<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Thomas Krause on Medium]]></title>
        <description><![CDATA[Stories by Thomas Krause on Medium]]></description>
        <link>https://medium.com/@krause-virtuell?source=rss-ec2b54236e52------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*DuYNP7_pf9O92otAmifedg.png</url>
            <title>Stories by Thomas Krause on Medium</title>
            <link>https://medium.com/@krause-virtuell?source=rss-ec2b54236e52------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Wed, 18 Nov 2026 18:32:53 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@krause-virtuell/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[How Prime Numbers Trigger Monstrous Moonshine]]></title>
            <link>https://medium.com/@krause-virtuell/how-prime-numbers-trigger-monstrous-moonshine-a3e6b819cf30?source=rss-ec2b54236e52------2</link>
            <guid isPermaLink="false">https://medium.com/p/a3e6b819cf30</guid>
            <category><![CDATA[string-theory]]></category>
            <category><![CDATA[mathematics]]></category>
            <category><![CDATA[geometry-of-reality]]></category>
            <category><![CDATA[physics]]></category>
            <category><![CDATA[number-theory]]></category>
            <dc:creator><![CDATA[Thomas Krause]]></dc:creator>
            <pubDate>Thu, 12 Nov 2026 10:05:04 GMT</pubDate>
            <atom:updated>2026-11-12T10:26:40.865Z</atom:updated>
            <content:encoded><![CDATA[<p><strong><em>A Mathematical Chain Reaction from a Clock Face to the j-Function</em></strong></p><p>Imagine standing in front of a massive, seemingly impenetrable cathedral of modern mathematics. High up in the rafters sits one of the strangest and most fascinating objects ever discovered: the modular j-function, a mathematical jewel mysteriously connected via so-called <em>Monstrous Moonshine</em> to the largest sporadic simple group — the “Monster,” a symmetry group with nearly 10⁵⁴ elements. Directly beneath it lie the 24-dimensional Leech lattice, where 196,560 spheres touch a single central sphere at once, and the 8-dimensional E8 lattice, which many theoretical physicists consider the geometric foundation of string theory.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*R6KvYt-e-56cs_w8UGumyg.jpeg" /></figure><p>Normally, you enter this cathedral through the main portal carrying heavy luggage from complex analysis, algebraic geometry, and representation theory. Yet a my new paper shows that there is a secret side entrance — and that to use it, you don’t need much more than a simple clock face up to 30 and the question of where the prime numbers hide on it. What is astonishing is not just that this path exists, but that once you step onto it, you barely have a choice. Once set in motion, pure arithmetic forces the walker from one structure to the next until, at the very end, the j-function stands inevitably before you.</p><p><strong>Key Takeaways</strong></p><ul><li><strong>A Geometric Clock Face as a Starting Point:</strong> All prime numbers greater than 5 land on exactly eight possible remainders when divided by 30. Geometrically, these eight numbers form the vertices of a four-dimensional object called the 16-cell.</li><li><strong>The Golden Ratio Splits the Primes:</strong> There is only one mathematical way to divide these eight remainders into two groups of four while preserving the symmetry of the 16-cell. This split leads directly to the number field of the golden ratio, where one half of the primes breaks into two pieces while the other half remains indivisible.</li><li><strong>Two Roads to the E8 Lattice:</strong> Whether you build a self-dual lattice out of the modulo-30 wheel or out of the symmetries of an icosahedron, mathematics forces the exact same 8-dimensional E8 lattice in both cases. Inside it, primes behave like standing waves — either as a double wave or a single wave, depending on which group of four they belong to.</li><li><strong>The Leech Lattice and Ramanujan’s Secret:</strong> Gluing three copies of this E8 lattice together at the prime 2 almost automatically produces the famous 24-dimensional Leech lattice. Its vibrations decompose into exactly two waves, from which Ramanujan’s famous Delta function and ultimately the j-function directly emerge.</li><li><strong>The Trail of the Prime 11:</strong> Along the entire journey, the prime number 11 repeatedly surfaces — as the fifth Lucas number, as the pacemaker for wave cancellations, and finally as a bridge to the symmetry of the Monster group.</li></ul><p><strong>First Station: The Wheel Modulo 30 and the Only Allowed Spli</strong>t</p><p>It all begins with an observation simple enough to sketch on the back of a coaster. Multiply the numbers 2, 3, and 5 together, and you get 30. Every prime number greater than 5 must leave a remainder when divided by 30 that is not divisible by 2, 3, or 5. There are exactly eight such remainders: 1, 7, 11, 13, 17, 19, 23, and 29. You can picture these eight numbers as the spokes of a wheel.</p><p>Things get even more interesting when you pair up opposite spokes so that each pair adds up to 30: 1 and 29, 7 and 23, 11 and 19, and 13 and 17. These four pairs form the four axes of a four-dimensional geometric body, the so-called 16-cell, which is the four-dimensional counterpart to an octahedron.</p><p>Now comes the first crucial question: can arithmetic sort these eight prime remainders in a meaningful way? If you look for a way to split them into two equal groups of four that is both compatible with multiplication and leaves the four axes of the 16-cell intact, out of all conceivable possibilities, only a single one remains. It is governed by quadratic residues modulo 5 and divides the eight spokes into two “great squares”: a “blue” square consisting of 1, 11, 19, and 29, and a “red” square consisting of 7, 13, 17, and 23. Behind this unique split hides a single real quadratic number field: the world of sqrt(5), and with it, the golden ratio phi = (1 + sqrt(5)) / 2.</p><p><strong>Second Station: How Primes Become Waves in the Golden Ring</strong></p><p>The moment you step into the golden ratio, something dramatic happens to prime numbers. They are no longer all alike; instead, they behave strictly according to which square of the 30-wheel they belong to. If a prime lands on the blue square — such as 11, 19, or 29 — it loses its indivisibility in the realm of the golden ratio and splits into two smaller prime ideals. If it lands on the red square — such as 7, 13, or 17 — it remains indestructible and prime.</p><p>To turn this yes-or-no event into something countable, you need a lattice. Once again, mathematics leaves no wiggle room. If you construct a number ring directly from the 30-wheel and simply demand that the resulting lattice be self-dual — meaning it is its own mirror image under a natural way of measuring length — you are inevitably forced into the eighth dimension, landing on the highly symmetric E8 lattice. Turning the 30-wheel by one spoke turns out to be the most symmetric rotation in E8, a so-called Coxeter element. Meanwhile, the number sqrt(5) hidden inside 30 cleanly slices the 240 shortest vectors of E8 into two nested four-dimensional 600-cells, whose size ratio equals the golden ratio.</p><p>It becomes even more fascinating when you view this lattice through the lens of the <em>icosians</em> — quaternions whose coordinates are built from the golden ratio and which are deeply tied to the symmetries of an icosahedron. If you count how many lattice vectors belong to a given prime p, the split from the very beginning transforms into a structure of sound and waves. A prime from the blue square generates exactly 120 * (1 + p)² vectors, whereas a prime from the red square generates 120 * (1 + p²). In the language of L-functions, this means every blue prime vibrates as a double wave with frequency log(p), while every red prime vibrates as a single wave with double the frequency. Because these waves are symmetric under reflection and do not travel away, Krause calls them <em>standing prime waves</em>.</p><p><strong>Third Station: From Gluing to the Leech Lattice</strong></p><p>However, the E8 lattice is still too simple to carry the full complexity of modular forms; it houses only a single, smooth wave known as an Eisenstein series. To capture a second, more finely structured wave, you have to climb into a higher dimension. The theory of Hilbert modular forms over the golden ratio dictates exactly where the journey must go: to rank 12 over the golden ring, which corresponds to 24 real dimensions.</p><p>So, you take three copies of the icosian E8 lattice and subtly “glue” them together along the smallest prime, 2. It turns out there are exactly five possible directions for this gluing, corresponding to the five points of a projective line over a finite field with four elements. If you pick the same direction twice, nothing new happens, and you are left with three separate copies of E8. As soon as you pick any two <em>distinct</em> directions, however, all 720 shortest vectors vanish — and out of nowhere emerges the legendary 24-dimensional Leech lattice, the densest and most symmetric sphere-packing lattice in its dimension.</p><p><strong>Fourth Station: Two Waves, Ramanujan, and the Node Mystery of 11</strong></p><p>Looking at the Leech lattice through the spectacles of the golden ratio reveals an astonishing simplicity. The space of its associated modular forms is only two-dimensional, meaning the entire geometry of the Leech lattice can be written as a superposition of exactly two standing prime waves: a smooth Eisenstein wave and a highly intricate cusp wave. Here, too, the original split of the 30-wheel makes its presence felt. On the blue spokes, the cusp wave always pushes the count of lattice points downward, while on the red spokes, it oscillates wildly between positive and negative values.</p><p>If you set the two variables of these golden waves equal to each other — projecting the whole picture back down to the ordinary rational numbers — a small miracle occurs. The two golden waves transform into the two most famous classical modular forms of weight 12: the Eisenstein series E_12 and Ramanujan’s discriminant Delta. Ramanujan’s mysterious tau(n) function reveals itself simply as the average value of the golden cusp wave along a straight trace line.</p><p>But where does this mysterious cusp wave actually come from? It is born via the so-called Doi–Naganuma lift from a classical modular form of level 5 whose coefficients live in the imaginary quadratic field Q(sqrt(-11)). Why 11 of all numbers? Here, Krause uncovers a general law he calls the <em>Lucas rule</em>. Lucas numbers are the siblings of Fibonacci numbers and arise as the sum of two opposing waves built from the golden ratio. Over the real numbers, these two waves never cancel out. However, if you compute modulo a prime number, special “nodes” appear where the two waves add up to exactly zero. The fifth Lucas number is L_5 = 11 — and that is precisely why 11 acts as the pacemaker in the coefficients of the cusp wave. In fact, this single prime 11 can be traced like a red thread through the entire construction: from how it splits in the golden ring, to fixed points of torus maps on the Leech lattice, all the way to a famous congruence of the j-function modulo 11.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*iCVrZePXaJxPcsrboocGQA.png" /></figure><p><strong>The Finale: Why the Road Ends Right Here</strong></p><p>At the end of the chain, you divide the theta series of the Leech lattice by Ramanujan’s Delta function. What pops out is j — 720 — directly giving Klein’s modular j-invariant, whose series expansion starts with the number 196,884 = 196,883 + 1, the famous clue that once put John McKay on the trail of the Monster group.</p><p>That this journey ends specifically in dimension 24 is no coincidence; it is the final forced link in the chain. Only for a lattice of rank 24 does the symmetry of the theta series match the 24 oscillators of the Dedekind eta function in the denominator, yielding a true, invariant modular function. And even beyond that, a door opens toward theoretical physics: the 24-dimensional Leech lattice sits as a light-like slice inside the 26-dimensional Lorentzian lattice II_{25,1} — the exact spacetime dimension in which bosonic string theory works without contradictions. What began as merely counting prime remainders on a 30-hour clock face ends right on the threshold of the deepest symmetry in mathematics.</p><p><strong>Read and check</strong></p><p>Standing Prime Waves on Icosian Lattices: From the Units Modulo 30 via the Leech Lattice to the j-Function and Moonshine:</p><p><a href="https://www.preprints.org/manuscript/202611.1271">Standing Prime Waves on Icosian Lattices: From the Units Modulo 30 via the Leech Lattice to the j -Function and Moonshine</a></p><p>See also in my Geometry of Reality — Book series:</p><ul><li><a href="https://www.amazon.com/dp/B0HL3W32DS/">Geometry of Reality: The Shape of Numbers: An Introduction to Quantum Number Theory</a></li><li><a href="https://www.amazon.com/dp/B0H7X876KM">Geometry of Reality - The Prime Machine:: From the Matchstick Game to the Half-Line</a></li><li><a href="https://www.amazon.com/dp/B0GZKGNPF4">The Geometry of Reality: From the First Tick to the Elementary Particles</a></li></ul><p>Thomas Krause is an independent researcher based in Lübeck, Germany, whose work explores the deep borderlands between pure number theory, higher-dimensional geometry, and theoretical physics. His central research question is how fundamental structures of our universe — from the distribution of prime numbers and classical constants like the golden ratio to root lattices like E8 and the behavior of elementary particles — emerge naturally from discrete geometric rules and symmetries.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a3e6b819cf30" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Counting Trees in Four Dimensions: A Lattice Constant That Turns Out to Be Built From Primes]]></title>
            <link>https://medium.com/@krause-virtuell/counting-trees-in-four-dimensions-a-lattice-constant-that-turns-out-to-be-built-from-primes-09cfff9ccb27?source=rss-ec2b54236e52------2</link>
            <guid isPermaLink="false">https://medium.com/p/09cfff9ccb27</guid>
            <category><![CDATA[prime-numbers]]></category>
            <category><![CDATA[number-theory]]></category>
            <category><![CDATA[mathematics]]></category>
            <category><![CDATA[physics]]></category>
            <category><![CDATA[spanning-tree]]></category>
            <dc:creator><![CDATA[Thomas Krause]]></dc:creator>
            <pubDate>Wed, 04 Nov 2026 16:29:47 GMT</pubDate>
            <atom:updated>2026-11-04T17:46:38.340Z</atom:updated>
