# References — Chapter ᚐ(A): The mechanic’s workshop *Teclam Áirme, the Mathematical Compendium — a copy for AI readers of https://www.teclamairme.com/#ref-ch-17, as of 2026-09-30. Links are relative to https://www.teclamairme.com/fortheaireaders/.* [Guide for AI readers](index.md) · [The chapter](ch17.md) *Each entry: one plain sentence saying exactly what it is cited for, its status, then every source with a note saying what that source actually carries. "Own computation" / "Own reading" entries are the author's own work, not outside sources.* ### 17.1 {#r-17.1} The chapter’s gamma, beta, hypergeometric, Bessel, Airy, error, Fresnel and orthogonal-polynomial functions are standard special-function families represented through integrals, series, differential equations and related constructions. *Status: confirmed* - [NIST Digital Library of Mathematical Functions](https://dlmf.nist.gov/) — the authoritative chapter index organizes the named families and their definitions. - [Andrews, Askey and Roy, Special Functions](https://www.cambridge.org/core/books/special-functions/1F1C575CEA780EE774F5518C7963BF08/listing) — the publisher’s contents and preface describe the subject’s long use and the book’s treatment of these families. ### 17.2 {#r-17.2} Gamma extends factorials, has no zeros and simple poles at the nonpositive integers, the pole at −n with residue (−1)ⁿ/n!, was introduced by Euler in 1729 correspondence with Goldbach, and received the Γ notation in Legendre’s 1811 volume. Sources differ on the year of Legendre’s Γ: 1811 (Miller, transcribing Legendre’s Exercices de calcul intégral, p. 277) or 1814 (Encyclopedia of Mathematics); corrections welcome. *Status: confirmed* - [NIST DLMF §5.2, Gamma Function](https://dlmf.nist.gov/5.2) — defines gamma, its meromorphic continuation, factorial relation and simple poles: “It is a meromorphic function with no zeros, and with simple poles of residue (−1)^{n}/n! at z = −n.” - [Encyclopedia of Mathematics, Gamma-function](https://encyclopediaofmath.org/wiki/Gamma-function) — records Euler’s 1729 introduction in a letter to Goldbach; it gives 1814 for Legendre’s symbol Γ and the name gamma-function (“proposed in 1814 by A.M. Legendre”). - [MacTutor, Earliest Uses of Function Symbols](https://mathshistory.st-andrews.ac.uk/Miller/mathsym/functions/) — transcribes the Γ notation from page 277 of Legendre’s 1811 volume. ### 17.3 {#r-17.3} Γ(1/2)=√π and Γ(3/2)=√π/2; the recurrence identifies these with (−1/2)! and (1/2)! respectively. *Status: confirmed* - [NIST DLMF §5.4, Special Values](https://dlmf.nist.gov/5.4) — gives the half-integer gamma values. - [NIST DLMF §5.5, Functional Relations](https://dlmf.nist.gov/5.5) — gives Γ(z+1)=zΓ(z). ### 17.4 {#r-17.4} The logarithmic integral uses a principal value across t=1, and the prime number theorem establishes its asymptotic relation to π(x); Hadamard and de la Vallée Poussin proved the theorem in 1896 after Gauss’s estimate. The prime number theorem can be written π(x) ∼ li(x). *Status: confirmed* - [NIST DLMF §6.2, Exponential and Logarithmic Integrals](https://dlmf.nist.gov/6.2) — defines li by the Cauchy principal value. - [NIST DLMF §27.2, Functions of Number Theory](https://dlmf.nist.gov/27.2) — (eq. 27.2.3): states the prime number theorem in the form π(x) ∼ x/ln x, credits the 1896 proofs of Hadamard and de la Vallée Poussin, and names Gauss and Legendre for the conjecture. It does not use li or the words logarithmic integral. - [NIST DLMF §27.12, Asymptotic Formulas: Primes](https://dlmf.nist.gov/27.12) — (eq. 27.12.5): bounds |π(x) − li(x)| by O(x exp(−c(ln x)^½)), the sharper form from which π(x) ∼ li(x) follows. This is the DLMF section that ties π(x) to li(x). - [Logarithmic integral function, Wikipedia](https://en.wikipedia.org/wiki/Logarithmic_integral_function) — read as raw wikitext: “The function 1/(ln t) has a singularity at t = 1, and the integral for x > 1 is interpreted as a Cauchy principal value”; “the prime number theorem states that: π(x) ∼ li(x).” ### 17.5 {#r-17.5} The Ramanujan–Soldner constant is the unique positive zero μ≈1.451369234883; Soldner’s table appeared in 1809, Ramanujan later used the associated expansion, and a 2019 arithmetic survey reports that irrationality of μ and log μ is unknown. Ramanujan computed μ (as 1.45136380…, the correct value being 1.45136923488…). *Status: confirmed* - [OEIS A070769, Ramanujan–Soldner