earth’s magnetic field wanders and occasionally reverses, but a magnetic reversal is not a pole shift [21]. Over long spans the magnetic pole averages out to about the spin axis [22]. In the sketch, the solid line is the spin axis and the dashed line the magnetic axis.
6 · Einstein’s role
In the last two and a half years of his life he corresponded with Charles Hapgood, the theory’s author, and wrote a foreword to his book [23][24]. He was impressed, and he named the idea’s weakest point.
Tells Hapgood he finds his arguments “very impressive” and his hypothesis likely correct [23][24]
Urges him to address the “centrifugal momentum” problem [23]
Einstein dies [23]
Earth’s Shifting Crust is published with Einstein’s foreword [24]
Hapgood concedes Einstein’s point: polar ice is too weak to move the crust [21][28]
“The only doubtful assumption is that the earth’s crust can be moved easily enough over the inner layers.”
Eocene rocks from these sites whose magnetism records a low latitude
Breaks in long ice cores on a 12,000-year rhythm
Rocks recording 90° swings within thousands of years, not millions
Bracketed numbers refer to this list; “p.” gives a page of the cited book. Links checked October 11, 2026.
IODP Expedition 302 Scientists (2006). Expedition 302 summary, Arctic Coring Expedition (ACEX). Proceedings of the Integrated Ocean Drilling Program 302.
Stein, R., and Backman, J. (2011). Lower Tertiary black shales near the North Pole: organic-carbon sources, paleoenvironment and source-rock potential (IODP Expedition 302 – ACEX). Conference abstract, 3P Arctic. Alfred Wegener Institute repository.
Donders, T. H., et al. (2011). The Eocene Arctic Azolla phenomenon: cause, duration and impact on organic carbon burial. AAPG Search and Discovery, conference abstract.
Christ, A. J., et al. (2021). A multimillion-year-old record of Greenland vegetation and glacial history preserved in sediment beneath 1.4 km of ice at Camp Century. PNAS 118(13), e2021442118. Free full text.
Eos (2023). Greenland was much greener 416,000 years ago. doi:10.1029/2023EO230333
Mastro, H., et al. (2023). Sediment beneath the Greenland Ice Sheet at Camp Century preserves plant and invertebrate evidence of past climates. AGU Fall Meeting poster C11D-1069.
Kjær, K. H., et al. (2022). A 2-million-year-old ecosystem in Greenland uncovered by environmental DNA. Nature 612, 283–291. Free full text.
Klages, J. P., et al. (2020). Temperate rainforests near the South Pole during peak Cretaceous warmth. Nature 580, 81–86.
EarthSky (2020). A temperate rainforest in Antarctica during the age of dinosaurs.
Miller, K., Tietjen, K., and Beard, K. C. (2023). Basal Primatomorpha colonized Ellesmere Island (Arctic Canada) during the hyperthermal conditions of the early Eocene climatic optimum. PLOS ONE 18(1), e0280114.
Jahren, A. H. (2007). The Arctic forest of the Middle Eocene. Annual Review of Earth and Planetary Sciences 35, 509–540.
Francis, J. E. (1988). A 50-million-year-old fossil forest from Strathcona Fiord, Ellesmere Island, Arctic Canada: evidence for a warm polar climate. Arctic 41(4), 314–318. doi:10.14430/arctic1738
Hays, J. D., Imbrie, J., and Shackleton, N. J. (1976). Variations in the Earth’s orbit: pacemaker of the ice ages. Science 194(4270), 1121–1132.
Raymo, M. E., and Huybers, P. (2008). Unlocking the mysteries of the ice ages. Nature 451, 284–285.
CNR, National Research Council of Italy (2025). Antarctica: historic drilling campaign reaches more than 1.2-million-year-old ice (Beyond EPICA – Oldest Ice, Little Dome C). Press release.
North Greenland Ice Core Project members (2004). High-resolution record of Northern Hemisphere climate extending into the last interglacial period. Nature 431, 147–151.