            <content:encoded><![CDATA[<p><strong>How many ways can you connect every point of a four-dimensional grid without making a loop? The answer seems to be a combination of two numbers from the theory of modular forms. It holds to more than 100 digits, and the proof is down to one missing piece.</strong></p><p><strong>Key message: </strong>The paper demonstrates that the exact number of loophole-free paths (spanning trees) in a four-dimensional grid can be calculated precisely through a highly symmetric mathematical structure known as a <strong>weight-4 modular form</strong>. The Asymptotic Lattice Constant in 4D for Spanning Trees is 1.999707644517312559687899304…</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ctlTZW6Q-s9N6_EiGZJsnA.png" /><figcaption>A random spanning tree on a square grid. In four dimensions, the growth rate of the number of such trees appears to be 12 L′(g₆, 0) − (56/3) ζ′(−2), where g₆ is a modular form of level 6</figcaption></figure><p>Take a grid of points and connect neighbours by edges. A spanning tree is a way of choosing some of these edges so that every point is reached and no loop is closed. On a large grid there are astronomically many such trees, and their number grows exponentially with the number of points N, roughly like e^(λN). The constant λ is the entropy of the lattice.</p><p>Physicists know this number well. It is the free energy of a limiting case of the Potts model, the entropy of the abelian sandpile, and the normalisation of the uniform random spanning tree. Gustav Kirchhoff showed in 1847 how to count spanning trees with a determinant ([Kirchhoff 1847](<a href="https://doi.org/10.1002/andp.18471481202">https://doi.org/10.1002/andp.18471481202</a>)), and today λ is known to many digits for many lattices ([Felker and Lyons 2003](<a href="https://doi.org/10.1088/0305-4470/36/31/301">https://doi.org/10.1088/0305-4470/36/31/301</a>)). The question in my new paper (P15) is not how large λ is, but which number it is.</p><p><strong>Where the idea came from</strong></p><p>In an earlier paper of this series ([Krause 2026, P9](<a href="https://doi.org/10.5281/zenodo.23805526">https://doi.org/10.5281/zenodo.23805526</a>)) I collected the entropies of flat lattices. For the square grid, λ = 4G/π, where G = 0.91596… is Catalan’s constant. For the triangular grid the corresponding constant comes from the field of √−3. In both cases a single periodic pattern of +1, 0 and −1, a Dirichlet character, decides which constant appears. In the language of number theory these are L-values of weight two.</p><p>In three dimensions, Guttmann and Rogers proved formulas of the next kind ([Guttmann and Rogers 2012](<a href="https://doi.org/10.1088/1751-8113/45/49/494001">https://doi.org/10.1088/1751-8113/45/49/494001</a>)). The simple cubic and the body-centred cubic lattices have entropies that are values of L-functions of modular forms of weight three.</p><p>So there is a ladder: weight two in the plane, weight three in space. The natural question was whether it goes on. Is the spanning-tree entropy of a four-dimensional lattice an L-value of weight four?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*9r9lNS5E5Q0RdIkA2hP-tA.png" /><figcaption>The known spanning-tree entropies. Each step up in dimension raises the weight of the modular form by one. Green: proved. Orange: the new relation for the hypercubic lattice ℤ⁴, numerical</figcaption></figure><p><strong>What we found</strong></p><p>There are two standard lattices in four dimensions. For the hyper-body-centred lattice the answer follows from a theorem of Papanikolas, Rogers and Samart ([PRS 2014](<a href="https://arxiv.org/abs/1305.2143">https://arxiv.org/abs/1305.2143</a>)), once it is written in the right form: the entropy is 8 L′(f₈, 0) − 28 ζ′(−2), with a modular form f₈ of weight four and level eight.</p><p>For the hypercubic lattice ℤ⁴, the ordinary four-dimensional grid, no formula was known. The first step was to find the right modular form. Among 464 candidate forms of weight four, a simple test with the number of closed random walks modulo primes left exactly one survivor: the form</p><p>g₆ = (η(τ) η(2τ) η(3τ) η(6τ))²,</p><p>the unique newform of weight four and level six. Here η is Dedekind’s eta function. Then an integer-relation search with a basis fixed in writing before the computation gave</p><p>**λ(ℤ⁴) = 12 L′(g₆, 0) − (56/3) ζ′(−2) = 1.99970764451731255968…**</p><p>Equivalently, λ(ℤ⁴) = (162/π⁴) L(g₆, 4) + (14/3) ζ(3)/π². The two sides agree to more than 100 digits, in two independent runs, and every other basis element got the coefficient zero. I did not find this relation in the literature I checked. It is not proved.</p><p>What it says in words: a constant that physics defines without any reference to primes is a rational combination of two numbers built from primes. One is the value of the L-function of g₆, whose coefficients are an Euler product over primes. The other is ζ′(−2) = −ζ(3)/(4π²), Apéry’s constant in disguise.</p><p><strong>How far the proof got</strong></p><p>A numerical identity to 100 digits is strong evidence, but it is not a proof. Most of the paper is about how far a proof can be pushed, and where exactly it stops.</p><p>The four-dimensional integral can be moved onto a single line in the upper half-plane, where g₆ and its relatives live. On that line the entropy splits into a part that is proved and one remaining integral K that runs from a special point to the boundary. The special point, τ* = 1/2 + i/(2√3), is a so-called CM point of the field of √−3, and it is the fixed point of a symmetry W₃ that acts like a mirror, Y ↦ 1/(12Y).</p><p>Under that mirror K splits into an even half and an odd half. The even half is now proved. The odd half is equivalent to the whole relation, and a lemma shows that no symmetry of the group can reach it. The missing step is therefore genuinely arithmetic. The entire problem has been reduced to one number:</p><p>K = 12 L′(g₆, 0) + (28/3) ζ′(−2) = 1.14715084688…</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*HMtkI3DcIjii7qxf_Fyx-g.png" /><figcaption>From a four-dimensional integral to one statement about the odd half of an integral at a CM point.</figcaption></figure><p>Two things did not work, and they are reported as negative results: the ladder does not continue for the four-dimensional face-centred lattice, and nothing of this kind was found for the five-dimensional body-centred lattice. The ladder is not a general law. It continues where the lattice has enough symmetry for a modular description.</p><p><strong>How this can help mathematicians</strong></p><ul><li><em>A small, precise open problem. </em>Number theorists now have one explicit iterated integral from a CM point to a cusp, with a conjectured value. Any method that evaluates such integrals in the odd sector can be tested on it. Candidates are Hecke correspondences, higher Green functions, and Brown’s non-holomorphic modular forms ([Brown 2018](<a href="https://doi.org/10.1007/s40687-018-0130-8">https://doi.org/10.1007/s40687-018-0130-8</a>)).</li><li><em>A new member of an open family</em>. The Mahler measures m(k − Σ(xᵢ + 1/xᵢ)) were listed as open by Papanikolas, Rogers and Samart. The case k = 8, which is λ(ℤ⁴), now has a conjectured closed form.</li><li><em>A new integral formula.</em> The paper proves an integral representation of the non-critical value L(g₆, 4), the kind of number that Beilinson’s conjectures describe as a regulator.</li><li><em>A pattern to look for elsewhere.</em> “Full path proved, half path open” should appear for other Calabi–Yau differential equations whose special point is a CM point fixed by such a mirror.</li></ul><p><strong>How this can help physicists</strong></p><ul><li><em>An exact free energy in the upper critical dimension</em>. Four is the upper critical dimension of the loop-erased random walk and the uniform spanning tree ([Lawler, Sun and Wu 2019](<a href="https://doi.org/10.1214/19-AOP1349">https://doi.org/10.1214/19-AOP1349</a>)). If the relation is proved, the spanning-tree free energy there has a closed form.</li><li>**A bridge to Feynman integrals.** The same form g₆ governs the six-Bessel moments of two-dimensional quantum field theory, the four-loop sunrise or “banana” integrals ([Broadhurst 2016](<a href="https://doi.org/10.4310/CNTP.2016.v10.n3.a3">https://doi.org/10.4310/CNTP.2016.v10.n3.a3</a>)). There its critical values appear, here the non-critical one. The tools used in one problem may help in the other.</li><li>**Lattice Green functions.** λ is an integrated Green function of the random walk on ℤ⁴, and the same level-six description controls the return probabilities. The methods carry over.</li><li>**A fast benchmark.** λ(ℤ⁴) can now be computed to 100 digits from a quickly converging q-series instead of a four-dimensional integral.</li></ul><p><strong>What this is not</strong></p><ul><li>The main relation is a conjecture with very strong numerical support, not a theorem.</li><li>Several steps use one cited input, a symmetry of a modular function proved by Shvets ([Shvets 2026](<a href="https://arxiv.org/abs/2604.06239">https://arxiv.org/abs/2604.06239</a>)), which I have not re-proved independently.</li><li>No physical prediction is made. The value of λ was known; the paper only asks which number it is.</li></ul><p><strong>A reading within the series</strong></p><p>This last part is an interpretation within my framework, not a result, and the mathematics does not depend on it.</p><p>The level of g₆ is 6 = 2 · 3, the second step of the primorial ladder 2, 6, 30 that organises earlier papers of this series. The special point belongs to the field of √−3, the same field that appeared in the triangular lattice. The factor 56/3 contains 7 = 8(1 − 2⁻³), the Euler factor at 2 of ζ(3), which keeps returning in this series, from electrons in metals to neutrino decoupling. The eight neighbours of a point in ℤ⁴ are the eight vertices of the 16-cell, the four-dimensional polytope on which earlier papers placed the residues modulo 30. And the mirror W₃ has exactly one fixed point; the half of the proof that is still open is precisely the half the mirror cannot reach.</p><p>If this reading were right, the four-dimensional grid would not just happen to produce these numbers. Its geometry would carry the same small primes 2 and 3 that organise the other papers, and the open half of the proof would be a place where a reflection alone is not enough. Whether that is more than a coincidence of small numbers would have to be shown; for now it is a guide for where to look.</p><p><strong>Read and check</strong></p><ul><li>Spanning Trees in Four Dimensions (P15). The paper with all proofs, figures, two primers for physicists and number theorists, and a companion code that checks every step at 50 and 100 digits: <a href="https://doi.org/10.5281/zenodo.23927482">https://doi.org/10.5281/zenodo.23927482</a></li><li>One Character per Laplace Symbol (P9), the two-dimensional rung of the ladder: [doi:10.5281/zenodo.23805526](<a href="https://doi.org/10.5281/zenodo.23805526">https://doi.org/10.5281/zenodo.23805526</a>)</li><li>A. J. Guttmann and M. D. Rogers, Spanning tree generating functions and Mahler measures: [doi:10.1088/1751–8113/45/49/494001](<a href="https://doi.org/10.1088/1751-8113/45/49/494001">https://doi.org/10.1088/1751-8113/45/49/494001</a>)</li><li>M. A. Papanikolas, M. D. Rogers and D. Samart, The Mahler measure of a Calabi–Yau threefold and special L-values: [arXiv:1305.2143](<a href="https://arxiv.org/abs/1305.2143">https://arxiv.org/abs/1305.2143</a>)</li><li>J. L. Felker and R. Lyons, High-precision entropy values for spanning trees in lattices: [doi:10.1088/0305–4470/36/31/301](<a href="https://doi.org/10.1088/0305-4470/36/31/301">https://doi.org/10.1088/0305-4470/36/31/301</a>)</li><li>D. Broadhurst, Feynman integrals, L-series and Kloosterman moments: [doi:10.4310/CNTP.2016.v10.n3.a3](<a href="https://doi.org/10.4310/CNTP.2016.v10.n3.a3">https://doi.org/10.4310/CNTP.2016.v10.n3.a3</a>)</li></ul><p>Thomas Krause is an independent researcher in Lübeck, Germany. He writes the book series Geometry of Reality.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=09cfff9ccb27" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Why Some Levels of Methane Are Missing: The Molecule Follows a Pattern From Number Theory]]></title>
            <link>https://medium.com/@krause-virtuell/why-some-levels-of-methane-are-missing-the-molecule-follows-a-pattern-from-number-theory-ac1140901651?source=rss-ec2b54236e52------2</link>
            <guid isPermaLink="false">https://medium.com/p/ac1140901651</guid>
            <dc:creator><![CDATA[Thomas Krause]]></dc:creator>
            <pubDate>Tue, 03 Nov 2026 10:46:21 GMT</pubDate>
            <atom:updated>2026-11-04T17:26:49.422Z</atom:updated>
            <content:encoded><![CDATA[<p>A rotating molecule with identical nuclei hides the same periodic patterns that mathematicians use to count primes. Only three of them can occur, and they explain an old rule of chemistry.</p><p><strong>Key message: </strong>The quantum mechanical statistical weights of rotating molecules correspond precisely to number-theoretic Dirichlet characters, allowing the thermodynamics of symmetrical gases to be calculated seamlessly and exactly for all temperatures using mathematical modular forms.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*5MvyDk9nh_VoVP9EX0qOXA.png" /><figcaption>The number by which a rotation acts on each rotational level J. For 2-, 3- and 4-fold axes it is always +1, 0 or -1, and it repeats with period 2, 3 or 4 in J. These are three Dirichlet characters, the patterns of number theory.</figcaption></figure><p>A methane molecule is a carbon atom with four hydrogen atoms at the corners of a tetrahedron. It spins like a tiny top. Quantum mechanics says that its rotational energy comes in levels, labelled J = 0, 1, 2, and so on. But the four hydrogen nuclei are identical, and identical particles obey a strict rule: swapping two of them must leave the state unchanged or flip its sign. Because of this rule, some combinations of rotation and nuclear spin are forbidden. Some levels are missing, and the others come with odd statistical weights.</p><p>Chemists have tabulated these weights since 1935 ([Wilson 1935](<a href="https://doi.org/10.1063/1.1749653">https://doi.org/10.1063/1.1749653</a>)). For methane, the first five levels have weights 5, 3, 5, 11 and 13. At first sight this looks like a list of exceptions.</p><p>My new paper (<a href="https://doi.org/10.5281/zenodo.23865579">https://doi.org/10.5281/zenodo.23865579</a>) shows that it is not. Written in the right variable, these weights are built from a few periodic patterns of +1, 0 and -1 that number theorists call Dirichlet characters. Those are the same patterns used to count primes in arithmetic progressions.</p><p><strong>Turning a molecule</strong></p><p>Start with a single rotation. Turn the molecule by a fraction 1/m of a full turn around one of its symmetry axes. Each rotational level responds with one number, its character. With n = 2J + 1 this number is</p><p>sin(nπ/m) / sin(π/m)</p><p>For most m this is an awkward number. For m = 5, a five-fold axis, the values involve the golden ratio, 1.618… and 0.618…. For m = 6 the value 2 shows up. But for m = 2, 3 and 4, and only for these, the value is always +1, 0 or -1, and it repeats</p><ul><li>Two-fold axis: +1, -1, +1, -1, … (period 2 in J)</li><li>Three-fold axis: +1, 0, -1, +1, 0, -1, … (period 3)</li><li>Four-fold axis: +1, +1, -1, -1, … (period 4)</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*hlMZ753dYpooRkbLI9lQLQ.png" /><figcaption>The number by which a rotation of order m acts on level J. Blue: for two-, three- and four-fold axes only +1, 0 and -1 occur. Red: for five-fold axes golden numbers appear, for six-fold axes the value 2.</figcaption></figure><p>These three patterns are not new inventions. They are exactly the characters that belong to three famous number systems: the Gaussian integers, built on i = √-1, and the number fields of √-3 and √-2. In the notation of number theory they are χ-4, χ-3 and χ-8, read “chi minus four” and so on; the number is the discriminant of the field.</p><p>The proof is elementary and fits in a few lines. It is a cousin of the crystallographic restriction, the classical fact that a crystal can only have two-, three-, four- or six-fold symmetry. Here the stricter requirement, that the rotation acts by a character, removes the six-fold case as well.</p><p><strong>Methane, written in characters</strong><br>A molecule has many rotations at once. Methane has three two-fold axes and four three-fold axes. The standard recipe of group theory averages over all of them, and with the patterns above the weight of level J becomes</p><p>w = ( 4n + 8·χ-3(n) + 3·χ-4(n) ) / 3, with n = 2J + 1.</p><p>Put in n = 1, 3, 5, 7, 9 and you get 5, 3, 5, 11, 13, the numbers in the tables. The formula has three parts with three meanings:</p><ul><li>4n/3 is the classical rotor. The factor 4/3 is 16/12: sixteen nuclear spin states divided by methane’s symmetry number, 12.</li><li>8·χ-3 comes from the three-fold axes.</li><li>3·χ-4 comes from the two-fold axes.</li></ul><p>So every rotation order of the molecule leaves its own fingerprint, and the fingerprint is a number-theoretic pattern. The same works for sulphur hexafluoride, SF6, an octahedron, where the four-fold axes bring in χ-8.</p><p><strong>Heat as a modular form</strong><br>What happens when the molecule is warm? The partition function is the sum over all levels, weighted by the Boltzmann factor. It contains everything about the heat content of a gas of rotating molecules.