constant](https://oeis.org/A070769) — gives the value and cites Soldner’s 1809 Théorie et tables d’une nouvelle fonction transcendante. - [Wikipedia, Logarithmic integral function](https://en.wikipedia.org/wiki/Logarithmic_integral_function) — defines the Ramanujan–Soldner constant μ ≈ 1.45136 92348 as the unique positive zero of li(x), and gives Ramanujan’s more rapidly convergent series. - [Johann Georg von Soldner, Wikipedia](https://en.wikipedia.org/wiki/Johann_Georg_von_Soldner) — “He also published on the logarithmic integral function”; lists “Soldner, J. G. v. (1809). Théorie et tables d’une nouvelle fonction transcendante. München: Lindauer.” - [Soldner’s Constant, Wolfram MathWorld](https://mathworld.wolfram.com/SoldnersConstant.html) — Ramanujan calculated “μ = 1.45136380...,” while the correct value is 1.45136923488…. ### 17.6 {#r-17.6} The reciprocal gamma function 1/Γ(z) is entire and has zeros at the simple poles of Γ, including (1/Γ)(0)=0. *Status: confirmed* - [NIST DLMF §5.2, Gamma Function](https://dlmf.nist.gov/5.2) — records that reciprocal gamma is entire with simple zeros at the nonpositive integers. ### 17.7 {#r-17.7} For the unnormalized convention sinc(x)=sin(x)/x, sinc(0)=1 by continuation and sinc(π)=0; sinc is the interpolation kernel in band-limited sampling. sinc(0) is defined as the limiting value 1. *Status: confirmed* - [NIST DLMF §4.14, Trigonometric Functions](https://dlmf.nist.gov/4.14) — supplies the sine zeros used in the elementary evaluation: “In ℂ the zeros of sin z are z = kπ, k ∈ ℤ” (the limiting value sinc(0) = 1 is on the Sinc function line). - [MIT OpenCourseWare, The Sampling Process](https://ocw.mit.edu/courses/2-693-principles-of-oceanographic-instrument-systems-sensors-and-measurements-13-998-spring-2004/91fd1c49b05bf805b0d7057cf45cf230_part2_smplingpro.pdf) — states that sinc is the ideal interpolation function in the sampling theorem. - [Sinc function, Wikipedia](https://en.wikipedia.org/wiki/Sinc_function) — read as raw wikitext: “In either case, the value at x = 0 is defined to be the limiting value” sinc(0) = lim sin(ax)/(ax) = 1. ### 17.8 {#r-17.8} For integer n≥1, J_n(0)=0 while J_0(0)=1, as follows from the defining Bessel series. Bessel’s equation comes from separating Laplace’s or Helmholtz’s equation in cylindrical coordinates; J is the solution finite at the origin; the zeros of J₀ set the axisymmetric modes of a circular membrane. For integer n, Jₙ(x) is the coefficient of tⁿ in the Laurent expansion of e^((x/2)(t − 1/t)). *Status: confirmed* - [NIST DLMF §10.2, Bessel Functions](https://dlmf.nist.gov/10.2) — gives the defining series (10.2.2), from which J₀(0) = 1 and Jₙ(0) = 0 for n ≥ 1 follow; the page does not state those values itself. - [Bessel function, Wikipedia](https://en.wikipedia.org/wiki/Bessel_function) — read as raw wikitext: “Bessel’s equation arises when finding separable solutions to Laplace’s equation and the Helmholtz equation in cylindrical or spherical coordinates”; applications include “Electromagnetic waves in a cylindrical waveguide,” “Heat conduction in a cylindrical object,” “Modes of vibration of a thin circular or annular acoustic membrane (such as a drumhead …)”; “For integer or positive α, Bessel functions of the first kind are finite at the origin”; “For integer order α = n, J_{n} is often defined via a Laurent series for a generating function: e^{(x/2)(t − 1/t)} = Σ_{n=−∞}^{∞} J_{n}(x) t^{n}.” - [Vibrations of a circular membrane, Wikipedia](https://en.wikipedia.org/wiki/Vibrations_of_a_circular_membrane) — the axisymmetric modes are set by the positive roots of J₀; the fundamental u₀₁ has α₀₁ = 2.40483. ### 17.9 {#r-17.9} The beta function satisfies B(a, b)=Γ(a)Γ(b)/Γ(a+b), hence B(1/2, 1/2)=π. *Status: confirmed* - [NIST DLMF §5.12, Beta Function](https://dlmf.nist.gov/5.12) — gives the beta integral and gamma quotient. ### 17.10 {#r-17.10} The digamma function is ψ=Γ′/Γ and ψ(1)=−γ; current literature still does not establish whether the Euler–Mascheroni constant is irrational. ψ(z) = Γ′(z)/Γ(z); ψ(1) = −γ and ψ(n + 1) = Σ 1/k − γ. *Status: confirmed* - [NIST DLMF §5.15, Polygamma Functions](https://dlmf.nist.gov/5.15) — defines the polygamma functions ψ⁽ⁿ⁾, n = 1, 2, …, whose properties “follow straightforwardly by differentiation of properties of the psi function,” and gives their special values (digamma’s definition and values are on the §5.2 and §5.4 lines). - [Marek Wolf, 