The Compass That Needs Light – How the European Robin Senses the Earth's Magnetic Field and What Remains Unproven

The Hook

Imagine a system whose behaviour you have been able to measure for half a century, but whose implementation you do not know.

You have a specification. It has been distilled from decades of behavioural observation; it is precise and falsifiable. You have two competing guesses about which component actually does the work. You can observe the system in operation, you can disrupt individual parts on purpose, you can remove components and put them on the test bench one at a time. But you cannot attach a debugger, cannot set a breakpoint, and cannot read a single line of source. Every statement about the implementation has to reach you indirectly – through failures, through interference experiments, through isolated parts in the lab.

This is not a developer facing an undocumented legacy system. It is sensory biology facing the magnetic compass of a European robin.

And the situation, after more than fifty years of research, is remarkably clear in the specification and remarkably open in the implementation. Hore and Mouritsen put it this way in their 2016 review: although it has been known for almost half a century that migratory birds can detect the direction of the Earth's magnetic field, the primary sensory mechanism behind it is "still unclear."

That is the genuinely interesting part of this story. Not the answer, but the way one works towards an answer one cannot see directly.

The Specification: A Compass That Works Differently From Yours

Let us begin with the finding on which everything else rests – and it is one that sounds at first like a footnote.

In 1972, Wolfgang and Roswitha Wiltschko published a short paper in Science titled "Magnetic Compass of European Robins." The finding, in translation from the abstract: the magnetic compass of European robins does not use the polarity of the magnetic field to determine the north direction. The birds derive their north direction from interpreting the inclination of the axial direction of the magnetic field lines in space – and they take as "north" that direction on the magnetic north-south axis where field lines and the gravity vector form the smaller angle.

Read that again, and translate it into engineering terms. The robin's compass is:

• axial, not polar. It yields an axis, not an arrow. From the field alone, the compass does not distinguish between "north is that way" and "south is that way."
• dependent on a second reference. Only the direction of gravity turns the axis into a direction. The bird compares two angles and takes the smaller one.
• an inclination compass, not a polarity compass. It evaluates the tilt of the field lines relative to the direction of gravity, not the polarity of the field.

This is not a minor detail but a hard constraint on every mechanistic hypothesis: every candidate must explain why the compass is blind to polarity. What response behaviour an iron-mineral-based receptor would have to show, however, is not stated by any of the sources opened here – and the obvious intuition "a compass needle points north, so it cannot be iron" is too quick. A magnetite receptor is not a freely rotating little needle; mechanically coupled or superparamagnetic particle arrangements can respond in an axially symmetric way.

What remains is this: that the compass works axially is the oldest property documented in this article, and every mechanistic hypothesis has to measure up to it. I am of the opinion that this 1972 finding is the underrated foundation of the entire research programme – and I am of the opinion that it makes iron-mineral-based explanations uncomfortable, even though it does not rule them out. No evidence for that assessment is to be found in the sources opened here.

Two Candidates, One Distinction

Into the 2000s, two biophysical hypotheses stood side by side. Zapka and colleagues summarise them in the abstract of their 2009 Nature paper as the two most promising candidates for magnetodetection:

The iron-mineral-based hypothesis. Magnetic information is said to be detected by magnetoreceptors in the upper beak and transmitted to the brain through the ophthalmic branch of the trigeminal nerve. Intuitively: a sense of touch for magnetic fields. How such a receptor would work in detail is not stated by the abstracts opened here.

The light-dependent hypothesis. Magnetic field direction is said to be sensed by radical-pair-forming photopigments in the eyes, and this visual signal to be processed in cluster N – a specialised, night-time active, light-processing forebrain region. Intuitively: a sense of sight for magnetic fields.

Two hypotheses, two entirely different signalling pathways, two different nerves. This is the situation in which an experiment can distinguish between two pathways – not by measuring the mechanism, but by cutting the wire.