</p><p>Once the weights are characters, the partition function changes character too. Each piece becomes a theta series, and for methane and SF6 these series are exact products of Dedekind’s eta function. Such objects are modular forms. They have a hidden symmetry: they turn into themselves, up to a simple factor, when the temperature is inverted.</p><p>That symmetry is useful. It connects the cold side, where only a few levels count, with the hot side, where very many levels contribute. One line of algebra then gives the full high-temperature behaviour. Each axis of order m leaves a correction of size</p><p>e^(-π² T / (m² Θ)),</p><p>where Θ is the rotational temperature of the molecule. The higher the order of the axis, the slower its fingerprint fades as the gas warms up. In methane the three-fold axes dominate, in SF6 the four-fold ones, in C60 the five-fold ones. These leading terms were found by direct calculation in 1987 (McDowell 1987). In the new picture they follow from the modular symmetry, and the formulas hold exactly at every temperature, not only in the limit.</p><p><strong>Why dividing by 12 works so well</strong><br>Every chemistry textbook has a simple rule. Take the classical partition function of a rotating molecule, divide it by the symmetry number (12 for methane), and multiply by the number of nuclear spin states. Above a few dozen kelvin this rule is astonishingly accurate. Why?</p><p>The answer turns out to be arithmetic. The correction to the rule is made of the character terms. Their behaviour at high temperature is controlled by a Dirichlet L-function, the generating series of the character. For the characters that occur here, all of the “odd” kind, these L-functions vanish at all negative odd integers. A classical result of number theory thus becomes a statement about heat: the correction has no power series at all, not even a tiny one. It is exponentially small.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*yVwLO8OiP56zAsRxfT8oDw.png" /><figcaption>Share of the two character terms in the partition function of freely rotating methane. Below about 20 kelvin they are large; at room temperature the three-fold share is about 3 × 10^-19.</figcaption></figure><p>At 10 kelvin the three-fold term carries 35 percent of the partition function of methane. At room temperature it is about 3 × 10^-19. This is why ordinary thermochemistry never notices it, and why the symmetry number rule works.</p><p><strong>A special temperature and Γ(1/4)</strong><br>Because of the modular symmetry, each term has one special temperature at which the cold and the hot description are equally good. For methane it lies at 3Θ/π = 7.20 kelvin. There the three-fold term takes an exact value:<br>Γ(1/4)³ · 2^(-9/4) · π^(-9/4) = 0.76249767…</p><p>Γ is Euler’s gamma function, and Γ(1/4) is the number that measures the arc length of the lemniscate, a figure-eight curve. I had expected Γ(1/3) here instead, because the three-fold axis belongs to the number field of √-3, and Γ(1/3) is that field’s natural constant. The expectation was wrong, and the reason is instructive. A real temperature puts the modular variable on the imaginary axis, and special points there belong to the field of i, not to the field of √-3. Where a character sits decides which constant appears.</p><p><strong>Ammonia: tunnelling as a dial</strong><br>Ammonia, NH3, is a flat pyramid that can tunnel through its own plane, like an umbrella turning inside out. This splits each level into two, by an energy Δ. Its symmetry has a three-fold axis, so one might expect χ-3. It is not there in the exact formula. Instead, tunnelling switches on the two-fold pattern χ-4, the same one that separates ortho and para hydrogen. Its strength is set by the factor 1 — e^(-Δ/kT). Without tunnelling the term vanishes; with strong tunnelling it is fully present. At 2 kelvin it is -38 percent of the partition function.</p><p><strong>C60 and the edges of a polyhedron</strong><br>The football-shaped molecule C60 has the full symmetry of the icosahedron. Its carbon-12 nuclei carry no spin at all, so only perfectly symmetric rotational states survive. Many levels are missing: J = 1 to 5, 7 to 9, 11, 13, 14, 17, 19, 23 and 29. From J = 30 on, every level is allowed.</p><p>Precision spectroscopy has confirmed this pattern directly (Changala et al. 2019; Chan et al. 2025).</p><p>The list has a mirror symmetry: below 30, level J is allowed exactly when 29 — J is missing. This was known for C60 (Petrov et al. 2026). The paper shows that it is part of a general rule. For the tetrahedron, the octahedron, the icosahedron and the prisms with an even number of sides, the allowed levels form what mathematicians call a symmetric numerical semigroup. The last missing level lies exactly one below the number of edges of the polyhedron: 5 for the tetrahedron with 6 edges, 11 for the octahedron with 12, 29 for the icosahedron with 30.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qZPHI2w51v2CCwKL02tORQ.png" /><figcaption>Allowed (blue) and missing (white) rotational levels for spinless nuclei on three polyhedral frames. Numbers in the boxes give the multiplicity when it is larger than 1. The red line marks the number of edges; the last missing level always lies just below it.</figcaption></figure><p>In words: the last forbidden rotational level of a polyhedral molecule with spinless nuclei lies one below the number of edges of its polyhedron. The proof uses only the rotation axes and a counting argument, no heavy machinery.</p><p><strong>What this is not</strong></p><ul><li>The molecules are treated as rigid rotors. Centrifugal stretching, small splittings and vibrations are ignored.</li><li>The exact formulas change no measured quantity. They are closed expressions for quantities that were already known, and the weights themselves go back to 1935.</li><li>Ammonia is the only non-rigid molecule in the paper. Benzene enters only partly, and one question about it is open.</li><li>For C60, the five-fold term could not be written as an eta product; this is open beyond the tested range.</li></ul><p><strong>What it might mean</strong><br>Where could one see the character terms directly? They are large only when the molecules are cold and still rotating freely. Methane trapped inside a C60 cage does exactly that (Bloodworth et al. 2019). If measured spin-isomer ratios of such trapped methane followed the closed formulas between 2 and 20 kelvin, one would see a Dirichlet character at work in a thermometer reading. For spectroscopy, the edge rule turns the list of missing lines of a polyhedral molecule into a simple counting question.</p><p>There is also a thought behind this work, and it is an interpretation, not a result. The rotation group of a molecule seems to select a number field, and the nuclear spin statistics would then be that field’s pattern made audible through heat. On this reading, methane carries the field of the triangle, ammonia borrows the field of the square through tunnelling, and C60 reaches the first pattern beyond them. If that view were right, the arithmetic would have been there before any physics was computed.</p><p>Whether it holds would have to be shown case by case; this paper is one such case, and the proofs do not depend on it.</p><p><strong>Read and check</strong></p><ul><li>The Symmetry of a Molecule Picks a Character: Nuclear spin statistics, eta quotients and Dirichlet L-functions. The paper with all proofs and figures, plus a companion code that checks every step at two precisions: doi: <a href="https://doi.org/10.5281/zenodo.23865580">https://doi.org/10.5281/zenodo.23865580</a></li><li>The Classical Side (P7), where the same two-fold pattern appears in ortho and para hydrogen: doi:10.5281/zenodo.23706677</li><li>One Character per Laplace Symbol (P9), where χ-3 lives in the triangular lattice instead: doi:10.5281/zenodo.23805526</li><li>E. B. Wilson, The statistical weights of the rotational levels of polyatomic molecules: doi:10.1063/1.1749653</li><li>R. S. McDowell, Rotational partition functions for spherical-top molecules: doi:10.1016/0022–4073(87)90028–8</li><li>P. B. Changala et al., Rovibrational quantum state resolution of the C60 fullerene: doi:10.1126/science.aav2616</li></ul><p>Thomas Krause is an independent researcher in Lübeck, Germany. He writes the book series Geometry of Reality.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ac1140901651" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How Often Can Heat Cancel the Vacuum? Ask a Continued Fraction]]></title>
            <link>https://medium.com/@krause-virtuell/how-often-can-heat-cancel-the-vacuum-ask-a-continued-fraction-6ee6803c90a2?source=rss-ec2b54236e52------2</link>
            <guid isPermaLink="false">https://medium.com/p/6ee6803c90a2</guid>
            <category><![CDATA[quantum-mechanics]]></category>
            <category><![CDATA[geometry]]></category>
            <category><![CDATA[physics]]></category>
            <category><![CDATA[quantum-physics]]></category>
            <category><![CDATA[mathematics]]></category>
            <dc:creator><![CDATA[Thomas Krause]]></dc:creator>
            <pubDate>Sun, 01 Nov 2026 18:48:54 GMT</pubDate>
            <atom:updated>2026-11-04T17:31:37.285Z</atom:updated>
            <content:encoded><![CDATA[<p><strong>On a twisted ring of space, the answer does not depend on size, only on how well one number can be written as a fraction.</strong></p><p><strong>Key message: </strong>The <strong>number of thermodynamic equilibrium temperatures</strong> of a quantum field on a twisted torus is determined exactly by the <strong>number theory of its continued fraction</strong>, where each exceptionally good rational approximation of the twist generates two new physical equilibria.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*5-URrzQ1OiImDTvlYaq5rg.png" /><figcaption>Ford circles, one for each fraction p/q. The vertical line is the twist π − 3. Wherever it cuts a highlighted circle, heat and vacuum get another chance to balance</figcaption></figure><p>Empty space is not quite empty. Between two metal plates, the quantum vacuum pushes them together. This Casimir effect has been measured many times, and its energy is negative. Heat pushes in the other direction: a warm gas of light between the same plates carries positive energy.</p><p>So there must be a temperature where the two cancel exactly. My new paper (<a href="https://doi.org/10.5281/zenodo.23858840">https://doi.org/10.5281/zenodo.23858840</a>) asks a simple question about that point: how many such balance points are there? The answer turned out to be a question about numbers, not about physics. For a twisted ring it can be one, nine, or infinitely many, and which of these you get is read off from a continued fraction.</p><p><strong>Time as a circle</strong></p><p>Physicists have a neat trick for describing anything at temperature T. They treat time as if it were a fourth direction of space and roll it up into a small circle. The circumference of that circle is β = ħc/kT. Hot means a small circle, cold means a large one.</p><p>Now close one direction of real space into a ring of length ℓ as well. Together, the time circle and the space ring form the surface of a doughnut, a torus. Its shape is captured by one number, the ratio y = β/ℓ.</p><p>On this torus, a massless field, the simplest kind of light, carries two competing energies. The Casimir energy of the ring is negative. The thermal energy of the time circle is positive. A balance point, or node, is where the total is exactly zero.</p><p><strong>Size drops out</strong></p><p>The first result is that the size of the system does not matter, only its shape. Make the ring ten times larger and the balance temperature drops by ten. The product of temperature and ring length stays fixed.</p><p>Without any twist, there is exactly one balance point, in every number of dimensions from three upward. This is proved. In our four-dimensional spacetime it sits at</p><p>kTℓ/ħc = 0.6578618…</p><p>For a ring of one micrometre this is about 1500 kelvin. Below it the vacuum wins, above it heat wins.</p><p><strong>The twist</strong></p><p>Now comes the interesting part. You can glue the doughnut with a twist. After one lap around the time circle, the field comes back shifted along the ring by a fraction θ of its length.</p><p>This is not a real rotation of the ring. It is how a heat bath moving along the ring shows up in the mathematics, with a formally imaginary velocity. Lattice simulations of particle physics use exactly this kind of boundary condition to measure thermodynamic quantities ([Giusti and Meyer 2011](<a href="https://doi.org/10.1007/JHEP11(2011)087)).">https://doi.org/10.1007/JHEP11(2011)087)).</a></p><p>What does the twist do? Think of a dial. Each lap in time moves the pointer forward by θ of a full turn. If θ were exactly 1/7, the pointer would be back at the start after seven laps. The path closes into a short, slightly tilted loop, and that loop acts like a new, shorter ring. A shorter ring has a stronger Casimir energy. In a window of temperatures, the vacuum wins again, and the energy crosses zero twice more: two new balance points.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*567ON7IVpG5xSDYtvq48rw.png" /><figcaption>Where the pointer lands after each lap in time. Left: with twist π − 3 the pointer is almost back after 7 laps. Right: with the golden ratio it never comes back well enough</figcaption></figure><p><strong>The test: laps times miss</strong></p><p>Of course θ is rarely a simple fraction. So the question becomes: how close does the pointer get to the start, and how many laps does it take?</p><p>The paper gives a sharp test. Take the number of laps q and the miss, how far from the start the pointer lands, measured in turns. If</p><p>laps × miss ≤ about 0.18,</p><p>the near return is good enough, and a new pair of balance points appears. In four dimensions the exact threshold is 0.1813.</p><p><strong>Two examples make this concrete.</strong></p><ul><li>For θ = π − 3 = 0.14159…, lap 7 misses the start by only 0.0089 of a turn. Laps × miss = 7 × 0.0089 = 0.062. Well below the threshold.</li><li>For the golden ratio, the pointer comes closest after 5, 8 and 13 laps. Laps × miss is 0.451, 0.446 and 0.448. It never drops; it settles at 1/√5 ≈ 0.447. Always above the threshold.</li></ul><p><strong>Continued fractions keep the score</strong></p><p>Mathematicians have a perfect bookkeeping tool for this, the continued fraction. Every number can be written as a tower of fractions,</p><p>π − 3 = 1/(7 + 1/(15 + 1/(1 + 1/(292 + …)))),</p><p>written [0; 7, 15, 1, 292, …]. Cutting the tower off gives the best fractions: 1/7, 15/106, 16/113, and so on, which is why 22/7 and 355/113 are famous approximations of π.</p><p>A large entry in the tower means an unusually good approximation comes next. Roughly: an entry of 6 or more gives a new pair of balance points, and an entry of 3 or less never does.</p><p>The golden ratio has the tower [0; 1, 1, 1, …], all ones. It is the number that is hardest to approximate by fractions, often called the most irrational number (a classical result going back to Hurwitz). The same golden number shows up elsewhere in physics for the same reason. In the theory of chaos, the orbit with golden winding number is the last one to break up [Greene 1979] (<a href="https://doi.org/10.1063/1.524170">https://doi.org/10.1063/1.524170</a>). And in Hofstadter’s butterfly, the energy spectrum of electrons in a magnetic field depends on the continued fraction of the magnetic flux [Hofstadter 1976](<a href="https://doi.org/10.1103/PhysRevB.14.2239">https://doi.org/10.1103/PhysRevB.14.223</a>).</p><p><strong>One, nine or infinitely many</strong></p><p>Putting it together gives three very different answers.</p><ul><li>Golden ratio: exactly one balance point. This is proved with computer-assisted interval arithmetic for three to six dimensions. The same holds for √2 − 1, √3 − 1, and every twist whose tower contains only entries of 3 or less.</li><li>π − 3: at least nine balance points, proved. Five of them lie in the range shown below, at kTℓ/ħc ≈ 0.658, 13.2, 29.1, 1377 and 8249. The others lie at even higher temperatures.