6+∞ New Expressions for the Euler–Mascheroni Constant (2025)](https://cmst.eu/wp-content/uploads/files/10.12921_cmst.2025.0000011_WOLF_acc.pdf) — the peer-reviewed paper explicitly states that irrationality of γ remains unknown. - [NIST DLMF §5.2, Gamma function: definitions](https://dlmf.nist.gov/5.2) — 5.2.2: ψ(z) = Γ′(z)/Γ(z). - [NIST DLMF §5.4, Gamma function: special values](https://dlmf.nist.gov/5.4) — 5.4.12: ψ(1) = −γ; 5.4.14: ψ(n+1) = Σ_{k=1}^{n} 1/k − γ. ### 17.11 {#r-17.11} Lambert W is the multivalued inverse of w↦we^w; W(1)=Ω≈0.5671433, and its real branches meet at W(−1/e)=−1. Ω = W(1) = 0.56714 32904… Corless, Gonnet, Hare, Jeffrey and Knuth proposed the standard notation for W’s branches in their 1996 paper. *Status: confirmed* - [NIST DLMF §4.13, Lambert W-Function](https://dlmf.nist.gov/4.13) — defines the inverse, its branches and the branch point W(−1/e) = −1 (the value Ω = W(1) is on the Omega constant line). - [Corless et al., On the Lambert W Function (1996)](https://cs.uwaterloo.ca/research/tr/1993/03/W.pdf) — the published paper’s author copy defines W and standardizes its branches and notation: “The text below is the same as that published in Advances in Computational Mathematics, Vol 5 (1996) 329 – 359”; authors R. M. Corless, G. H. Gonnet, D. E. G. Hare, D. J. Jeffrey and D. E. Knuth; “We use the letter W for this function, following early Maple usage.” and “An important part of this paper is our proposal of a standard notation for all the branches of W in the complex plane.” - [Omega constant, Wikipedia](https://en.wikipedia.org/wiki/Omega_constant) — read as raw wikitext: “It is the value of W(1), where W is Lambert’s W function”; Ω = 0.56714 32904 09783 87299 … ### 17.12 {#r-17.12} W(1) is transcendental: if algebraic nonzero Ω satisfied Ωe^Ω=1, Lindemann–Weierstrass would make e^Ω transcendental rather than algebraic 1/Ω. The same theorem makes π transcendental, and transcendence survives a nonzero rational scaling and a square root, so √π, π/2 and π itself are transcendental together. *Status: confirmed* - [Encyclopedia of Mathematics, Lindemann–Weierstrass theorem](https://encyclopediaofmath.org/wiki/Lindemann%E2%80%93Weierstrass_theorem) — states the required transcendence of e^α for nonzero algebraic α. - [NIST DLMF §4.13, Lambert W-Function](https://dlmf.nist.gov/4.13) — supplies Ωe^Ω=1. ### 17.13 {#r-17.13} The first positive zero j_{0,1}≈2.4048256 determines the fundamental axisymmetric circular-drum mode, and Siegel’s 1929 work implies every nonzero J_{0} zero is transcendental. *Status: confirmed* - [NIST DLMF §10.21, Zeros of Bessel Functions](https://dlmf.nist.gov/10.21) — defines the positive Bessel zeros j_{ν,m} and gives their properties and asymptotic expansions (numerical tables are referred elsewhere). - [MacTutor, Carl Ludwig Siegel](https://mathshistory.st-andrews.ac.uk/Biographies/Siegel/) — records Siegel’s 1929 result in these words: “for any non-zero algebraic integer r he showed that J0(r) is transcendental,” so no non-zero algebraic integer is a zero of J₀ (the statement for every nonzero root is on the Bessel function line). - [Vibrations of a circular membrane, Wikipedia](https://en.wikipedia.org/wiki/Vibrations_of_a_circular_membrane) — (§ The axisymmetric case, § Animations of several vibration modes): the axisymmetric modes are set by the positive roots of J₀, and the fundamental mode u₀₁ has α₀₁ = 2.40483. - [Wikipedia, Bessel function](https://en.wikipedia.org/wiki/Bessel_function) — its section on the zeros states that “Siegel proved in 1929 that when ν is rational, all nonzero roots of Jν(x) and J′ν(x) are transcendental,” citing Lorch and Muldoon, Int. J. Math. Math. Sci. 18 (1995) 551–560 (doi:10.1155/S0161171295000706). - [D. S. Karachalios, I. V. Gosea, Q. Zhang, A. C. Antoulas, Case study: Approximations of the Bessel function, arXiv:1801.03390](https://arxiv.org/abs/1801.03390) — read from the PDF: “The first 6 solutions of the equation J0(s) = 0, with 15 digits precision are: … 2.40482555769577.” ### 17.14 {#r-17.14} erf(1)≈0.8427008 equals the standard-normal probability within √2 standard deviations under erf(x)=2/√π∫_0^x e^(−t²)dt. erf(a/(σ√2)) is the probability of an error between −a and a; at a = σ√2 this is erf(1) = 0.842700793… *Status: confirmed* - [NIST DLMF §7.2, Error Functions](https://dlmf.nist.gov/7.2) — defines erf and gives its normalization. - [NIST/SEMATECH e-Handbook, Normal