Fremnant (June 3, 2024). The mammoth conundrum: problems with the 90-degree Earth flip theory. Catastrophism (Substack). Quotes and discusses Ben Davidson’s book Earth Disaster Cycle: The Cycle Resets Soon (2023).
Vaes, B., and van Hinsbergen, D. J. J. (2025). Slow true polar wander around varying equatorial axes since 320 Ma. AGU Advances 6, e2024AV001515.
Adhikari, S., and Ivins, E. R. (2016). Climate-driven polar motion: 2003–2015. Science Advances 2(4), e1501693. Summary: NASA JPL (April 8, 2016), NASA study solves two mysteries about wobbling Earth.
Mitchell, R. N., et al. (2021). A Late Cretaceous true polar wander oscillation. Nature Communications 12, 3629.
Wikipedia. Cataclysmic pole shift hypothesis. Accessed October 2026.
Prothero, D. R. (February 13, 2013). Magnetic myths. eSkeptic, the Skeptics Society.
Physics Forums (2006). The Einstein–Hapgood papers. A forum post copying the account of crust-displacement author Rand Flem-Ath, who obtained the letters (1952–1955) from the Einstein Archives.
Hapgood, C. H. (1958). Earth’s Shifting Crust: A Key to Some Basic Problems of Earth Science, with a foreword by Albert Einstein. Pantheon Books. Internet Archive. Einstein’s letters are printed in the appendix.
Eberle, J. J., and Greenwood, D. R. (2012). Life at the top of the greenhouse Eocene world: a review of the Eocene flora and vertebrate fauna from Canada’s High Arctic. GSA Bulletin 124(1–2), 3–23. Paleolatitudes from Irving and Wynne (1991).
Royer, D. L., Osborne, C. P., and Beerling, D. J. (2005). Contrasting seasonal patterns of carbon gain in evergreen and deciduous trees of ancient polar forests. Paleobiology 31(1), 141–150.
Christ, A. J., et al. (2023). Deglaciation of northwestern Greenland during Marine Isotope Stage 11. Science 381(6655), 330–335.
Hapgood, C. H. (1970). The Path of the Pole. Chilton Book Company. Internet Archive. See its note on the Mather and Einstein forewords.
Moran, K., Backman, J., Brinkhuis, H., et al. (2006). The Cenozoic palaeoenvironment of the Arctic Ocean. Nature 441, 601–605.
Martínez-Frías, J., Hochberg, D., and Rull, F. (2006). Contributions of Albert Einstein to Earth sciences: a review in commemoration of the World Year of Physics. Naturwissenschaften 93, 66–71.
Fu, H., Zhang, S., Condon, D. J., and Xian, H. (2022). Secular change of true polar wander over the past billion years. Science Advances 8, eabo2753.
Jouzel, J., et al. (2007). Orbital and millennial Antarctic climate variability over the past 800,000 years. Science 317(5839), 793–796. The EPICA Dome C record.
Method
On October 11, 2026, we checked every statement, number, quotation and citation on this page against its source, about 75 items in all. The check led to corrections: how the plant and ice layers sit in each core, the length of polar night at the two forest sites, the name and status of the oldest Antarctic ice core, the description of a proposed 3° polar-wander swing, the dates of Einstein’s letters, and several citations. The change log has the details.
The verdicts, the items tagged “Our analysis,” the rate conversions, the count of claimed flips and the lengths of polar night are our own work, built on the cited figures. They aren’t findings of the sources. We counted polar night as the days when the Sun doesn’t rise above the horizon even at noon; the same method gives Svalbard’s observed polar night at 78°N. Drawings marked schematic are not to scale.
The popular claim this page checks circulates in several versions. Its parts about 90° flips, polar forests and Einstein also come up in the Disaster Cycle video, which we check separately; see Related checks.
This page was researched and written with Claude, an AI assistant made by Anthropic, and then re-checked against its sources in a separate verification pass.
Corrections
If we get something wrong, we fix it and record the change below with the date. If a correction changes a verdict, the log says what the old verdict was.