</li><li>e − 2: the tower [0; 1, 2, 1, 1, 4, 1, 1, 6, …] contains ever larger entries, so there are infinitely many balance points, proved.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*RGgPR0t3STVCdK8mGFsxQQ.png" /><figcaption>Which side wins as the system gets hotter, for three twists. Blue: the vacuum wins. Red: heat wins. Gold dots: balance points. The vertical scale is compressed</figcaption></figure><p>There is a geometric way to see these three answers. Heating the torus traces a path on the surface of the modular group, and each entry of the continued fraction records how far this path dips into the surface’s one infinitely long “throat” (Series 1985). For the golden ratio the path settles into the shortest closed loop on that surface, for e − 2 it dips ever deeper by a fixed rule, and for π − 3 no rule is known.</p><p>What about a twist picked at random? Here classical number theory answers. Almost every number has arbitrarily large entries in its tower (Borel and Bernstein), and about 22 percent of all entries are 6 or more (Gauss and Kuzmin). So almost every twist gives infinitely many balance points. The golden ratio, with its single balance point, is the rare exception.</p><p>Two special shapes are never in balance at all: the square torus and the hexagonal torus.</p><p><strong>What this is not</strong></p><ul><li>The paper treats a single free massless boson field. A mass would bring a length scale back in. Fermions, such as electrons, are not covered yet.</li><li>The upper bound on the number of balance points is proved only for three to six dimensions. A narrow band of thresholds, less than 0.01 wide, is still open.</li><li>Whether the balance points are stable equilibria is not addressed.</li><li>The twist is a boundary condition, not a spinning object.</li></ul><p><strong>What it might mean</strong></p><p>If this picture carried over to experiment, the cleanest test would perhaps come from lattice simulations rather than from the laboratory. With a twist near 1/7 and a sufficiently hot, short torus, such simulations ought to show the first extra window in which the vacuum wins again. If they did, the number of equilibria of a physical system would depend on the arithmetic of one of its parameters. That would be an unusual kind of count in classical physics: not set by forces or masses, but by how well a number can be written as a fraction. Higher-dimensional tori would then lead to simultaneous approximation of several numbers at once, a much harder problem of number theory, and a version for fermions could show whether the same rule survives once spin enters.</p><p>There is also a thought behind this work, and it is an interpretation, not a result. If geometry alone decided whether heat and vacuum balance once or infinitely often, then the arithmetic of a number would not merely describe the physics. It would be part of it. One might then wonder whether other physical counts, too, are set less by what things are made of than by the shape of the space they live in. Whether that view holds would have to be shown case by case, and this paper would be only one small case.</p><p><strong>Read and check</strong></p><ul><li>Where Heat Balances the Vacuum: How continued fractions count the equilibria of a twisted thermal torus. The paper with all proofs and figures, plus a companion code that checks every step in ball arithmetic: <a href="https://doi.org/10.5281/zenodo.23858840">https://doi.org/10.5281/zenodo.23858840</a></li><li>Euler Products, the Reflection s ↔ 1−s, and Zeta Values at Integers (P0, the toolkit of the series): <a href="https://doi.org/10.5281/zenodo.23301196">https://doi.org/10.5281/zenodo.23301196</a></li><li>The Casimir Constant as a Bernoulli Number (P1): <a href="https://doi.org/10.5281/zenodo.23301747">https://doi.org/10.5281/zenodo.23301747</a></li><li>L. Giusti, H. B. Meyer, Thermal momentum distribution from path integrals with shifted boundary conditions: <a href="https://doi.org/10.1007/JHEP11(2011)087)">https://doi.org/10.1007/JHEP11(2011)087</a></li><li>F. Alessio, G. Barnich, Modular invariance in finite temperature Casimir effect: <a href="https://doi.org/10.1007/JHEP10(2020)134)">https://doi.org/10.1007/JHEP10(2020)134</a></li><li>L. S. Brown, G. J. Maclay, Vacuum stress between conducting plates: <a href="https://doi.org/10.1103/PhysRev.184.1272">https://doi.org/10.1103/PhysRev.184.1272</a></li></ul><p>Thomas Krause is an independent researcher in Lübeck, Germany. He writes the book series Geometry of Reality.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=6ee6803c90a2" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[One Formula for Superconductors, Built from Riemann’s Zeta Function]]></title>
            <link>https://medium.com/@krause-virtuell/one-formula-for-superconductors-built-from-riemanns-zeta-function-78e1678208fc?source=rss-ec2b54236e52------2</link>
            <guid isPermaLink="false">https://medium.com/p/78e1678208fc</guid>
            <category><![CDATA[physics]]></category>
            <category><![CDATA[mathematics]]></category>
            <category><![CDATA[number-theory]]></category>
            <category><![CDATA[geometry]]></category>
            <category><![CDATA[superconductor]]></category>
            <dc:creator><![CDATA[Thomas Krause]]></dc:creator>
            <pubDate>Sat, 31 Oct 2026 00:34:27 GMT</pubDate>
            <atom:updated>2026-11-08T17:07:52.888Z</atom:updated>
            <content:encoded><![CDATA[<p><strong>From ordinary metals to graphene, the strength of electron pairing follows a single curve — and it contains ζ(½)</strong></p><p><strong>Key message: </strong>The two poles of the completed Riemann zeta function strictly govern the quantum instabilities of Bose-Einstein condensation s=1 and Cooper pairing s=0, yielding a single closed formula that exactly determines the measurable gap-to-T_{c} ratio of superconductors across any power-law density of states.</p><p>In 1957 John Bardeen, Leon Cooper and Robert Schrieffer explained superconductivity. Their theory contains a number that every student of the subject learns by heart: 1.764.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Lu87p6vo2S-IFt8095BhKQ.png" /></figure><p>It is the ratio between two quantities you can measure in any superconductor. One is the energy gap Δ₀, the minimum energy needed to break a pair of electrons at zero temperature. The other is the critical temperature T_c, where superconductivity disappears. For an ordinary metal at weak attraction, the theory predicts</p><p>Δ₀ / T_c = π e^(−γ) ≈ 1.764,</p><p>where γ ≈ 0.5772 is Euler’s constant (Bardeen, Cooper, Schrieffer 1957). The remarkable thing is that the ratio does not depend on the metal. The strength of the attraction and the details of the material cancel. Measuring a value near 1.76 is still a standard sign that a superconductor is “conventional”.</p><p>Fifty years later a second number of this kind appeared. For superconducting graphene at the point where its two energy cones touch, Nikolai Kopnin and Edouard Sonin found</p><p>Δ₀ / T_c = 2 ln 2 ≈ 1.386 (Kopnin and Sonin 2008).</p><p>The two values look unrelated. One has π, an exponential and Euler’s constant. The other has a logarithm of 2. My new paper shows that they are two points on one curve, and that the curve is written in terms of one of the most famous functions in mathematics.</p><p><strong>What changes from one material to the next</strong><br>To see the curve you need one idea: the density of states. It counts how many electron states are available near a given energy ξ.</p><p>In an ordinary metal the electrons that matter sit near the Fermi energy, and there the density of states is practically constant. In graphene the electrons behave like massless particles. Their energy grows linearly with momentum, and the number of states grows linearly with energy, N(ξ) ∝ ξ.<br>Newer materials give other powers. In general one writes</p><p>N(ξ) ∝ ξ^a.</p><p>The exponent a is the only property of the material that enters the result. There is a simple rule for it. If the energy grows with momentum like p in some directions and like p² or p³ in others, then</p><p>a = (sum of 1/n over all directions) − 1,</p><p>where n is the power in each direction. Graphene has two linear directions: a = 1 + 1 − 1 = 1. A semi-Dirac material, linear in one direction and quadratic in the other, has a = 1 + ½ − 1 = ½. Such points were predicted in oxide multilayers by Victor Pardo and Warren Pickett (Pardo and Pickett 2009; Banerjee et al. 2009), and an anisotropic Dirac state with linear and quadratic directions was observed in potassium-doped black phosphorus (Kim et al. 2015). A triple-Weyl node in three dimensions, cubic in two directions and linear in the third, has a = ⅓ + ⅓ + 1 − 1 = ⅔ (Fang et al. 2012).</p><p>There is one important difference between a = 0 and a &gt; 0. In an ordinary metal, any attraction, however weak, produces a superconductor. When a is positive, there are too few states near the crossing point, and pairing needs a minimum strength of attraction. In the mean-field theory used here, that threshold is exactly the coupling λ_c = a. The ratio Δ₀/T_c becomes universal again in the limit where the attraction just barely exceeds this threshold, and this is the number in the table.</p><p><strong>One formula</strong></p><p>The main result of the paper is a closed formula for that number, for every a between 0 and 2:</p><p>(Δ₀ / T_c)^a = 2^(a+1) · η(a) · Γ(1 + a/2) / Γ(1 − a/2).</p><p>Γ is the gamma function, the continuous version of the factorial. The new ingredient is η, the alternating zeta function:</p><p>η(a) = 1 − 1/2^a + 1/3^a − 1/4^a + …</p><p>It is a close relative of Riemann’s zeta function ζ. Both are built from the same reciprocal powers of the whole numbers; η only flips every second sign. The exact relation is</p><p>η(a) = (1 − 2^(1−a)) · ζ(a).</p><p>The alternating signs are not decoration. Electrons are fermions, and in thermal physics fermions carry a minus sign for each turn around the imaginary time circle. That minus sign is what turns ζ into η.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*xsBD-NppvIWfj9AshQTZrA.png" /><figcaption>The gap-to-critical-temperature-ratio at the threshold of pairing, plotted against the exponent a of the density of states. The shaded band is the range in which the formula uses values of η inside the critical strip. The open triangle is the earlier approx. for semi-Dirac materials. At a = 2 the curve ends at a pole of the gamma function.</figcaption></figure><p><strong>Old numbers, one new one</strong><br>The formula reproduces both known values and adds new ones.</p><p><strong>a → 0, the ordinary metal</strong>. Both sides of the formula tend to 1, and taking the limit carefully gives back π e^(−γ) = 1.764, the number from 1957. Its universality at weak coupling was proved rigorously by Christian Hainzl and Robert Seiringer (Hainzl and Seiringer 2008), and extended to low dimensions by Henheik, Lauritsen and Roos (Henheik, Lauritsen, Roos 2024).</p><p><strong>a = 1, graphene</strong>. Here η(1) = 1 − ½ + ⅓ − ¼ + … = ln 2, the gamma factors give 1, and the formula reduces to 2 ln 2 = 1.386. This is the Kopnin–Sonin value.</p><p><strong>a = ½, semi-Dirac</strong>. This case is new. The formula gives an exact constant,<br>Δ₀ / T_c = π² η(½)² / Γ(¾)⁴ = 1.6015…</p><p>Bruno Uchoa and Kangjun Seo had worked out the zero-temperature gap for this case and used an approximation for T_c, which gives 1.393 (Uchoa and Seo 2017). The exact value is 13 percent higher. Their gap formula is confirmed; only the shortcut for the temperature was too rough.</p><p><strong>a = 2, the end</strong>. Γ(1 − a/2) has a pole at a = 2. The denominator becomes infinite and the ratio drops to zero. The curve does not just fade out; it ends at a pole of a classical function.</p><p><strong>Where Riemann enters</strong><br>So far η could be just a convenient name for a series. The paper shows that it is more than that, and the reason is a pair of poles.</p><p>In 1859 Bernhard Riemann showed that his zeta function becomes perfectly symmetric if you complete it with a gamma factor:</p><p>Λ(s) = π^(−s/2) · Γ(s/2) · ζ(s), with Λ(s) = Λ(1 − s).</p><p>This completed function has exactly two poles, at s = 0 and at s = 1, and nowhere else.</p><p>A cold gas of quantum particles can become unstable in two classical ways. Bosons can collapse into a single quantum state, Bose–Einstein condensation. Fermions can form pairs and become a superconductor. For a density of states ξ^a, each of these instabilities is controlled by one factor that comes from a standard integral:</p><ul><li>Bose condensation: Γ(a + 1) · ζ(a + 1), which reads Riemann’s function at s = a + 1.</li><li>Cooper pairing: 2 Γ(a) · η(a), which reads it at s = a.</li></ul><p>At a = 0, both factors hit a pole at the same time. Condensation sits on the pole of ζ at s = 1. Pairing sits on the pole of Γ at s = 0. Both blow up like 1/a, with the same strength. That is the mathematical reason why an ordinary metal becomes superconducting for any attraction at all: the pairing factor is infinite there.</p><p>The constants left over after removing the 1/a are also telling. For condensation the constant is exactly 0. For pairing it is</p><p>ln(π/2) − γ = −0.1256…</p><p>and this is minus the logarithm of 2e^γ/π = 1.1339, the prefactor that appears in the BCS formula for the critical temperature. The famous BCS constants are what remains of Riemann’s function right next to its pole.<br>A word of caution about naming. The two factors do not sit at mirror images s and 1 − s of each other. They read the same function at s and at s + 1, shifted by one. What they share is the boundary a = 0, where each meets one of the two poles of Λ.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*EoxPEUFWhwsQS9nMsAPpqg.png" /><figcaption>The two poles of Riemann’s completed zeta function, at s = 0 and s = 1. Pairing of electrons reads the function at s = a, condensation of bosons at s = a + 1. For ordinary metals, a = 0, both sit on a pole.</figcaption></figure><p><strong>A measurable number from the critical strip</strong><br>There is one more detail that I find striking. For 0 &lt; a &lt; 1 the formula uses values of η between 0 and 1. In number theory this region has a name: the critical strip. It is where all the nontrivial zeros of the zeta function lie, and where the Riemann hypothesis lives. For a = ½ the formula uses η(½) = (1 − √2) · ζ(½), a value right on the critical line.</p><p>To be clear: nothing in this work says anything about the zeros, and nothing in it bears on the Riemann hypothesis. The formula uses real values of η, not complex ones. But to my knowledge this is the first time that values from inside the critical strip appear in a measurable quantity of this kind. A ratio you could measure in a laboratory contains ζ(½).</p><p><strong>What it could be used for</strong><br>The formula is not only a curiosity. It turns two routine measurements into a statement about a material, and it gives experiments a fixed target.</p><p><strong>A fingerprint of the electron structure</strong>. When a new superconductor is discovered, one of the first questions is whether it is conventional or whether its electrons have an unusual structure, such as Dirac or Weyl points. Two standard measurements answer this. The pair energy Δ₀ comes from tunnelling spectroscopy or from photoemission (ARPES). The critical temperature T_c comes from the electrical resistance. Their ratio, read against the curve, points to the exponent a:</p><ul><li>about 1.76: an ordinary metal,</li><li>about 1.60: a semi-Dirac material,</li><li>about 1.54: a triple-Weyl semimetal,</li><li>about 1.39: Dirac or double-Weyl nodes.</li></ul><p>Mapping the band structure directly is demanding. With the formula, a measurement that is done for almost every new superconductor becomes a cross-check of the band structure. Reading a off uniquely assumes that the curve falls steadily, which so far is checked numerically.