That is exactly what Zapka and colleagues did. According to their abstract: European robins with bilateral lesions of cluster N showed no oriented, magnetic-compass-guided behaviour – but were still able to perform sun compass and star compass orientation. In contrast, bilateral section of the ophthalmic branch of the trigeminal nerve did not influence the birds' ability to use their magnetic compass for orientation.

The authors' conclusion, in indirect speech: cluster N is required for magnetic compass orientation in this species and may be specifically involved in processing magnetic compass information. The data strongly suggest that a vision-mediated mechanism underlies the magnetic compass of this migratory songbird, and that the putative iron-mineral-based receptors in the upper beak, connected to the brain by the trigeminal nerve, are "neither necessary nor sufficient" for magnetic compass orientation in European robins.

Note the structure of this finding, because it is the methodological core. It is a dissociation – the magnetic compass fails, the sun and star compasses remain – plus a null result on the competing pathway. If only the cluster N lesion had abolished the magnetic compass, one could object: perhaps the birds are simply disoriented in general after the procedure. That the sun and star compasses remain intact argues strongly against that objection. The trigeminal arm, in turn, weakens the suspicion that the loss is merely a consequence of surgery – although that was a different procedure on a different structure, and strictly speaking a null result there says nothing about the side effects of the forebrain lesion. Nor do the two findings jointly carry one claim; they carry two different ones: one about cluster N, the other about the beak receptor.

One limitation stated in the abstract itself: the claim holds "in this species," in European robins. It is a statement about one species, not about birds in general. I am of the opinion that this limitation usually drops out when the finding is retold.

Three years later, a paper appeared whose title addresses the beak hypothesis directly: in 2012 Treiber and colleagues published in Nature "Clusters of iron-rich cells in the upper beak of pigeons are macrophages not magnetosensitive neurons." For this article I was able to open only the bibliographically confirmed title and metadata of that paper, not its abstract or text; its substantive claim is therefore cited here explicitly as a title claim only, and not interpreted further.

The Mechanism: How Spin Is Supposed to Become a Direction

If the compass is vision-mediated and needs light, then the question arises how a molecule can measure a direction at all.

The answer that research has pursued for decades is called the radical-pair mechanism. Hore and Mouritsen in 2016 explicitly call it "the leading hypothesis" in the abstract of their review – the leading hypothesis, not the established mechanism. Its core, in translation: radical pairs, that is, magnetically sensitive chemical intermediates formed by photoexcitation of cryptochrome proteins in the retina.

A modelling paper on photoreceptor-based magnetoreception was presented in 2000 by Thorsten Ritz, Salih Adem and Klaus Schulten in the Biophysical Journal, under the title "A Model for Photoreceptor-Based Magnetoreception in Birds." A transparency note on this: in this run I was able to open only the bibliographic data and the title from Crossref, not the text. That it is a model for photoreceptor-based magnetoreception is established by the title; which specific predictions it makes I cannot document here from the primary source, and therefore do not claim.

What can be documented from the abstracts actually opened is the toolkit: the hypothesis rests on the quantum spin dynamics of photoinduced radical pairs in cryptochrome flavoproteins in the retina – so says the abstract of the 2021 Nature paper by Xu and colleagues, and so, in other words, does Bassetto and colleagues 2023 ("the quantum spin dynamics of transient radical pairs formed in cryptochrome proteins in the retina").

The following account is textbook knowledge. It has not been checked in this article against an accessible primary source – neither the electron transfer, nor the two-state system, nor the oscillation, nor the orientation-dependent product yield, nor the axial response curve derived from it. The full text of the Hore and Mouritsen review, in which this chain is worked out, was not reachable in this run. So read the next lines as an account of the model, not as physics documented in this article:

Light hits the molecule, an electron changes place, and two unpaired electron spins are left behind at two different sites in the protein. This pair can be in either of two states, and it oscillates between them. How fast and how far depends, among other things, on how the molecule is oriented relative to the external magnetic field. Which state ends up predominating determines which chemical product is formed – and thereby which signal the cell passes on.