Distribution](https://www.itl.nist.gov/div898/handbook/eda/section3/eda3661.htm) — gives the normal density and cumulative-probability standardization used for the probability interpretation. - [Error function, Wikipedia](https://en.wikipedia.org/wiki/Error_function) — read as raw wikitext: “The defining integral cannot be evaluated in closed form in terms of elementary functions”; erf(x) = 2Φ(x√2) − 1; for a normal distribution with standard deviation σ and mean zero, erf(a/(σ√2)) “is the probability that the error of a single measurement lies between −a and a”; table: erf 1 = 0.842700793. ### 17.15 {#r-17.15} Ai(0)=1/(3^(2/3)Γ(2/3))≈0.355028; Airy functions solve the Airy equation and occur in linear-potential quantum problems. For x > 0 Ai decays exponentially and Bi grows; for x < 0 both oscillate. *Status: confirmed* - [NIST DLMF §9.2, Airy Functions](https://dlmf.nist.gov/9.2) — gives the Airy equation and Ai(0) special value. - [NIST DLMF §9.16, Physical Applications](https://dlmf.nist.gov/9.16) — documents Airy-function applications including quantum-mechanical turning-point problems; Ai “first appears as an integral in two articles by G.B. Airy on the intensity of light in the neighborhood of a caustic.” - [Airy function, Wikipedia](https://en.wikipedia.org/wiki/Airy_function) — read as raw wikitext: “When x is positive, Ai(x) is positive, convex, and decreasing exponentially to zero, while Bi(x) is positive, convex, and increasing exponentially. When x is negative, Ai(x) and Bi(x) oscillate around zero”; “the Airy equation is the simplest second-order linear differential equation with a turning point (a point where the character of the solutions changes from oscillatory to exponential)”; “The Airy function underlies the form of the intensity near an optical directional caustic, such as that of the rainbow.” ### 17.16 {#r-17.16} K(k)=∫_0^(π/2)dθ/√(1−k²sin²θ), so K(0)=π/2; K gives the finite-amplitude pendulum period, whereas ellipse arc length uses the second-kind integral E. *Status: confirmed* - [NIST DLMF §19.2, Legendre’s Integrals](https://dlmf.nist.gov/19.2) — defines K and E. - [NIST DLMF §19.6, Special Cases](https://dlmf.nist.gov/19.6) — gives K(0)=π/2. - [NIST DLMF §22.19, Physical Applications](https://dlmf.nist.gov/22.19) — gives the pendulum period as 4K(sin(α/2)). - [NIST DLMF §19.30, Lengths of Plane Curves](https://dlmf.nist.gov/19.30) — gives ellipse arc length using E, resolving the inherited K/E mix-up. ### 17.17 {#r-17.17} The Fresnel integral S(x)=∫_0^x sin(πt²/2)dt tends to 1/2, and S together with C parametrizes the Cornu spiral used in diffraction. Cornu showed in 1874 that diffraction intensity can be read off a graph of the spiral; the same curve is Euler’s of 1744. *Status: confirmed* - [NIST DLMF §7.2, Fresnel Integrals](https://dlmf.nist.gov/7.2) — defines S and C, and 7.2.9 gives lim_{x→∞} C(x) = lim_{x→∞} S(x) = ½. - [NIST DLMF §7.12(ii), Fresnel Asymptotics](https://dlmf.nist.gov/7.12.ii) — gives the large-z asymptotic expansions of the auxiliary functions f and g, referring to 7.5.3–7.5.4 for C and S (the limits at infinity are on the §7.2 line). - [NIST DLMF §7.20, Mathematical Applications](https://dlmf.nist.gov/7.20) — identifies the parametric Cornu spiral. - [NIST DLMF §7.21, Physical Applications](https://dlmf.nist.gov/7.21) — connects Fresnel integrals and the Cornu spiral to diffraction. - [Euler spiral, Wikipedia](https://en.wikipedia.org/wiki/Euler_spiral) — read as raw wikitext: named for “elastic springs (“Euler spiral,” 1744), graphical computations in light diffraction (“Cornu spiral,” 1874)”; “In 1874, Alfred Marie Cornu showed that diffraction intensity could be read off a graph of the spiral”; the diagram of a Cornu spiral is used for plane-wave diffraction from a knife-edge. - [R. D. Blandford and K. S. Thorne, Applications of Classical Physics, Ch. 8 “Diffraction,” §8.4.3 (Caltech Ph136, 2012)](http://www.pmaweb.caltech.edu/Courses/ph136/yr2012/1208.1.K.pdf) — for a straight edge, “Long before the occultation, we can approximate σmin = −∞ and σmax = +∞ …, so ∆Eσ is given by the arrow from (−1/2, −1/2) to (1/2, 1/2)”; and “The flux |ψ|2 ∝ |∆Eσ|2 is proportional to the squared length of the arrow on the Cornu spiral.” ### 17.18 {#r-17.18} The Gauss hypergeometric series is Σ(a)_n(b)_nz^n/((c)_nn!); parameter choices produce logarithmic and inverse-trigonometric special cases and negative integer