</p><p><strong>A target for tunable experiments</strong>. In many materials the attraction between electrons can be tuned, through pressure, strain, an electric gate or doping. For a nodal material the theory then predicts two things close to the onset of superconductivity. The critical temperature should rise like (λ − λ_c)^(1/a) above the threshold, and the ratio Δ₀/T_c should settle at the value of the curve. Candidate platforms are potassium-doped black phosphorus, oxide multilayers and multi-Weyl semimetals (Roy, Goswami, Juričić 2017). If an experiment lands on the curve, it confirms both the band structure and the mean-field picture. If it misses, that is informative too: it would show that fluctuations or effects beyond mean field are at work.</p><p><strong>An outlook</strong>. Further uses are conceivable, though nothing in the paper establishes them. Numerical codes for superconductors could use the family of exact values as reference points to test their accuracy. And the ratio Δ₀/T_c might matter in devices. In superconducting detectors and qubits, the number of electron pairs broken by heat falls exponentially with the pair energy relative to the temperature (Zmuidzinas 2012; Glazman and Catelani 2021). A material with a lower ratio would therefore behave differently from a conventional one at the same critical temperature. Whether that would be a drawback, or whether it could possibly be turned into an advantage somewhere, is an open question.</p><p><strong>What this is not</strong><br>The limits matter as much as the result. The formula is a mean-field result. In two dimensions, fluctuations and the Berezinskii–Kosterlitz–Thouless transition shift the measured T_c.</p><ul><li>The ratio is universal only in the limit where the coupling just exceeds the threshold. For a close to 2 this limit is approached slowly.</li><li>Disorder and doping move the chemical potential away from the node and smear out the power law.</li><li>That the curve falls steadily from 1.764 to 0 is checked numerically, not proved.</li><li>No new fundamental physics is claimed. The integrals behind the formula are classical. What is new is the general formula, its form through η, and the exact constant at a = ½.</li></ul><p><strong>Try it yourself</strong></p><p>To make the curve easy to explore, I built a small companion app, the Gap Ratio Explorer. It runs in the browser and needs no account.</p><p><a href="https://smex-ctp.trendmicro.com/wis/clicktime/v1/query?url=https%3a%2f%2fp12%2dgap%2dexplorer.pplx.app&amp;umid=671a0c38-a2d0-4d2c-92be-921fc78a4c2d&amp;auth=7c6a26a4c55d7c72a255a55e8b4b30c7438d56f1-677a76d7bf69bb0bbbece0c11ec2dbe566d11f08"><strong>p12-gap-explorer.pplx.app</strong></a></p><blockquote>Note when yo open the App: The app runs entirely in your browser and stores no data. Some security tools flag new web addresses as untested; this is not a malware warning.</blockquote><p>On the first tab you choose a material, or enter your own measured gap and critical temperature, and see where it sits relative to the curve. A slider moves the coupling and shows how the ratio follows. The second tab works the other way round: you click a point on the curve, fix one energy scale, and the app solves the gap equations for the rest. The third tab lists the formulas, the numerical method and all data sources.</p><p>A word of caution: the app uses the same simple model as the paper, one band with a sharp energy cutoff. Strongly coupled metals such as lead and mercury lie above the range it can reach, and two-band materials such as MgB₂ do not fit a single curve. The app shows these cases openly rather than hiding them.</p><p>— -</p><p><strong>Read and check</strong></p><ul><li>Space and Matter: The Two Sides of One Equation. A closed gap-to-T_c ratio for power-law densities of states and the two poles of the completed zeta function, with LaTeX source and a companion code that checks every number: <a href="https://www.preprints.org/manuscript/202611.0643">https://www.preprints.org/manuscript/202611.0643</a></li><li>The Superconducting Side: Odd Zeta Values and the Pole at s = 1 in BCS Theory: doi:10.5281/zenodo.23770403</li><li>J. Bardeen, L. N. Cooper, J. R. Schrieffer, Theory of Superconductivity: doi:10.1103/PhysRev.108.1175</li><li>N. B. Kopnin, E. B. Sonin, BCS superconductivity of Dirac electrons in graphene layers: doi:10.1103/PhysRevLett.100.246808</li><li>B. Uchoa, K. Seo, Superconducting states for semi-Dirac fermions: doi:10.1103/PhysRevB.96.220503</li></ul><p>Thomas Krause is an independent researcher in Lübeck, Germany. He writes the book series Geometry of Reality.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=78e1678208fc" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Big Bang Left a Second Glow — and It’s ∛(4/11) Colder]]></title>
            <link>https://medium.com/@krause-virtuell/the-big-bang-left-a-second-glow-and-its-4-11-colder-f1d1219cec68?source=rss-ec2b54236e52------2</link>
            <guid isPermaLink="false">https://medium.com/p/f1d1219cec68</guid>
            <category><![CDATA[mathematics]]></category>
            <category><![CDATA[big-bang]]></category>
            <category><![CDATA[physics]]></category>
            <category><![CDATA[number-theory]]></category>
            <category><![CDATA[geometry]]></category>
            <dc:creator><![CDATA[Thomas Krause]]></dc:creator>
            <pubDate>Fri, 30 Oct 2026 13:15:52 GMT</pubDate>
            <atom:updated>2026-11-04T17:39:18.225Z</atom:updated>
            <content:encoded><![CDATA[<p><strong>Why that number is almost, but not quite, exact — and why physicists roll time into a circle to find out</strong></p><p><strong>Key message: </strong>The cosmological temperature ratio and the quantum electrodynamic corrections of relic neutrino decoupling are fundamentally governed by number-theoretic Euler factors of the prime 2, causing the leading correction to the effective number of neutrino species N_{eff} to possess an exactly rational coefficient in front of alpha/pi due to the properties of weight-2 zeta moments.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*XQ7UKnBdvpWmlP1au7DtXA.png" /></figure><p>Everyone has heard of the afterglow of the Big Bang. The cosmic microwave background fills the whole sky, and its temperature today is 2.7255 kelvin, measured to four decimal places.</p><p>Fewer people know that there is a second afterglow. It is made not of light but of neutrinos, the ghostly particles that pass through the Earth as if it were empty space. About 336 of them per cubic centimetre, left over from the first second, are flying through you as you read this. Their temperature is predicted to be 1.945 kelvin.</p><p>The ratio between the two temperatures is a number you can write on a napkin:</p><p>T(neutrinos) / T(light) = ∛(4/11) ≈ 0.714.</p><p>This article is about where that number comes from, why it is not quite exact, and the strange trick physicists use to compute the difference. The trick turns time into a circle. Six steps later, it runs into a wall.</p><p><strong>A room that is heated after a guest has left</strong><br>In its first second the universe was a hot soup of light (photons), electrons, their antiparticles the positrons, and neutrinos. Everything was in contact with everything else, and everything had the same temperature.</p><p>As the universe expanded and cooled, the neutrinos stopped talking to the rest. They interact only through the weak force, and at a temperature of about ten billion degrees those interactions became too rare. From that moment on the neutrinos simply flew on, cooling with the expansion like everything else, but no longer exchanging energy (Particle Data Group).</p><p>Shortly afterwards the temperature dropped below the mass of the electron. Electrons and positrons met and annihilated into light. Their energy went into the photons, and only into the photons, because the neutrinos had already left the room.</p><p>So the light got a boost the neutrinos never received. How big a boost? You only need to count.</p><p><strong>Counting is enough</strong><br>The heat content of a hot gas, more precisely its entropy, is proportional to T³ times the number of ways its particles can exist. Photons come in two polarisations: 2. Electrons and positrons have two spin states each: 4. But electrons are fermions, and fermions in a hot gas count a little less, by a factor of exactly 7/8. So electrons and positrons count 4 × 7/8 = 3.5.</p><p>Before annihilation the photon side holds 2 + 3.5 = 5.5. Afterwards it holds 2. The entropy is conserved, so the photons’ T³ must grow by the factor 5.5/2 = 11/4 compared with the neutrinos. Take the cube root, and you have ∛(4/11).</p><p>That is the whole calculation, and it is in every cosmology textbook. It is also not quite right.</p><p><strong>Why “exactly” is not exact</strong><br>The napkin calculation makes two idealisations. The neutrinos did not leave the room at one sharp instant, and the electrons, positrons and photons are not a gas of free particles. They interact. Electrons push each other away, photons scatter, the plasma screens electric charges.<br>Physicists put the result into one number, N_eff, the “effective number of neutrino species”. In the idealised picture N_eff = 3, one for each kind of neutrino. The full Standard Model calculation gives N_eff = 3.044 (Bennett et al. 2021). The measurement from the Planck satellite is 2.99 ± 0.17 (Planck 2018), so the extra 0.044 is still below what we can see, but future surveys are closing in.</p><p>Most of the 0.044 comes from the messy part, the neutrinos leaving gradually, which can only be computed numerically. But about a fifth of it, almost exactly 0.01, comes from the interactions in the plasma, and that part has a closed formula. To get it, physicists do something that sounds like science fiction.</p><p><strong>The trick: time becomes a circle</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*y6cMAOLgGEcRx5Izk6mJsg.png" /><figcaption>Left: space and time as a flat strip. Middle: glue the top edge to the bottom edge, and time runs around a circle. Right: the hotter the gas, the thinner the cylinder.</figcaption></figure><p>Quantum theory has a remarkable identity. If you want the average behaviour of a system at temperature T, you can replace the averaging by something that looks completely different. You let the system evolve for a fixed stretch of “imaginary time”, a mathematical continuation of ordinary time, and you demand that it ends exactly where it started. The length of that stretch is set by the temperature: the hotter, the shorter.</p><p>“Ending where it started” means that the beginning and the end of the time interval are glued together. Time becomes a circle. Space stays as it is. The hot plasma of the early universe, in this language, lives on a cylinder: space along its length, time around it, with a circumference of one over the temperature (Laine and Vuorinen).</p><p>There is a small twist. Light, being made of bosons, comes back to itself after one turn. Electrons, being fermions, come back with a minus sign. That minus sign is the whole difference between bosons and fermions in a hot gas, and we will meet it again at the end.</p><p>Now every calculation becomes a drawing on this cylinder.</p><p><strong>Six steps on the cylinder</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*z_Lt4swFsyhk7M_Qzjwimg.png" /><figcaption>Each panel is one step of the calculation. The step number is the power of the electron charge e. Blue: light. Orange: electrons. Red: a loop frozen in time.</figcaption></figure><p>Physicists compute interactions as a series: a first answer, then a correction, then a correction to the correction. Each step carries one more power of the electron charge e, and e is small, so the steps get smaller. The picture shows what each step looks like on the cylinder.</p><p><strong>Step 0, the free gas. </strong>Particles that do not interact simply run around the time circle, once, twice, many times. Every winding contributes, with a weight that falls off with the number of turns. Adding up all windings gives the famous Stefan–Boltzmann law. For electrons the signs alternate, plus, minus, plus, because of the minus sign at each turn, and that alternation is exactly where the factor 7/8 comes from.</p><p><strong>There is no step 1. </strong>Every time an electron emits or absorbs a photon, a factor e appears, and photons always come with two ends. So the ordinary series only has even powers: e², e⁴, e⁶. Remember this, because it is about to break.</p><p><strong>Step 2, one photon exchanged</strong>. An electron loop runs around the circle, and a photon jumps across it. The result is a product of two windings, and the coefficient in front is a simple fraction. This term alone shifts N_eff by about 0.01.</p><p><strong>Step 3, screening</strong>. Here something unexpected happens: an odd power, e³. It cannot come from the ordinary series. It comes from a loop that does not wind around the time circle at all. It is frozen in time and sees only space. In a plasma, electric fields are screened: a charge is surrounded by a cloud of opposite charges, much as a celebrity is hidden by a crowd. The frozen loop is exactly this effect, and it brings a square root into the formula. That is why the series suddenly has e³.</p><p><strong>Step 4, three loops. </strong>Two photons are exchanged. Now logarithms appear, slower and subtler numbers than fractions. Still, the coefficient is known exactly, in closed form (Arnold and Zhai 1995).</p><p><strong>Step 5, deeper screening.</strong> The frozen loop gets electrons inside it. A famous constant appears: Euler’s γ ≈ 0.5772, the one every mathematics student meets in the harmonic series. Still known exactly (Andersen 1996).<br>Step 6, the wall. This is where the story splits in two.</p><p><strong>A wall that is not about computing power</strong><br>For light and electrons, step 6 is hard, but possible. The frozen loop has an electric part, which is screened, and a magnetic part, which is not. Magnetic photons are not screened, but photons do not interact with each other, so the unscreened part stays free and harmless. The step 6 coefficient has been partly computed only this year (Gorda and Majumder 2026). The rest is a matter of effort.</p><p>For quarks and gluons, the particles of the strong force, the same step is a wall of principle. Gluons, unlike photons, carry the charge of their own force, so they interact with each other. In the frozen, magnetic part of the loop they form a tangled, strongly coupled mess. Every extra loop costs a factor that is no longer small; it is of order one. The series, which relied on each step being smaller than the previous one, has nothing left to expand in.</p><p>This was found by Andrei Linde in 1980, in a short paper in Physics Letters B. Perturbation theory for the hot quark–gluon plasma can be pushed exactly to the logarithmic part of step 6, and no further (Kajantie et al. 2003). The missing number is not hidden behind a large computation. The series itself does not contain it. It has to be supplied from outside, from brute-force simulations of the field on a lattice.</p><p>When I first saw this, I had assumed that the only limits in calculations like these are limits of computing power. Here the limit is in the structure of the problem. On the cylinder you can even see where it sits: in the one loop that refuses to wind around time.</p><p><strong>Back to the second glow</strong><br>For the neutrino background none of this is a problem. The plasma there is made of light and electrons, and its interaction part is known through step 5. About a fifth of the deviation from the napkin value, N_eff − 3, is explained by steps 2 and 3 alone. The rest comes from the neutrinos leaving gradually, which needs a computer, but no new principle.</p><p>So the second glow is ∛(4/11) colder than the first, almost exactly. The “almost” is a sum of drawings on a cylinder of time.</p><p><strong>A number theorist’s footnote</strong><br>One detail kept me busy for a while. The factor 7/8 that makes electrons count less is 1 − 1/²³. In number theory this is not just any fraction. It is the factor of the prime 2 in the Euler product of the Riemann zeta function, the product over all primes that Euler wrote down in 1737. The alternating signs of fermions remove exactly the prime 2 from that product and put it back with a sign. The 11/4 in the neutrino temperature is 1 + 2 × (1 − 2⁻³).</p><p>The same happens one step further. Steps 2 and 3 are built from the same zeta values, Euler’s γ at step 5 is the constant of the zeta function at its pole, and at step 4 the functional equation of the zeta function appears. Why number theory sits so deep inside a hot plasma is the subject of the paper below.</p><p><strong>Read and check</strong></p><ul><li>Euler Factors at 2 in Neutrino Decoupling, with LaTeX source and a companion code that checks every number: doi:10.5281/zenodo.23805730</li><li>The Fermi Side: The Euler Factor at 2, the Sommerfeld Expansion and the Lorenz Number: doi:10.5281/zenodo.23322882</li><li>Particle Data Group, Neutrinos in cosmology: review</li><li>M. Laine, A. Vuorinen, Basics of Thermal Field Theory: doi:10.1007/978–3–319–31933–9</li></ul><p>Thomas Krause is an independent researcher in Lübeck, Germany. He writes the book series Geometry of Reality.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f1d1219cec68" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Two Sides of the Same Coin (EML & GoU)]]></title>