I am of the opinion that an orientation-dependent sensor fits an axial compass well: a mechanism whose sensitivity depends on the orientation of the molecule rather than on the polarity of the field would, without additional assumptions, yield an axis and not an arrow. Evidence for this correspondence between model and behavioural finding is, however, not to be found in the abstracts opened here. Take it for the plausibility bridge that it is.

This is not, incidentally, a rhetorical genuflection. It is precisely the distinction this topic is about: between an account that is true, and an account that can be shown to be true.

The Molecule: Cryptochrome 4 in the Robin's Eye

In 2021, the paper that biophysically characterised cryptochrome 4 as a candidate appeared in Nature: "Magnetic sensitivity of cryptochrome 4 from a migratory songbird," by Jingjing Xu, Lauren E. Jarocha, Tilo Zollitsch and 33 further authors – 36 in total – with Henrik Mouritsen and P. J. Hore as last authors. Volume 594, pages 535–540.

The finding, according to the abstract: the photochemistry of cryptochrome 4 (CRY4) from the night-migratory European robin (Erithacus rubecula) is magnetically sensitive in vitro – and more so than CRY4 from two non-migratory bird species, chicken (Gallus gallus) and pigeon (Columba livia). Furthermore, site-specific mutations of ErCRY4 revealed the roles of four successive flavin–tryptophan radical pairs, both in generating magnetic field effects and in stabilising potential signalling states, in a way that could enable sensing and signalling functions to be independently optimised in night-migratory birds.

Three things about this wording deserve attention.

First: "in vitro." The measurement was made in vitro, not in the eye of a flying bird. That is not a shortcoming of the work but its method – yet it is a different claim from "the bird uses this to measure the Earth's magnetic field."

Second: the species comparison is the actual argument. That a protein responds in a magnetic field is not by itself a compass. That the photochemistry of the migratory bird's protein responds more strongly than that of two non-migrants suggests a connection with function – an inference the abstract itself does not draw. It is a correlational argument across three species.

Third: the conditional is in the original. "could enable," "potential signalling states." The abstract contains no claim that cryptochrome 4 is the receptor in the living bird. That statement is not found there, and it is not implied here.

A note on nomenclature that will matter shortly: the 2021 abstract speaks of "ErCRY4." The distinction between splice variants – ErCry4a and ErCry4b – does not appear there; it shows up in the later work we come to next. Why the spelling changed is not stated by the sources opened here; I merely observe the difference.

The Counter-Evidence: When the Model Organism's Support Fell Away

This is where the story becomes interesting – and where it becomes clear why I chose this topic for an article about evidence, and not merely for one about birds.

Bassetto and colleagues state the field's fundamental problem in the abstract of their 2023 Nature paper with remarkable candour: the exact mechanism of this light-dependent magnetic compass sense is "not fully understood." And further, verbatim: "Frustratingly, much of the supporting evidence for this theory is circumstantial, largely because of the extreme challenges posed by genetic modification of wild birds." A large part of the supporting evidence, then, is circumstantial – essentially because genetically modifying wild birds is extremely difficult.

That is precisely why the field had taken a detour. Drosophila, the fruit fly, has according to the authors been recruited as a model organism, and several influential reports of cryptochrome-mediated magnetic field effects on fly behaviour have been widely interpreted as support for a radical-pair-based mechanism in birds.

Methodologically this detour is entirely reasonable. Flies can be genetically manipulated; robins essentially cannot. If you can show that a fly lacking cryptochrome loses its magnetic field sensitivity, you hold a causal argument that cannot be obtained from the bird.

Bassetto and colleagues set out to replicate those findings. The scale: magnetic field effects on 97,658 flies moving in a two-arm maze and on 10,960 flies performing the spontaneous escape behaviour known as negative geotaxis.