parameters give polynomial cases. *Status: confirmed* - [NIST DLMF §15.2, Gauss Series](https://dlmf.nist.gov/15.2) — gives the defining series, analytic continuation and polynomial cases. - [NIST DLMF §15.4, Special Cases](https://dlmf.nist.gov/15.4) — gives logarithm, arctangent and arcsine parameter choices. ### 17.19 {#r-17.19} Li_s(z)=Σz^n/n^s specializes to −ln(1−z) at s=1 and ζ(s) at z=1 where convergent; Li_2(1/2)=π²/12−(ln2)²/2. Li₁(z) = −ln(1 − z). *Status: confirmed* - [NIST DLMF §25.12, Polylogarithms](https://dlmf.nist.gov/25.12) — gives the definition and the special case “ζ(s) = Li_{s}(1)”; its reflection formula 25.12.6, Li₂(x) + Li₂(1 − x) = π²/6 − (ln x) ln(1 − x), gives Li₂(½) = π²/12 − (ln 2)²/2 at x = ½ (Li₁ is on the Polylogarithm line). - [Polylogarithm, Wikipedia](https://en.wikipedia.org/wiki/Polylogarithm) — read as raw wikitext: “Li₁(z) = −ln(1−z)” and “Li₁(½) = ln 2.” ### 17.20 {#r-17.20} For the standard Jacobi convention, θ₃(0;i)=Γ(1/4)/(√2π^(3/4))=π^(1/4)/Γ(3/4); theta functions connect to elliptic functions and the zeta functional equation. In the convention ϑ(z; τ) = Σ exp(πin²τ + 2πinz), the relation ϑ(0; −1/τ) = (−iτ)^½ ϑ(0; τ) is what Riemann used for the zeta functional equation. *Status: confirmed* - [MathWorld, Jacobi Theta Functions](https://mathworld.wolfram.com/JacobiThetaFunctions.html) — explicitly gives θ₃(e^−π)=π^(1/4)/Γ(3/4); “The Jacobi theta functions are the elliptic analogs of the exponential function, and may be used to express the Jacobi elliptic functions.” - [NIST DLMF §5.5, Functional Relations](https://dlmf.nist.gov/5.5) — Euler’s reflection formula makes the two displayed gamma forms identical. - [NIST DLMF §20.9, Relations to Other Functions](https://dlmf.nist.gov/20.9) — connects theta functions to elliptic functions and the Riemann zeta function. - [Theta function, Wikipedia](https://en.wikipedia.org/wiki/Theta_function) — read as raw wikitext: “ϑ(z; τ) = Σ_{n=−∞}^{∞} exp(πin²τ + 2πinz)”; “The relation ϑ(0; −1/τ) = (−iτ)^½ ϑ(0; τ) was used by Riemann to prove the functional equation for the Riemann zeta function, by means of the Mellin transform.” ### 17.21 {#r-17.21} Legendre, Hermite and Laguerre are classical orthogonal-polynomial families; Hermite occurs in oscillator eigenfunctions, generalized Laguerre in hydrogenic radial eigenfunctions, and Legendre in spherical harmonics. Legendre polynomials are orthogonal under weight 1 on [−1, 1], Laguerre under e^(−x)x^α, Hermite under e^(−x²). *Status: confirmed* - [NIST DLMF §18.3, Orthogonal Polynomials: Definitions](https://dlmf.nist.gov/18.3) — defines the classical families and their weights (Table 18.3.1: the orthogonality intervals and weight functions) (the link formerly pointed to the chapter’s table of contents). - [NIST DLMF §18.39, Applications in the Physical Sciences](https://dlmf.nist.gov/18.39) — gives the oscillator and hydrogen-atom eigenfunction applications. - [NIST DLMF §14.30, Spherical and Spheroidal Harmonics](https://dlmf.nist.gov/14.30) — relates spherical harmonics to the Ferrers (associated Legendre) functions 𝖯_{l}^{m}(cos θ); it does not name the Legendre polynomials themselves. - [NIST DLMF §18.3, Classical orthogonal polynomials: definitions](https://dlmf.nist.gov/18.3) — “Table 18.3.1 provides the traditional definitions of Jacobi, Laguerre, and Hermite polynomials via orthogonality and standardization”; its weights: Legendre w(x) = 1, Laguerre w(x) = e^(−x)x^α, Hermite w(x) = e^(−x²). ### 17.22 {#r-17.22} The upper and lower incomplete gamma functions sum to Γ(s); γ(1/2, x²)=√π erf(x), and regularized incomplete gamma functions supply chi-square distribution functions. *Status: confirmed* - [NIST DLMF §8.2, Incomplete Gamma Functions](https://dlmf.nist.gov/8.2) — defines upper and lower functions and their sum. - [NIST DLMF §8.4, Special Values](https://dlmf.nist.gov/8.4) — gives the error-function identity. - [NIST DLMF §8.23, Statistical Applications](https://dlmf.nist.gov/8.23) — identifies chi-square distributions as particular regularized incomplete-gamma forms. ### 17.23 {#r-17.23} The arithmetic classifications of μ, γ and erf(1) are treated conservatively: explicit recent sources leave μ and γ irrationality unresolved; the cited 2023 analysis proves ∫_0^1e^(−x²)dx transcendental but does not settle its ratio 2/√π=erf(1). *Status: judgement* - [Wikipedia, List of mathematical constants](https://en.wikipedia.org/wiki/List_of_mathematical_constants) — the