            <link>https://medium.com/@krause-virtuell/two-sides-of-the-same-coin-eml-gou-d60a37efb543?source=rss-ec2b54236e52------2</link>
            <guid isPermaLink="false">https://medium.com/p/d60a37efb543</guid>
            <category><![CDATA[mathematics]]></category>
            <category><![CDATA[number-theory]]></category>
            <category><![CDATA[geometry]]></category>
            <category><![CDATA[eml-operator]]></category>
            <category><![CDATA[mathematical-constants]]></category>
            <dc:creator><![CDATA[Thomas Krause]]></dc:creator>
            <pubDate>Sun, 25 Oct 2026 19:55:37 GMT</pubDate>
            <atom:updated>2026-11-04T17:50:49.041Z</atom:updated>
            <content:encoded><![CDATA[<p><strong>What a single operator for all elementary functions learns from continued fractions, and what that changes for the Geometry of Unity</strong></p><p>Imagine a calculator with one button and one constant. The button takes two numbers x and y and returns eˣ − ln y. The constant is 1. Earlier this year Andrzej Odrzywołek showed that this is enough: by nesting the button you can build every elementary function, addition, multiplication, roots, sine and cosine, and the constants e, π and i (Odrzywołek 2026). He calls the operator EML, for exp-minus-log.</p><p><strong>Key message: </strong>The paper proves that the <strong>EML calculator operator</strong> and <strong>classic continued fractions</strong> are two readings of the exact same mathematical mechanism, demonstrating that when continued fractions fail to converge, they oscillate as a perfect standing wave around the precise complex value hit by the EML operator.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*FQO8kOcFgWQcS0_H_gT2EA.png" /><figcaption><strong>Two Sides of the Same Coin: </strong>On the left, the golden growth of nature phi &amp; e; on the right, the oscillating interference pattern of the quantum world pi — united by a single mathematical engine.</figcaption></figure><p>I came to this result from the other end. In The Geometry of Unity (Zenodo) I build the three most famous constants, φ, e and π, from the unit alone, by three simple continued-fraction rules. Three rules, three constants. EML has one rule and gets all three, plus i.</p><p>So the natural question was: are these two different ways of making the same numbers, or are they the same thing seen from two sides? My new paper, Two Sides of the Same Coin, answers that they are the same mechanism, and it makes the common core explicit. This article is about what that means, for EML and for my own approach. The proofs, and the scripts that check every statement, are in the paper (link at the end).</p><p><strong>Two machines</strong><br>A continued fraction is a development. You start with a number, add a fraction, inside that fraction you add another one, and so on. Each stop along the way is a convergent, an approximation that gets better step by step. The golden ratio is the simplest case: 1 + 1/(1 + 1/(1 + …)).</p><p>EML looks different at first. It has no fractions, only exp and log. But a computer cannot evaluate a logarithm in one step either. It develops it, and the classical development of ln y is again a continued fraction, the Euler–Padé fraction. The same holds for eˣ. So inside every EML tree there are two kinds of developing nodes: exponential nodes and logarithmic nodes.<br>Once you look at EML this way, the two machines can be compared node by node.</p><p><strong>What EML gains</strong></p><p><strong>A map: one number decides the speed</strong><br>The first result is a single number, t, that tells you how any of these continued fractions converges. It is computed from the long-run shape of the fraction, and it sorts everything into three regimes:</p><ul><li>at t = 1, convergence is faster than any geometric rate. This is where the exponential node and my e-rule sit;</li><li>for t off the negative axis, convergence is geometric, with a rate fixed by t. The golden rule sits at t = 5;</li><li>on the negative axis, t ≤ 0, there is no geometric convergence at all.<br>For the logarithmic node the surprise is that t is simply y, the argument of ln y. So the speed of a logarithm is read off from the place where you evaluate it. And my three rules turn out to be three EML nodes at three definite places on this one axis: the golden rule is the node at t = 5, the e-rule is the exponential node at t = 1, and the π-rule is Brouncker’s 17th-century continued fraction, which reappears inside the logarithmic node at y = −1.</li></ul><p><strong>How EML reaches π</strong><br>This is the part I find most beautiful. EML gets π from ln(−1) = iπ. The button does it in one step. But what does the underlying development do at y = −1, where the convergents are real numbers and the target is not?<br>It does not converge, and it does not fail at random either. The paper proves an exact formula for every convergent on the negative axis, in terms of Legendre polynomials and their companions, and from it an interference law. The convergents form a standing wave. Its centre line is ln|y|, the real part of the logarithm, and its amplitude is π, the imaginary part. The development cannot reach its complex target, so it circles it. Geometrically, the convergents turn about the point ln y like a point on a rotating wheel, and every reasonable average of the wave returns exactly ln y.</p><p>So the one-step answer of EML, iπ, is the centre of a wave that the slow development keeps circling. Seen statistically, the convergents form a Lorentz profile, the familiar bell-like line shape of physics, whose position is the real part and whose width is the imaginary part of ln y. For the even convergents this structure is a special case of a theorem of Bowman and McLaughlin; the explicit centre and width are new as far as I know.</p><p>The wave repeats exactly when a certain angle belongs to a root of unity. For rational y this happens only at y = −1, −3 and −1/3. The golden ratio shows up here too, at y = −φ³/√5 and y = −√5·φ³, where the pattern has periods 10 and 5.</p><p><strong>Trees, rates and a cancellation</strong><br>For a whole EML tree, the paper proves that the error falls at the rate of its slowest logarithmic node. That is the expected rule. The unexpected part is that it can fail in a controlled way. If two equally slow nodes enter with opposite signs, their leading errors cancel, and the tree converges at the squared rate. The paper gives an explicit example built on the omega constant Ω, the solution of Ωe^Ω = 1.</p><p><strong>The exponential node has a wave of its own</strong><br>On the imaginary axis the continued-fraction approximants of eˣ never leave the unit circle, just as e^(iθ) itself does not. Their whole error is a phase. And that phase is a known quantity of classical physics: the phase shift that a hard sphere imprints on the n-th partial wave when a plane wave scatters off it. Nothing was put in to get this. It comes out of the development of e.</p><p>The same computation shows something about Bernoulli numbers. The n-th stage of Lambert’s continued fraction for tan z contains exactly the Bernoulli numbers B₂ up to B₂ₙ, and misses the next one by a precise amount. Bernoulli numbers are the numbers behind the Casimir constant (P1) and behind the values of the zeta function at negative integers (ZG).</p><p><strong>One expansion behind both nodes</strong><br>Physics students meet the Rayleigh expansion early: a plane wave written as a sum of partial waves, each a product of a radial factor (spherical Bessel functions) and an angular factor (Legendre polynomials). The paper shows that the two EML nodes are the two halves of this one expansion. The logarithmic node on its cut is a one-sided Fourier transform of the radial factor. The exponential node on the circle is a quotient of two one-sided transforms of the angular factor. Each node is written in the other node’s functions.</p><p><strong>Where EML-like operators stop</strong><br>Not every single operator works. Stachowiak studied the purely angular operator arccot x − cot y (Stachowiak 2026). The paper shows that it does generate addition, multiplication and division, contrary to a remark in the literature. But every closed tree is real, so i and iπ are out of reach, and neither eˣ nor ln x can be built from it on any interval. The angular gear alone is not enough.</p><p>A golden version of EML, φˣ − log_φ y, is built on φ instead of e. Assuming Schanuel’s conjecture, the paper proves that it cannot produce its own scale, ln φ, in its first layer. Whether this holds for all layers is open.</p><p><strong>What my approach gains</strong><br>The Geometry of Unity started from a simple idea: every number is a development from the unit, and φ, e and π are three ways of developing. What the comparison with EML adds is precision in three places.<br>One axis instead of three rules. The three rules are no longer three separate gifts. They are three positions on a single axis, the parameter t, and EML supplies the operator that moves between them. The golden ratio is the geometric regime, e is the super-fast regime, and π lives on the cut where geometric convergence breaks down.</p><p>Interference as a theorem. In my earlier work I spoke of an interference reading of numbers: a development leaves a remainder that behaves like a wave. That was an interpretation. On the logarithmic cut it is now a theorem with an exact formula. π is the amplitude of a standing wave whose centre is a real logarithm. The phrase “π as interference” now has a precise meaning that anyone can check.</p><p>A bridge to classical physics without new physics. My research is not about new forces or predictions. It is about an arithmetic explanation of why the same few numbers keep returning in classical physics. The hard-sphere phase shift, the Rayleigh expansion and the Bernoulli numbers of the Casimir effect are textbook physics. Here they appear as properties of the developments of e and ln, with no physical assumption added. That is the kind of explanation I am after: the physics does not change, but it becomes clearer why these numbers appear.</p><p><strong>What is not claimed</strong><br>Every result in the paper carries a status label: proved, proved and cited, conditional, framework-internal, numerical or open. The golden-operator result depends on Schanuel’s conjecture, which is unproved. The arithmetic reading of the results, as one machine with an angular and a radial gear, is an interpretation and is labelled as such. No new physical mechanism is proposed. And several pieces are classical: the Euler–Padé fraction, Brouncker’s fraction, the Bessel form of the Padé approximants of eˣ, the hard-sphere phase shift. What is new is how they fit together.</p><p>Three questions are left open in the paper. Does the golden-operator result extend beyond the first layer? Do the small correction terms of the wave have a direct reading through the one-sided transforms? And is there one joint transform from which both node theorems and the tree rate follow at once?</p><p><strong>Read and check</strong><br>The paper and all scripts with their logs are open:</p><ul><li>Two Sides of the Same Coin: The EML Operator, Continued-Fraction Rules of Unity, and an Interference Law on the Logarithmic Branch Cut, Zenodo: <a href="https://doi.org/10.5281/zenodo.23648701">https://doi.org/10.5281/zenodo.23648701</a></li><li>The Geometry of Unity: <a href="https://doi.org/10.5281/zenodo.22857699">https://doi.org/10.5281/zenodo.22857699</a></li><li>The geometry of zeta values: <a href="https://doi.org/10.5281/zenodo.23647944">https://doi.org/10.5281/zenodo.23647944</a></li><li>The Casimir Constant as a Bernoulli Number: <a href="https://doi.org/10.5281/zenodo.23301747">https://doi.org/10.5281/zenodo.23301747</a></li><li>A. Odrzywołek, All elementary functions from a single binary operator: arXiv:2603.21852</li></ul><p>If you have worked on EML-type operators or on continued fractions on the branch cut, I would be glad to hear from you.</p><p>Thomas Krause is an independent researcher in Luebeck, Germany. He writes the book series <strong>Geometry of Reality.</strong></p><p><strong>The Shape of Numbers:</strong> <a href="https://www.amazon.com/dp/B0HL3W32DS">https://www.amazon.com/dp/B0HL3W32DS</a></p><p><strong>The Prime Machine:</strong> <a href="https://www.amazon.com/dp/B0H7X876KM">https://www.amazon.com/dp/B0H7X876KM</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d60a37efb543" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[On Primes, Kissing Numbers & the Monstrous Moonshine]]></title>
            <link>https://medium.com/@krause-virtuell/primes-on-the-kissing-shell-e5ead7da3acc?source=rss-ec2b54236e52------2</link>
            <guid isPermaLink="false">https://medium.com/p/e5ead7da3acc</guid>
            <category><![CDATA[geometry]]></category>
            <category><![CDATA[mathematics]]></category>
            <category><![CDATA[number-theory]]></category>
            <category><![CDATA[physics]]></category>
            <category><![CDATA[prime-numbers]]></category>
            <dc:creator><![CDATA[Thomas Krause]]></dc:creator>
            <pubDate>Sun, 11 Oct 2026 08:43:37 GMT</pubDate>
            <atom:updated>2026-11-04T17:54:48.237Z</atom:updated>