The result, verbatim: "Under meticulously controlled conditions and with vast sample sizes, we have been unable to find evidence for magnetically sensitive behaviour in Drosophila." And further – this is the hard part – after reassessing the statistical approaches and sample sizes used in the studies they tried to replicate, the authors suggested that many – "if not all" – of the original results were false positives.

Their own conclusion is stated precisely: the findings cast considerable doubt on the existence of magnetic sensing in Drosophila and thus strongly suggest that night-migratory songbirds remain the organism of choice for elucidating the mechanism of light-dependent magnetoreception.

Read carefully what happened here. The same authors who in 2021 presented the biophysically characterised candidate devalued in 2023 a line of evidence that the field had widely read as support for that hypothesis: Hore and Mouritsen are authors on both papers. Whether it was their own earlier work that was torn down here – that is, whether they were authors of the replicated fly studies – is not something the sources opened here establish; the 2023 abstract says only that the reports had been "widely interpreted as support," that is, read that way by the field. I am of the opinion that this shows science at its best.

Important for the bookkeeping of evidence: the 2023 abstract does not claim that the cryptochrome hypothesis in birds has been refuted. On the contrary, it refers to it as background as "the most promising hypothesis." What falls away is not the hypothesis but the Drosophila line of evidence. The assessment that much of the supporting evidence is circumstantial stands unchanged alongside it in the same abstract.

2025: The Circle Narrows – and One Variant Becomes Unlikely

In February 2025 a paper appeared as a bioRxiv preprint titled "Cryptochrome 4b protein is likely irrelevant for the radical pair based magnetoreception in the European robin" – 16 authors, first in the list again Jingjing Xu, last again Henrik Mouritsen, posted on 27 February 2025.

A framing note first, and it is not incidental: this is a preprint. Crossref lists the record as "posted-content" with the subtype "preprint"; no peer-reviewed version could be found in this run via either Crossref or Semantic Scholar. Everything that follows is therefore the state of a manuscript that has not yet been reviewed.

The starting point, according to the paper's introduction: while ErCry4a has been characterised as a magnetosensory molecule, it is unknown whether the alternative splice variant ErCry4b could play any role in magnetoreception. So this is housekeeping: which of two variants of the same gene is the candidate?

The findings the authors bring against ErCry4b:

• Colourless in the test tube. The purified protein was "colorless"; the characteristic absorption peak at 450 nm was missing. The authors conclude from this that it did not bind FAD in vitro – or that it was not well folded in vitro. They explicitly leave both readings open.
• Simulations. Molecular dynamics calculations found FAD less stably bound in ErCry4b than in ErCry4a; the discussion speaks of an increased binding energy barrier in ErCry4b.
• Not detectable in three tissues examined. Mass spectrometry of immunoprecipitates from the retina did not detect the ErCry4b-specific peptide; the paper reports the same for cerebellum and liver.
• Considerably rarer. The discussion gives the abundance of ErCry4b in the retina (PacBio Iso-Seq) as roughly ten times lower than that of ErCry4a. The absolute counts are presented inconsistently in the paper with respect to direction and are therefore not reproduced here.

And the conclusion, verbatim from the paper's closing section: "ErCry4a remains the most likely radical pair based magnetic sensory molecule in European robins."

What convinces me about this work is not the result but how cleanly the authors bound their own reach. Two sentences from the discussion, in indirect speech: they do not yet know whether ErCry4b or other alternative splice variants are completely meaningless in other tissues. And: they cannot completely exclude ErCry4b protein expression and function.

That is, in a developer's language, an honest changelog. Not: variant b removed. But: variant b not detectable in retina, cerebellum and liver; relevance in other tissues open.

What We Do Not Know

Let us summarise what remains open after more than fifty years. Not as a list of defects, but as a map of the gap between specification and implementation.

The receptor in the living bird has not been demonstrated. What has been shown: a protein whose photochemistry responds magnetically in the test tube, more strongly in the migratory bird than in two non-migrants. What has not been shown: that precisely this protein, in the eye of a robin, supplies the information the bird orients by.