Soldner constant μ’s row marks its rationality, algebraicity and period status all as unknown (“?”). - [Wolf, 6+∞ New Expressions for the Euler–Mascheroni Constant (2025)](https://cmst.eu/wp-content/uploads/files/10.12921_cmst.2025.0000011_WOLF_acc.pdf) — explicitly states that γ irrationality is unknown. - [MathOverflow, arithmetic of ∫₀¹e^(−x²)dx (2023)](https://mathoverflow.net/questions/461028/is-the-integral-of-e-x2-from-0-to-1-known-to-be-irrational) — supplies a detailed Siegel–Shidlovsky proof for the unnormalized integral; it does not classify erf(1), so the page makes only a documented literature-survey judgement for that ratio. ### 17.24 {#r-17.24} Chudnovsky’s algebraic-independence work in the 1970s makes π algebraically independent of Γ(1/4), and likewise of Γ(1/3); this proves the gamma values and the displayed Ai(0) combination transcendental. The result is Chudnovsky’s 1976 Doklady paper (Waldschmidt, Philibert); the 1975 that Wikipedia’s Lemniscate constant page gives is the date of his abstract in the Notices of the AMS, which announced the result (two events, not a dispute); the 1984 date is the later AMS monograph that collected it. *Status: judgement* - [G. V. Chudnovsky, Algebraic independence of the values of elliptic function at algebraic points](https://doi.org/10.1007/BF01390068) — the primary author’s 1980 paper cites the relevant 1976 Doklady paper by title, issue and pages. - [Michel Waldschmidt, Transcendental Number Theory: the State of the Art (KIAS, Seoul, February 2004)](https://webusers.imj-prg.fr/~michel.waldschmidt/articles/pdf/KIAS-02-2004.pdf) — (under the heading “G.V. Chudnovskij (1976)”): “The two numbers Γ(1/4) and π are algebraically independent and the two numbers Γ(1/3) and π are algebraically independent.” - [Wikipedia, Lemniscate constant (revision of 13 September 2026)](https://en.wikipedia.org/w/index.php?title=Lemniscate_constant&oldid=1374703971) — a permanent link to the revision read. Read raw on 24 September 2026: “In 1975, Gregory Chudnovsky proved that π and ϖ are algebraically independent over Q,” citing his abstract in the Notices of the AMS 22 (1975), p. A-486; the book dates the result by the 1976 Doklady paper, which that abstract announced. History: a 23 September 2026 reading reported this sentence gone; the page has had no edit since 13 September, so that reading was mistaken and the sentence was there throughout. - [Georges Philibert, Une mesure d’indépendance algébrique, Annales de l’Institut Fourier 38 (1988), no. 3, 85–103](https://doi.org/10.5802/aif.1142) — cited for its reference list, which distinguishes Chudnovsky’s 1976 Doklady paper (entry [2]) from the 1984 AMS monograph (entry [4]). ### 17.25 {#r-17.25} The epigraph is quoted from the 1999 Cambridge University Press preface by George Andrews, Richard Askey and Ranjan Roy, where the authors attribute the remark to Paul Turán. *Status: confirmed* - [Cambridge University Press, Special Functions — Preface](https://www.cambridge.org/core/books/abs/special-functions/preface/7A3FFF3EE7FEB76075E47BA4C4A44152) — prints the sentence, author list and attribution to Turán; print publication year 1999. ### 17.26 {#r-17.26} The gamma function is the unique log-convex extension of the factorial: Bohr and Mollerup, 1922. Γ is not the only analytic extension of the factorial: any analytic function vanishing on the positive integers, such as k sin(mπx), can be added. Substituting t = u² in Euler’s integral turns Γ(½) into the Gaussian integral, 2∫₀^∞ e^(−u²) du = √π. *Status: confirmed* - [Bohr–Mollerup theorem, Wikipedia](https://en.wikipedia.org/wiki/Bohr%E2%80%93Mollerup_theorem) — proved in 1922 by Harald Bohr and Johannes Mollerup; “Γ(x) is the only function that satisfies f(x + 1) = x f(x) with log(f(x)) convex and also with f(1) = 1,” for positive f on x > 0. - [Gamma function, Wikipedia](https://en.wikipedia.org/wiki/Gamma_function) — read as raw wikitext: “it is not the only analytic function that extends the factorial, as one may add any analytic function that is zero on the positive integers, such as k sin(mπx) for an integer m.”; “Perhaps the best-known value of the gamma function at a non-integer argument is Γ(1/2) = √π, which can be found … simply by making the substitution t = u² in the integral definition of the gamma function, resulting in a Gaussian integral”; with t = cx², “if z = 1/2, c = 1 we get