            <content:encoded><![CDATA[<p><strong>What the densest sphere packings know about prime numbers, and how far that leads toward the Monster</strong></p><p><strong>Key message: </strong>The underlying paper to this article proves that the minimal generation lengths of prime ideals in the densest sphere packings E8 and the Leech lattice are strictly determined by number-theoretic residue classes, showing that 91% of the Leech lattice’s kissing shell consists of prime generators directly linked to the algebraic structure of the Monster group.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*mTgBExoA9HbXsIZYE9Bbmw.png" /><figcaption>Illustration of Kissing Spheres</figcaption></figure><p>Put a ball on a table and ask how many balls of the same size can touch it at once. In the plane the answer is six, the familiar honeycomb. In three dimensions it is twelve, a question Newton argued about. In eight dimensions the answer is 240, and in twenty-four dimensions it is 196 560.</p><p>These two are special. In both dimensions there is a single, perfectly regular arrangement that achieves this kissing number and packs space as densely as possible: the lattice E8 and the Leech lattice. Both records are now proved theorems.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*KXhYxiedttH6f_SWDODpcg.png" /><figcaption>Remainder of mod30</figcaption></figure><p>These two lattices are among the most symmetric objects in mathematics. The symmetry group of the Leech lattice leads, by a short chain of constructions, to the Monster, the largest of the sporadic simple groups. And the Monster, through “monstrous moonshine”, is tied to the coefficients of one of the oldest functions in number theory, the modular j-function.</p><p>So a natural question is whether prime numbers have a place in all of this. Not in a vague sense. Can you point at a prime and say where it sits in the Leech lattice?</p><p>That is the question of my new paper, First Shells of Prime Ideals in Cyclotomic Lattices. This article is about what the answer means, not about the proofs. The paper and all code are open (link at the end).</p><p><strong>A lattice that is also a number system</strong><br>The starting observation is old, and it is the key to everything else. E8 and the Leech lattice are not only arrangements of points. They can be written as number systems.</p><p>Take the complex numbers built from a root of unity: for E8 a 30th root, for the Leech lattice a 35th root. Their integer combinations form a ring of “cyclotomic integers”. You can add and multiply in it as with ordinary integers. With the right notion of length, the points of this ring are exactly the points of E8 or of the Leech lattice. The 240 or 196 560 kissing points then become particular numbers in that ring.</p><p>In such a ring, primes do not disappear. They split. An ordinary prime p breaks into “prime ideals”, and each prime ideal is generated by numbers, which are points of the lattice. So every prime gets a set of points that represent it, and among them there is a closest one to the centre.</p><p>I call the distance of that closest point the first shell of the prime. Rescaled by the size of the prime, it becomes a single number, ρ, which tells you how efficiently a prime manages to sit close to the centre. ρ = 1 means the prime sits as close as the geometry of the lattice allows.</p><p><strong>E8: the remainder decides</strong><br>For E8 the picture is complete, and it is surprisingly clean. Every prime larger than 5 leaves one of eight remainders when divided by 30. These are the same eight numbers that label the corners of the 16-cell in my earlier books. The remainder alone decides how the prime behaves:</p><ul><li>For remainders 7, 13, 17 and 23 the prime is rigid: it sits at exactly the closest possible distance, and the set of its points is a perfect scaled copy of E8 itself.</li><li>For 11 and 19 the first shell follows a golden law. It is read off from the way p can be written as a² + ab − b², the number system of the golden ratio. Fibonacci appears inside E8 without being put there.</li><li>For 29 it is a sum of four squares inside a smaller lattice shaped by the 16-cell.</li><li>For remainder 1 there is no simple formula. The values scatter in a pattern governed by Hecke’s equidistribution theorem from 1918.<br>And in every class ρ ≤ 1. A prime never sits farther out than the densest arrangement allows. The paper proves this in three independent ways. One proof can be checked by hand, one uses certified computation, and one rests on Blichfeldt’s theorem from 1935 that E8 is the densest lattice in eight dimensions.</li></ul><p><strong>The Leech lattice: primes on the kissing shell</strong><br>In twenty-four dimensions the same idea gives a first striking fact. Of the 196 560 kissing points, 178 920 represent primes or products of primes. So primes such as 71 and 211 do not just live somewhere in the Leech lattice. They lie on the kissing shell itself, among the balls that touch the central one.</p><p>For most remainders modulo 35 the paper proves exact laws for the first shell. Some classes are rigid again. Some follow the same golden law as E8. For others the answer is the corner of an explicit polytope, found by a method I call the facet criterion. The method turns an infinite search into finitely many exact checks. Its hardest case needed 194 different facet types.</p><p>Two classes remain open: the primes with remainder 1 or 34 modulo 35. There, the space that has to be searched has eleven dimensions, and none of the methods reaches that far yet.</p><p><strong>Interpretation: three layers</strong><br>The most useful thing to come out of the work is not a single theorem but a way of seeing why the theorems hold. There are three layers.</p><p><strong>Geometry.</strong> E8 and the Leech lattice are the densest lattices in their dimensions. That alone forces every subset of them shaped like a prime to be no denser than the whole. This part comes for free.</p><p><strong>Arithmetic</strong>. The real question is whether the shortest point inside a prime’s region actually represents the prime, or is merely a point that happens to lie there. In E8 the answer is always yes, and the reason fits in one line: there are no small numbers in the ring that could get in the way. The smallest non-trivial “number size” is 16, which is too large to spoil anything. In the Leech lattice the smallest one is 71, and that is not large enough for the same argument.</p><p><strong>Symmetry</strong>. The lattice has symmetries that mix the different prime ideals of one prime. These make the first shell depend only on the prime, not on which of its ideals you look at, and they reduce the work in the hard cases.</p><p>The arithmetic layer is where the open problem now sits. It can be stated without any polytopes or tori:</p><p><strong>Conjecture. In the Leech lattice, the shortest point of every prime ideal represents that prime ideal.</strong></p><p>I checked it for 1674 prime ideals, without a single exception. If it is true, the open classes close at once, because the Cohn–Kumar theorem (2009), which says that the Leech lattice is the densest lattice in 24 dimensions, then does the rest.</p><p>There is a telling contrast. For lattices that are not the densest in their dimension, such as the family A_{p−1}, the bound ρ ≤ 1 fails, infinitely often, for every prime p from 7 to 43. In some of them the shortest point of a prime ideal is indeed not a representative. Primes behave best exactly where the packing is best.</p><p><strong>And the Monster?</strong><br>This brings me to the question I was asked most often while writing: so have we connected the primes to monstrous moonshine?</p><p>The honest answer is: partly, and it is important to say which part.<br>Here is what is true and proved. Multiplying by the 35th root of unity is itself a symmetry of the Leech lattice. Inside the symmetry group of the Leech lattice, Conway’s group Co₀, it has a fingerprint called its Frame shape, 1·35/(5·7). Through the correspondence of Conway and Norton, this fingerprint belongs to a specific element class of the Monster, called 35B. Its “moonshine function” is a clean expression in Dedekind’s eta function, η(5τ)η(7τ)/(η(τ)η(35τ)), minus 1.</p><p>The new part is this. The symmetries that the paper had to construct to handle the primes are twisted versions of the Galois maps of the number system. They turn out to be, up to roots of unity, all of the symmetries that respect the 35th root of unity: 1680 of them. So the remainder of a prime modulo 35 picks out a definite family of Leech symmetries. Each family carries Frame shapes, and these point to Monster classes. In the rigid classes all 560 symmetries share one fingerprint.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*UI9BHDdikco0fUvL9xULFQ.png" /></figure><p>What this gives is a dictionary. Every prime can be translated, through its remainder, into a set of Frame shapes of Conway’s group, and for most remainders also into a set of Monster classes; every step of the translation is a theorem. The translation does not run backwards to individual primes: a class that occurs determines an infinite family of primes, defined by remainders modulo 35, and most Monster classes do not occur at all.</p><p>What it does not give, yet, is a bridge. A bridge would mean that the numbers we computed for an individual prime, its first shell, show up in the moonshine functions themselves, or are predicted by them. I have not found such a relation, and I have not seriously tested for one. Until someone does, it would be wrong to say that the primes have been connected to the Monster in the way that moonshine connects the Monster to the j-function.</p><p>It is also fair to say what is not new. That the Leech lattice can be written over the 35th roots of unity goes back to Craig in 1978. The correspondence between that symmetry and the Monster class 35B is in the tables of Conway–Norton (1979) and Lang (1989). What the paper adds is the link from the prime-by-prime laws to that symmetry.</p><p><strong>A coincidence worth defusing</strong><br>One number tempts every reader: 71. It is the smallest prime that splits in the Leech number system, and it is also a factor of 196 883, the dimension of the smallest non-trivial representation of the Monster (196 883 = 47 · 59 · 71). Is that a sign?</p><p>№71 is simply the smallest prime that leaves remainder 1 on division by 35, since 71 = 2 · 35 + 1. That is why it is the smallest size of a prime ideal in this ring, and it has nothing to do with the Monster. I mention it because this is exactly the kind of coincidence that can carry a theory away. In the paper it is listed as a coincidence.</p><p><strong>How the work was done</strong><br>Every statement in the paper carries a label: proved, proved relative to a cited theorem, computer-assisted, numerical, conditional, framework-internal, open, refuted or withdrawn. There is a separate section of negative results and corrections. It includes ideas that did not work, a claim I had to withdraw, and errors in my own test scripts that were caught and fixed along the way. One example: a naive expectation that larger open search spaces come with more Frame shapes turned out to be false. All computations are in an open archive with expected outputs and checksums, so anyone can re-run them.</p><p>None of this makes a physical claim. Moonshine is sometimes discussed in connection with string theory. This paper is pure mathematics, and the dictionary above is a statement about lattices and groups, nothing more.</p><p><strong>What it implies</strong><br>For number theory, the work turns a geometric question into an arithmetic one. Whether a prime sits as close to the centre as possible depends on whether the shortest points of its ideals represent it. That question makes sense for every number system of this kind, and the computations show that the answer really varies: yes for E8, D4 and, so far, the Leech lattice, and no for some others.</p><p>For the larger programme behind my books, it is a confirmation of a principle I keep testing: a geometric picture of the primes is only worth something if it computes things that can be checked independently. Here the 16-cell remainders modulo 30 decide the laws in E8, the golden ratio appears on its own, and the densest lattices turn out to be the most “prime-friendly” ones.</p><p>For the Monster, the door is open but not walked through. There is a precise dictionary from primes to moonshine classes. Whether anything flows back across it, whether a moonshine function knows the first shell of a prime, is now a well-posed question. I would be glad to hear from anyone who wants to try.<br><br>The paper, its LaTeX source and the complete verification code are available on Zenodo: <a href="https://zenodo.org/records/23293398">https://zenodo.org/records/23293398</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e5ead7da3acc" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Numbers Have a Geometric Shape — and It Computes Something Real]]></title>
            <link>https://medium.com/@krause-virtuell/numbers-have-a-shape-and-it-computes-something-real-4f6e80a3048c?source=rss-ec2b54236e52------2</link>
            <guid isPermaLink="false">https://medium.com/p/4f6e80a3048c</guid>
            <category><![CDATA[quantum-number-theory]]></category>
            <category><![CDATA[number-theory]]></category>
            <category><![CDATA[ontology]]></category>
            <category><![CDATA[geometry]]></category>
            <category><![CDATA[mathematics]]></category>
            <dc:creator><![CDATA[Thomas Krause]]></dc:creator>
            <pubDate>Sat, 10 Oct 2026 10:14:01 GMT</pubDate>
            <atom:updated>2026-10-13T10:49:46.486Z</atom:updated>
            <content:encoded><![CDATA[<p><strong>Inside “The Shape of Numbers”, the third volume of Geometry of Reality</strong></p><p>Take a handful of matchsticks. Lay one on the table, then a second at its end, then a third that closes the two into a triangle. From then on add one matchstick at a time, following four fixed rules: close a figure if you can, otherwise echo a figure you already have, never count a mere repetition as something new, and open a new direction only when nothing else applies.</p><p>The game records just one thing: which steps open something new. And from the third step on, those steps fall exactly on the prime numbers.<br>That observation was the starting point of the second volume of my series Geometry of Reality. The third volume, The Shape of Numbers: An Introduction to Quantum Number Theory, asks the question that follows from it, and it is a harder one.</p><p><strong>The question</strong><br>The matchstick game leads naturally to the number 30 = 2 · 3 · 5. Every prime larger than 5 leaves one of only eight remainders when divided by 30:</p><p>1, 7, 11, 13, 17, 19, 23, 29</p><p>These eight numbers fit exactly onto the eight corners of the 16-cell, a regular four-dimensional polytope, the four-dimensional relative of the octahedron. Opposite corners pair up as 1 with 29, 7 with 23, 11 with 19 and 13 with 17. Each pair adds up to 30.</p><p>Volume II built a deterministic “prime machine” on this figure. But a picture can be decoration. So The Shape of Numbers starts from a simple test:</p><p>If this shape is real, it should not only describe the primes. It should compute something that number theorists already know how to compute by completely different means.</p><p>If the shape is decoration, the two computations will have nothing to say to each other. If the shape is structure, they will agree, and the agreement will explain something.</p><p>They agree.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*iAr3edJXFU-tdfWaH2q_aA.png" /></figure><p><strong>One map, many names</strong><br>The whole book turns on one surprisingly modest operation: the map x ↦ −x.</p><ul><li>On the 16-cell it sends every corner to the opposite corner.</li><li>On the remainders modulo 30 it gives the four pairs above.</li><li>On the complex plane it is complex conjugation.</li><li>On the zeta and L-functions it decides, through the sign χ(−1) of each character, how the reflection s ↦ 1 − s acts across the famous critical line.</li></ul><p>These look like four different things. The book shows that they are one involution seen from four sides, and that this one map organises the prime machine of Volume II and a large part of classical algebraic number theory at the same time.</p><p><strong>A facet that turns out to be a class number</strong><br>This was the moment that made me write the book.</p><p>Choose one corner from each opposite pair of the 16-cell and you have chosen a facet, one of its three-dimensional faces. That is pure geometry.</p><p>Now go to the 19th century. When unique factorisation into primes breaks down in number systems built from roots of unity, the size of the failure is measured by the class number. Kummer showed that a key part of it, the “minus part”, is controlled by certain half sums: add up the values of a character over the first half of the residues. Maillet later turned this into a determinant.</p><p>The book shows that these half sums are exactly the Fourier coefficients of a facet of the cross-polytope. The facet is a geometric object. The class number is an arithmetic one. They are the same data.</p><p>A small example you can check by hand: modulo 30 the four relevant half sums are 1, 1 + i, 1 − i and 2. Their sizes, read through what the book calls the ladder identity, give the class number of the corresponding cyclotomic field: exactly 1. In Chapter 7 a whole class number comes out of a 3 × 3 matrix that you can write down on a napkin.</p><p><strong>The even half and the odd half</strong><br>Once this is seen, the prime machine of Volume II finds its place. It turns out to see only what is even under x ↦ −x: orbits of squares. The book proves that it is blind to everything that is odd.</p><p>And the odd half is where the class numbers live, together with the Bernoulli numbers, the signs of cyclotomic units and the 2-parts of class groups. I call this odd half quantum number theory.