The causal test is missing – and for a technical reason. Bassetto and colleagues name it explicitly: the extreme challenges of genetically modifying wild birds. You cannot knock out the gene in a robin and then watch whether the compass fails. That is why a large part of the evidence consists, in their words, of circumstantial evidence.

The detour via the model animal has been called into question. After the 2023 replication study, the fly as a stand-in is strongly in doubt; the authors consider many of the original fly findings to be false positives.

The bridge between model and behaviour is not documented. That an orientation-dependent radical-pair sensor is what explains the axial inclination compass of 1972 is, in this article, plausibility and not evidence. It is the step at which the account most easily appears firmer than it is.

The signalling chain behind the molecule is open. Zapka and colleagues showed that cluster N is required for magnetic compass orientation in robins. How the signal gets there from a molecule in the retina is not thereby answered.

Species specificity is a real limit. The Zapka findings apply explicitly to European robins. The 2012 pigeon-beak paper concerns pigeons. Generalisations to "birds" go beyond the sources opened here.

There is also a paper I can only mention here and not interpret: in 2014 Engels and colleagues published in Nature a paper titled "Anthropogenic electromagnetic noise disrupts magnetic compass orientation in a migratory bird." Title, author list, volume 509, pages 353–356 and DOI are bibliographically confirmed; its abstract was not accessible in this run. I therefore cite the work exclusively as a title claim and draw from it no figures, no frequency ranges and no conclusions.

A Grid: Levels of Evidence in This Topic

The most useful thing to take from this story is not a fact but a sorting. The left column names the level of evidence, the middle the finding with its source, the right what it does not carry.

The right-hand column is the actual work in this table. Each row on its own sounds like progress – and is progress. But not a single row carries the sentence "birds see the magnetic field with cryptochrome 4." That sentence only comes into being when you add the rows up, and it is exactly in the adding that the error appears.

The Central Takeaway

The usable core of this story is not ornithological. It is this: a hypothesis can be simultaneously the best available one and unproven – and the supports it rests on differ greatly in load-bearing capacity. If you do not label them individually, sooner or later you will mistake a chain of circumstantial evidence for a proof.

The mechanism is the same in every domain. Four findings, each one correct and cleanly published, add up in the reader's head to a claim that none of them carries:

1. The European robin has an inclination compass and not a polarity compass: documented (Wiltschko & Wiltschko 1972). That this compass is light-dependent is presupposed as background by the more recent papers – in this article it is not documented from a finding.
2. Cluster N is required for magnetic compass orientation, the ophthalmic branch of the trigeminal nerve is not: documented, for this species.
3. Radical pairs in cryptochromes are the leading explanatory approach: documented, as a hypothesis.
4. The photochemistry of the robin's cryptochrome 4 protein is magnetically sensitive: documented, in vitro.

Sum in the reader's head: "robins navigate using quantum mechanics." Not documented – and not because any of the addends is wrong, but because the addition produces a claim that goes beyond each individual one.

From this follows a concrete course of action, and it is more useful for your work than for your birdwatching. The next time you build a justification – an architecture decision backed by benchmarks, an incident postmortem, a risk assessment, a proposal with references to studies – make the right-hand column of the table above explicit. Not just: "this evidence supports X." But, for each piece of evidence: what level is it on, and what does it explicitly not carry? Lab condition or production? One component or the system? One instance or the class? A correlation across three cases, or a causal test?

It should be noted that this procedure was not tested in any of the papers discussed here; it follows from the structure of the evidence, not from a finding. I am of the opinion that separating levels of evidence is the most rewarding transfer from this research into technical decisions – but that is a derivation, not a measurement.

And a second point, which strikes me as almost more important: the same names appear on the 2021 finding and on the 2023 null result. I am of the opinion that this increases the credibility of that research group rather than damaging its hypothesis. Whether you can do this in your team – publicly marking a line of evidence that favoured your own position as not load-bearing – is not a question of philosophy of science but of culture.