Γ(1/2) = 2∫₀^∞ e^(−t²) dt = √π” (formulas as typeset on the page). ### 17.27 {#r-17.27} The gamma reflection formula is Γ(z)Γ(1−z) = π/sin(πz) for z ≠ 0, ±1, …; at z = 1/3 it gives Γ(1/3)Γ(2/3) = 2π/√3, the identity the chapter’s Ai(0) transcendence argument uses beside Chudnovsky’s theorem. *Status: confirmed* - [NIST DLMF §5.5(ii), Functional Relations: Reflection](https://dlmf.nist.gov/5.5) — equation 5.5.3 states Γ(z)Γ(1−z) = π/sin(πz) for z ≠ 0, ±1, …; §5.5(i) carries the recurrence and §5.5(iv) the Bohr–Mollerup theorem. Editorial note: Own arithmetic checked: sin(π/3) = √3/2, so the reflection formula at z = 1/3 gives Γ(1/3)Γ(2/3) = 2π/√3 ≈ 3.6275987, which matches the direct product of the two gamma values to full double precision. Added in the V14 prose pass because Chapter ᚐ(A)’s paragraph 7 invokes “the reflection formula” by name, and no reference already in the chapter states it: r-17.3 cites DLMF §5.5 but only for the recurrence Γ(z+1) = zΓ(z). Card ai0 invokes the same identity under 17.24 and should cite this entry too. ### 17.28 {#r-17.28} The defining equations of the four families the chapter names and had no family card for: Bessel’s equation z²w″ + zw′ + (z² − ν²)w = 0 with Jν its solution of the first kind; the Airy equation w″ = zw with Ai and Bi its standard solutions; erf z = (2/√π)∫₀ᶻ e^(−t²) dt with erfc z = 1 − erf z; and the Fresnel integrals C(z) = ∫₀ᶻ cos(πt²/2) dt and S(z) = ∫₀ᶻ sin(πt²/2) dt. *Status: confirmed* - [NIST DLMF §10.2, Bessel Functions: Definitions](https://dlmf.nist.gov/10.2) — equation 10.2.1 gives Bessel’s differential equation z²(d²w/dz²) + z(dw/dz) + (z² − ν²)w = 0, and 10.2.2 gives the ascending series defining Jν(z), the Bessel function of the first kind. - [NIST DLMF §9.2, Airy Functions: Differential Equation](https://dlmf.nist.gov/9.2) — equation 9.2.1 gives the Airy equation d²w/dz² = zw, names Ai(z) and Bi(z) as its standard solutions, records that all solutions are entire, and gives Ai(0) = 1/(3^(2/3)Γ(2/3)) = 0.35502 80538…. - [NIST DLMF §7.2, Error Functions and Fresnel Integrals: Definitions](https://dlmf.nist.gov/7.2) — 7.2.1 gives erf z = (2/√π)∫₀ᶻ e^(−t²) dt, 7.2.2 gives erfc z = (2/√π)∫ᶻ^∞ e^(−t²) dt = 1 − erf z, and 7.2.7 and 7.2.8 give the Fresnel integrals C(z) = ∫₀ᶻ cos(πt²/2) dt and S(z) = ∫₀ᶻ sin(πt²/2) dt. Editorial note: Added for Q17.4, answered by the default. The glossary carried family cards for hypergeometric, polylog, theta, orthogonal polynomials and incomplete gamma — none of which the prose mentions — and none for Bessel, Airy, error or Fresnel, which it does. The four value cards (j01, j1zero, ai0, erf1, fresnelS) each cite an entry for their own value, and no entry carried the defining equations the family cards need. This one does. ### 17.29 {#r-17.29} The Cauchy principal value, named after Augustin-Louis Cauchy, assigns a value to an improper integral that is otherwise undefined by excluding a symmetric neighbourhood of the singularity and taking the limit as that neighbourhood shrinks; the two divergent halves are made to cancel by being cut away at the same rate. It is also the way the distribution p.v.(1/x) is defined. Approached asymmetrically, the integral can be made to equal any real number; li(x) for x > 1 is a Cauchy principal value. *Status: confirmed* - [Cauchy principal value, Wikipedia](https://en.wikipedia.org/wiki/Cauchy_principal_value) — read as raw HTML: “the Cauchy principal value, named after Augustin-Louis Cauchy, is a method for assigning values to certain improper integrals which would otherwise be undefined. In this method, a singularity on an integral interval is avoided by limiting the integral interval to the non-singular domain,” with the symmetric-limit formulation and the distribution p.v.(1/x). - [Cauchy Principal Value, MathWorld](https://mathworld.wolfram.com/CauchyPrincipalValue.html) — the symmetric-exclusion limit written out. - [Alan Reynolds, Introduction to Complex Analysis (PHYS30041 course notes): Cauchy principal value](https://alan-reynolds.quarto.pub/introduction-to-complex-analysis-phys30041/cpv.html) — “By approaching the singularity asymmetrically in other ways, we can make the value of the integral equal to any real number we choose!” - [Logarithmic integral function, Wikipedia](https://en.wikipedia.org/wiki/Logarithmic_integral_function) — read as raw wikitext: “The function 1/(ln t) has a singularity at t = 1, and the integral