</p><p>The word needs an honest explanation, and the book gives one in Chapter 21. It has nothing to do with physics. No constant of nature is derived, and no claim is made about the physical world. The word is borrowed because the central invariants behave like quantised quantities. They take discrete values, they are governed by an integer level, and they obey selection rules that forbid the values in between. A reader who dislikes the word can replace it by “cyclotomic 2-adic” and lose nothing.</p><p><strong>What the book proves</strong><br>The book collects ten main results. They fall into three groups:</p><p><strong>Strcture:</strong> what the geometry actually is</p><ul><li>The class number is a facet determinant.</li><li>The prime machine is the even quotient of a larger picture.</li><li>The spectral theorem of the Witt tower: at every stage of the tower the spectrum is that of the 16-cell, only scaled, which explains in one line why symmetric operators, however cleverly chosen, cannot see class numbers</li></ul><p><strong>Measurement:</strong> how large the 2-part of a class number is</p><ul><li>A half sum on one facet measures it, so a question that usually needs heavy machinery becomes a parity count.</li><li>The level theorem: each prime of the conductor that splits contributes one factor of 2. This is the selection rule that gives the theory its name, and it is proved in full generality.</li><li>The second source and the chirality law, and the ladder identity that connects the levels of the tower.</li></ul><p><strong>Detectors:</strong> deciding field by field whether a specific piece of a class group or unit group is present</p><ul><li>The witness prime system.</li><li>The Gauss lemma vector as a block detector.</li><li>Dual blocks and the quarter formula: an exact rule that decides which of two opposite facets a field “selects”, by counting residues in the first quarter of an interval. With a Chinese-remainder version of the count, fields with conductors up to 5 × 1⁰¹² are decided in seconds.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*a8HclRK11ywDDHGKxYslBQ.png" /></figure><p><strong>And the Riemann Hypothesis?</strong><br>The book has a dedicated outlook chapter, and it is careful. It does not prove the Riemann Hypothesis or the Hodge conjecture, and it says so plainly.</p><p>What it does show is where the Riemann Hypothesis “hides” in the objects of the book. A classical theorem of Nicolas (1983) says that the hypothesis holds if and only if the share of units in the primorial wheels 30, 210, 2310, 30030, … stays below an explicit bound at every stage. Read on the wheel, the Riemann Hypothesis becomes, among other things, a statement about the stability of the prime machine’s tower. The book gives three elementary reformulations of it in the language of its own constructions. It keeps clear which equivalences are classical, which translations are exact, and which questions remain open.</p><p><strong>A book you can check yourself</strong><br>The book is written for three kinds of reader at once: the curious reader who wants to know where the story goes; the professional mathematician who wants to know whether there is anything to it; and the specialist who already knows characters, Bernoulli numbers and class field theory and is looking for a new, intuitive geometry behind them. In that geometry a class number is a facet, a unit is a chord, and a block of a class group is a direction on a polytope.</p><p>To serve all three, roughly seventy per cent of the text is explanation, figures and worked examples, and thirty per cent is proof sketches. Each sketch points to the research paper where the full proof and the code are deposited.</p><p><strong>Three features matter to me more than anything else:</strong></p><ul><li>Every statement carries exactly one status label: proved, proved relative to a cited theorem, numerical, open, refuted, withdrawn and a few more. You always know what has been established and what has not. If you follow only the statements labelled proved and proved relative, you get a coherent chain from the matchstick game to the 2-adic structure of class groups.</li><li>Every chapter ends with exercises and full solutions.</li><li>Every number in the book can be recomputed. A computational companion contains one short Python script per chapter. Wherever the text says compute it yourself, it names the script that does it. You are invited to change the parameters and test the statements on your own examples.</li></ul><p>The book closes with a research programme of open problems, sorted by entry level from bachelor projects to genuine research questions. If you have ever wanted to work on a question in number theory that you can actually start on this weekend, this is meant for you.</p><p><strong>Who it is for</strong><br>The Shape of Numbers does not require you to have read the earlier volumes. Part I recalls everything needed from Volume II in about thirty pages. It does not ask you to believe anything either. It asks you to look, compute and check.</p><p>If you have ever suspected that numbers are not just points on a line, but that they have a shape, this book shows what that shape computes.<br><br>The Shape of Numbers: An Introduction to Quantum Number Theory<br>Geometry of Reality, Volume III — Thomas Krause</p><p><em>→ Hardcover on Amazon: </em><a href="https://www.amazon.com/dp/B0HL3W32DS"><em>amazon.com/dp/B0HL3W32DS</em></a></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/350/1*bfnFz6DvqFWvHXTWd595rw.jpeg" /><figcaption>Volume 3: The Shape of Numbers</figcaption></figure><p><strong>Also in the series</strong><br>Geometry of Reality, Volume II — The Prime Machine: From the Matchstick Game to the Half-Line</p><p>Where the story of the shape begins. A matchstick game with four simple rules turns out to find the prime numbers by itself. Its addresses lead to the eight residue classes modulo 30 and to the 16-cell, and from there to a deterministic machine whose orbit minima are the primes. The book then reads classical questions on this geometry, from Bertrand’s postulate and twin primes to Goldbach, Polignac gaps, Hardy–Littlewood constellations and Collatz, and it ends on the half-line of the zeta function with an observation that Volume III takes up.</p><p><em>→ Hardcover: </em><a href="https://www.amazon.com/dp/B0H7X876KM"><em>amazon.com/dp/B0H7X876KM</em></a><em><br>→ Paperback: </em><a href="https://www.amazon.com/dp/B0HL4ND8WT"><em>amazon.com/dp/B0HL4ND8WT</em></a></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/350/1*a8yHOMURySzdJNEOlWM39g.jpeg" /><figcaption>Volume 2: The Prime Machine</figcaption></figure><p>Geometry of Reality, Volume I — From the First Tick to the Elementary Particles</p><p>The book that started the series, and the one that follows a physical thread. Starting from nothing more than a point and a beat, it builds lines, triangles, the pentagon and the golden ratio, the icosahedron and its symmetry group, and it asks whether such structures could underlie the numbers of physics: Sommerfeld’s 137, the spread of the fermion masses, three generations of matter. It is written as a proposal, not a proof, and it contains its own chapter on what would overturn it. Volumes II and III were written to secure the mathematics on which a later return to that physical thread will rest.</p><p><em>→ Hardcover</em><a href="https://www.amazon.com/dp/B0GZKGNPF4"><em>: amazon.com/dp/B0GZKGNPF4</em></a></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/350/1*Eto6vuMODh2sxPMD-LmlqQ.jpeg" /><figcaption>Volume 1: Geometry of Reality</figcaption></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=4f6e80a3048c" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[I Usually Write About Prime Numbers. This Time, It’s About You!]]></title>
            <link>https://medium.com/@krause-virtuell/i-usually-write-about-prime-numbers-this-time-its-about-you-b863a78f59b9?source=rss-ec2b54236e52------2</link>
            <guid isPermaLink="false">https://medium.com/p/b863a78f59b9</guid>
            <category><![CDATA[philosophy]]></category>
            <category><![CDATA[infj]]></category>
            <category><![CDATA[mtbi]]></category>
            <category><![CDATA[social-change]]></category>
            <category><![CDATA[personality-types]]></category>
            <dc:creator><![CDATA[Thomas Krause]]></dc:creator>
            <pubDate>Mon, 21 Sep 2026 22:49:55 GMT</pubDate>
            <atom:updated>2026-09-21T22:49:55.662Z</atom:updated>
            <content:encoded><![CDATA[<h3>I Usually Write About Prime Numbers. This Time, It’s About You</h3><p><strong>A thought experiment about roles, signatures, and a tree instead of a pyramid.</strong></p><p>If you’ve read anything of mine before, it was probably about the Riemann Hypothesis, the geometry of prime numbers, or some attempt to trace physical constants back to arithmetic structure. Zeta zeros, prime gaps, symmetry groups — that’s usually my territory.</p><p>This piece is not about any of that. It’s about something much closer to home: how you see yourself, how you relate to the people around you, and — as a thought experiment, nothing more — what a society might look like if it took those things seriously.</p><p>I call the idea the Sacred Tree of Civilization. It started as a personal exercise in self-understanding, grew into a relationship framework, and eventually turned into an open question about how power and responsibility could be distributed differently. None of it requires you to believe in my other work. It stands entirely on its own.</p><p><strong>The Problem With Titles</strong><br>Most of the structures we live inside — companies, families, even entire political systems — are organized like a pyramid. Few people at the top hold titles; many people below execute. Worth gets measured by position rather than by what someone actually contributes. That produces a familiar kind of quiet pressure: the sense that you’re “not enough” whenever you don’t fit into the drawer that was built for you.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*TCCyV3lPvIyn5FqdfjdVpw.jpeg" /></figure><p>The Sacred Tree proposes a different image. Instead of a pyramid, picture a tree: roots (values and history), trunk (structure), branches (domains like education, economy, relationships), and leaves (individual people, each with their own signature). In this picture, everyone has a place — not because of where they came from or what their title says, but because of the role they actually carry.</p><p><strong>Your Personal Signature</strong><br>The core idea is that every person carries an individual “signature” — a combination of personality type and role profile. The role profile itself comes from a simple 3×3 matrix:</p><p>• Three layers of personality — Core (your inner center of gravity), Shell (how you protect yourself and connect with others), and Light (how you show up in the world).</p><p>• Three basic functions — Carrying (holding, stabilizing), Connecting (mediating, building bridges), and Moving (changing, driving things forward).</p><p>Combine the two and you get nine role fields — from the calm, dependable “Core-Carrying” type to the creative, idea-driven “Light-Moving” type. The point isn’t to sort people into boxes. It’s to give people a language for understanding what comes naturally to them, and where they burn out if they spend years living against their actual role.</p><p><strong>Importantly</strong>: assignment happens exclusively through your own honest self-recognition. No test, no biological marker, no external authority decides your type for you — and you’re always free to revise it as you keep developing.</p><p><strong>Relationships, Reimagined</strong><br>The same logic extends to relationships. A Relationship Matrix describes how two role profiles interact energetically — as strong attraction with a lot of friction (Red), as a stable everyday foundation (Green), or as a calm, relieving quality (Blue). Once you understand which combination you’re in with a partner, conflicts become easier to place: not as personal failure, but as a familiar dynamic of a particular role pairing. The same thinking extends to parenting — accompanying a child’s own developing role instead of turning them into a project that has to succeed.</p><p><strong>Part II: What If a Whole Society Worked This Way?</strong><br>Here’s where the thought experiment gets bigger — and where I want to be extremely clear about its limits before I describe it.</p><p>What if political power followed role instead of status, origin, or volume? What if a democracy elected people, but only into positions that genuinely matched their type and demonstrated suitability — a kind of “democracy with fit”? What if all sixteen personality types were understood as equally necessary, with none ranked above another? What if every person could revise their role at any point in their life, the same way they can revise their own signature?</p><p>That’s the whole vision, stripped to its structure: not a pyramid with a narrow peak, but a tree with roots, a trunk, branches, and leaves, where power distributes the way a tree distributes growth — outward and upward at once, never concentrated at a single point.</p><p><strong>What this explicitly is not, and this matters more than the vision itself:</strong></p><p>• It is not a prediction with a date attached, and not a claim about an imminent development. It’s a possible direction to think through, not a countdown.</p><p>• It involves no coercion and no claim of superiority. Nobody is forced to have their type determined or to take on a role that isn’t genuinely their own. A society that used this model to exclude, sort, or rank people would have fundamentally misunderstood and perverted it.</p><p>• It has no biological or deterministic basis. Type assignment happens exclusively through self-knowledge — never through birth, origin, or a supposed biological formula.</p><p>• It is not a substitute for the rule of law, separation of powers, an independent judiciary, or a free press. A functioning version of this vision would presuppose and strengthen all of that, not replace it.</p><p>You don’t need to believe the full vision could ever be implemented to get something out of it. Just sitting with the question “What if power followed role instead of status?” changes how you look at your own work, your own relationships, your own organization. That’s the actual value of a thought experiment: not as a blueprint, but as a compass for a lot of small, voluntary decisions.</p><p><strong>Why I’m Sharing This</strong><br>There’s no business behind this. No coaching offer, no subscription, no consulting funnel. The full text — both the personal framework in Part I and the societal thought experiment in Part II, along with worksheets for individuals, couples, and parents, plus notes for coaches, educators, and therapists — is written up as a complete, freely downloadable book:</p><p><strong>The Sacred Tree of Civilization: </strong><a href="https://a-toe.org/"><strong>https://a-toe.org/</strong></a></p><p>You can read the whole thing, share it with anyone it might be useful for, or even load it into an AI conversation as a knowledge base and let it question you about your own signature. None of those uses require anything from you in return — that’s the entire idea of a seed rather than an institution.<br>If this piece gave you a new way to talk about yourself, your relationships, or the people you’re raising — even just a small one — it’s done what it set out to do. The rest, as always, is written by you.<br><br>The full book, “The Sacred Tree of Civilization,” is available free as a PDF at <a href="http://www.a-toe.org/">www.a-toe.org</a> — no login, no paywall, no email required.</p><p>I’m the same author behind a-toe.de and the Arithmetic Theory of Everything, an unrelated research program on prime numbers and fundamental physics. The two projects are deliberately independent — you don’t need one to make sense of the other.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b863a78f59b9" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>