Reflection Question

Which claim that counts as settled in your team actually rests only on the addition of several correct individual findings – and which single finding would carry it, if you had it?

Fact Check

Automated AI check against abstracts and metadata of the primary sources; three papers opened bibliographically only; no human expert review.

122 checkable claims were extracted from the draft and checked against the sources opened in this run: 29 hard facts, 19 quotations, 21 source attributions and 53 synthesis claims. The check was carried out by a separate reviewer pass that received only the draft and the source list, with the brief of finding errors. Result: 64 claims confirmed, 57 corrected or weakened, 8 of those struck as unverifiable or marked as the author's own assessment. Of 53 synthesis claims, only 10 were phrased at the strength of the sources – so the errors sat almost entirely where correct individual facts had been assembled into an over-strong claim.

The three most important corrections:

1. The draft inferred from the 1972 finding that iron-mineral-based explanations were thereby undermined ("a compass needle points north, not along an axis"). None of the sources opened says what response behaviour an iron receptor would have to show – and a magnetite receptor is not a freely rotating little needle. The inference has been struck and replaced by the documentable constraint ("every candidate must explain the blindness to polarity") plus an assessment explicitly marked as opinion.
2. The draft claimed that an orientation-dependent radical-pair sensor delivers "in principle exactly what Wiltschko and Wiltschko measured in 1972." This bridge between model and behavioural finding appears in none of the sources opened; it was at the same time the article's rhetorically load-bearing connection. It is now marked as a plausibility assumption and additionally listed as an open point in the section "What We Do Not Know."
3. An entire paragraph on cryptochromes as components of the circadian clock was covered by none of the sources opened and was struck. Also corrected: the portrayal that the authors had torn down "the strongest support for their own hypothesis" – the sources establish only that the fly findings were read by the field as support, not that they originated with these authors.

Further corrections concerned, among other things: labelling the Zapka finding a "double dissociation" (it is a dissociation plus a null result); "declination compass" as a wrongly chosen counterpart to "inclination compass" (the counterpart in the abstract is polarity) and referring the tilt angle to the horizontal instead of to the gravity vector; "the two most established candidates" instead of "most promising"; the count "five findings" instead of four; the claim "the bird has a light-dependent inclination compass: documented," which contradicted the article's own species-specificity rule; several German renderings of English original quotations presented as verbatim; the missing second reading ("or not well folded") in the ErCry4b finding; and seven unmarked value judgements, now marked as the author's own assessment.

Not verifiable from primary sources were the substantive claims of Ritz et al. (2000), Treiber et al. (2012) and Engels et al. (2014) – these three papers were accessible in this run only bibliographically and are cited exclusively as a title reference or title claim. Likewise inaccessible was the full text of the Hore and Mouritsen (2016) review, which is why the spin-chemistry account is explicitly marked as an unchecked textbook presentation.