for x > 1 is interpreted as a Cauchy principal value”; “the prime number theorem states that: π(x) ∼ li(x).” ### 17.30 {#r-17.30} Cites the definition of the Lambert W function as the multivalued inverse of w↦we^w and its two real branches. W solves equations such as x^x = z and is used in combinatorics and in delay differential equations; Lambert studied the related equation in 1758, Euler a special case in 1783, and Corless, Gonnet, Hare, Jeffrey and Knuth gave the modern treatment in 1996. *Status: confirmed* - [Lambert W function, Wikipedia](https://en.wikipedia.org/wiki/Lambert_W_function) — defines W as the inverse of w e^w, its branches W0 and W-1, and the branch point at (-1/e, -1); “It is useful in combinatorics, for instance, in the enumeration of trees” and “also occurs in the solution of delay differential equations, such as y′(t) = a y(t − 1).” - [Lambert W functions, B. Ćurgus, Western Washington University](https://faculty.curgus.wwu.edu/Courses/Math_pages/Math_138/Lambert_W_function.html) — carries x^x solved by W: “Solve x^x = y for x > 0 and appropriate y > 0.” with solution “x = exp(W0(ln y)) = ln y / W0(ln y)” - [Corless, Knuth and Lambert W, Nick Higham](https://nhigham.com/2016/08/23/corless-knuth-and-lambert-w/) — carries the 1996 date of the canonical reference: “Robert M. Corless, Gaston N. Gonnet, D. E. G. Hare and David J. Jeffrey and Donald Knuth, On the Lambert W Function, Adv. in Comput. Math. 5, 329-359, 1996” ### 17.31 {#r-17.31} Cites the beta function’s definition as an integral and its identity B(a, b)=Γ(a)Γ(b)/Γ(a+b). Euler introduced the integral; the function normalizes the beta distribution, appears in the Veneziano amplitude, and has B(½, ½) = π. *Status: confirmed* - [Beta function, Wikipedia](https://en.wikipedia.org/wiki/Beta_function) — gives Euler’s integral definition and the identity relating B to the gamma function; “it was the first known scattering amplitude in string theory, first conjectured by Gabriele Veneziano.” - [Introduction to the Gamma Function, P. Sebah and X. Gourdon](https://scipp-legacy.pbsci.ucsc.edu/~haber/archives/physics116A10/gamma.pdf) — carries Euler’s integral and B(½, ½): “Euler gave (30) in 1730” and “8.1 Special values B(1/2, 1/2) = π” - [Beta distribution, Wikipedia](https://en.wikipedia.org/wiki/Beta_distribution) — carries the normalization: “The beta function, B, is a normalization constant to ensure that the total probability is 1.” ### 17.33 {#r-17.33} Cites the digamma function’s definition as the logarithmic derivative of the gamma function and its special values. It obeys ψ(x+1) = ψ(x) + 1/x, with ψ(1) = −γ and ψ(n+1) = Hₙ − γ, and has the asymptotic expansion ψ(z) ∼ ln z − 1/(2z) − …. *Status: confirmed* - [Digamma function, Wikipedia](https://en.wikipedia.org/wiki/Digamma_function) — defines ψ = Γ’/Γ, gives ψ(1) = -γ and the harmonic-number recurrence. - [§5.11 Asymptotic Expansions, NIST Digital Library of Mathematical Functions](https://dlmf.nist.gov/5.11) — carries the two expansions side by side: 5.11.1 “Ln Γ(z) ∼ (z−1/2) ln z − z + 1/2 ln(2π) + Σ B2k/(2k(2k−1)z^(2k−1))” and 5.11.2 “ψ(z) ∼ ln z − 1/(2z) − Σ B2k/(2k z^(2k))” ### 17.35 {#r-17.35} Describes the Cornu spiral as the curve traced by plotting the Fresnel integrals against each other, with curvature proportional to arc length, and its use in diffraction analysis. Euler studied its curvature property in 1744 and Alfred Cornu used it for diffraction; with C and S the integrals of cos(πt²/2) and sin(πt²/2) it winds toward (½, ½) and (−½, −½). Also called a clothoid, it is used for transition curves in railway and highway engineering. *Status: confirmed* - [Euler spiral, Wikipedia](https://en.wikipedia.org/wiki/Euler_spiral) — describes the Cornu spiral / Euler spiral, its curvature-proportional-to-arc-length property, and Cornu’s diffraction application; “This curve is also referred to as a clothoid or Cornu spiral”; “They are also widely used in railway and highway engineering to design transition curves between straight and curved sections of railways or roads,” so that “the centripetal acceleration increases smoothly with the traveled distance” instead of causing “lateral jerk.” - [The Cornu Spiral, HyperPhysics, Georgia State University](http://hyperphysics.gsu.edu/hbase/phyopt/cornu.html) — carries the two limit points: “The eyes of the spirals at (.5,.5) and (-.5, -.5) represent image plane points at infinity.”