Sources and Further Reading

1. Wiltschko, W., & Wiltschko, R. (1972). Magnetic Compass of European Robins. Science, 176(4030), 62–64. https://doi.org/10.1126/science.176.4030.62 — abstract opened in full in this run.
2. Ritz, T., Adem, S., & Schulten, K. (2000). A Model for Photoreceptor-Based Magnetoreception in Birds. Biophysical Journal, 78(2), 707–718. https://doi.org/10.1016/S0006-3495(00)76629-X — opened only bibliographically in this run; abstract and text not accessible; cited exclusively as a title reference.
3. Zapka, M., Heyers, D., Hein, C. M., Engels, S., Schneider, N.-L., Hans, J., Weiler, S., Dreyer, D., Kishkinev, D., Wild, J., & Mouritsen, H. (2009). Visual but not trigeminal mediation of magnetic compass information in a migratory bird. Nature, 461(7268), 1274–1277. https://doi.org/10.1038/nature08528 — abstract opened in full in this run.
4. Treiber, C. D., Salzer, M. C., Riegler, J., Edelman, N., Sugar, C., Breuss, M., Pichler, P., Cadiou, H., Saunders, M., Lythgoe, M., Shaw, J., & Keays, D. A. (2012). Clusters of iron-rich cells in the upper beak of pigeons are macrophages not magnetosensitive neurons. Nature, 484(7394), 367–370. https://doi.org/10.1038/nature11046 — opened only bibliographically in this run; cited exclusively as a title claim.
5. Engels, S., Schneider, N.-L., Lefeldt, N., Hein, C. M., Zapka, M., Michalik, A., Elbers, D., Kittel, A., Hore, P. J., & Mouritsen, H. (2014). Anthropogenic electromagnetic noise disrupts magnetic compass orientation in a migratory bird. Nature, 509(7500), 353–356. https://doi.org/10.1038/nature13290 — opened only bibliographically in this run; cited exclusively as a title claim.
6. Hore, P. J., & Mouritsen, H. (2016). The radical-pair mechanism of magnetoreception. Annual Review of Biophysics, 45(1), 299–344. https://doi.org/10.1146/annurev-biophys-032116-094545 — abstract opened in full in this run; full text not accessible.
7. Xu, J., Jarocha, L. E., Zollitsch, T., Konowalczyk, M., Henbest, K. B., Richert, S., … Mouritsen, H., & Hore, P. J. (2021). Magnetic sensitivity of cryptochrome 4 from a migratory songbird. Nature, 594(7864), 535–540. https://doi.org/10.1038/s41586-021-03618-9 — 36 authors; abstract opened in full in this run.
8. Bassetto, M., Reichl, T., Kobylkov, D., Kattnig, D. R., Winklhofer, M., Hore, P. J., & Mouritsen, H. (2023). No evidence for magnetic field effects on the behaviour of Drosophila. Nature, 620(7974), 595–599. https://doi.org/10.1038/s41586-023-06397-7 — abstract opened in full in this run.
9. Xu, J., Ameerjan, A. B. P., Hungerland, J., Langebrake, G., Ravnsborg, T., Jensen, O. N., Schmidt, J., Bartölke, R., Kasahara, T., Satish, B., Schwigon, L., Dedek, K., Nolte, A., Liedvogel, M., Solov'yov, I. A., & Mouritsen, H. (2025). Cryptochrome 4b protein is likely irrelevant for the radical pair based magnetoreception in the European robin. bioRxiv (preprint, posted 2025-02-27). https://doi.org/10.1101/2025.02.21.639466 — not peer-reviewed; full text (introduction through methods) and metadata opened in this run, its own abstract not.

Cross-References in the Vault

• Which One of You Was It? The Duhem-Quine Problem and the Underdetermination of Theory — the philosophy-of-science foundation of this article: why an experiment never tests a single hypothesis but always a bundle of assumptions.
• The Infrastructure Beneath the Forest: Mycorrhizal Networks, the Wood Wide Web, and the Limits of Evidence — the same case in botany: a fascinating mechanism, a popular narrative, and an evidence base that does not quite carry either.
• The Clock in Every Cell: The Molecular Rhythm of Life from Fruit Fly to Chrono-Medicine — there, cryptochromes appear in a different role, and there the fruit fly was the model organism. The connection is a reading pointer, not a finding documented in this article.
• Spooky Action at a Distance: Quantum Entanglement from Einstein to the Quantum Internet — what electron spins are capable of when prepared deliberately, and why "quantum effect" alone is not yet an explanation.
• The 38 Witnesses Who Never Were: The Bystander Effect Between Laboratory Finding and Video Evidence — how a narrative outruns the evidence, and what happens when you check it.
• The Peak of Not Knowing: Why Psychology's Most Famous Effect May Be Just a Statistical Mirage — a second case in which a robust core and an over-strong popularisation come apart.

← All articles