Urbain Le Verrier entered the École Polytechnique in 1831 and initially pursued chemistry. By the late 1830s he had turned to celestial mechanics, publishing memoirs on the secular perturbations of planetary orbits and the stability of the solar system.
His first astronomical memoir, presented on 16 September 1839 and printed in the additions to the Connaissance des Temps, took up the secular variations of the orbital elements of the seven principal planets. Earlier treatments had kept only the terms in the first power of the eccentricities and inclinations; Le Verrier carried the calculation to the third order.
It is worth being precise about what that achieved, because the usual summary gets it backwards. He did not improve the convergence of the perturbation series — he tested it, and for half the solar system the test failed. The third-order terms introduce small divisors, and with the masses then adopted “quelques diviseurs sont assez petits pour rendre les séries divergentes”. His conclusion is a disclaimer rather than a result: “Il paraît donc impossible, par la méthode des approximations successives, de prononcer si… le système composé de Mercure, Vénus, la Terre et Mars jouira d’une stabilité indéfinie”, and he asks geometers to settle the question by rigorous integration instead. Stability was confirmed only for Jupiter, Saturn and Uranus.
That is a sharper piece of work than “extending Laplace” suggests. Knowing exactly where a method stops being trustworthy is the same instinct that later let him insist a residual in Uranus was real rather than an artefact of his own approximations.
Every planet pulls on every other planet. In a solar system with only one planet, Kepler’s laws give an exact elliptical orbit. Add a second planet and the orbits are no longer exactly elliptical — each planet perturbs the other.
The key insight of perturbation theory: if the perturbing planet is much less massive than the Sun, its effect is small and can be computed as a correction to the Keplerian orbit. The correction is expressed as a series expansion in the mass ratio \(m'/M_\odot\).
The gravitational potential of a perturbing planet can be expanded in a Fourier series involving the orbital elements of both planets. For circular, coplanar orbits this simplifies to:
\[ R = \frac{G m'}{a'} \sum_{j=-\infty}^{\infty} b_{1/2}^{(j)}(\alpha) \cos(j(\lambda - \lambda')) \]
where \(\alpha = a/a'\) is the ratio of semi-major axes, \(\lambda\) and \(\lambda'\) are the mean longitudes, and \(b_{1/2}^{(j)}(\alpha)\) are the Laplace coefficients — integrals that Le Verrier and his contemporaries computed by hand.
The Laplace coefficient is defined as:
\[ b_s^{(j)}(\alpha) = \frac{1}{\pi} \int_0^{2\pi} \frac{\cos(j\theta)}{(1 - 2\alpha\cos\theta + \alpha^2)^s} \, d\theta \]
Le Verrier evaluated these by expanding the integrand in power series and summing term by term — tedious but tractable with patience. We can compute them numerically in a line of code.
import numpy as np
from discoverneptune.perturbation import laplace_coefficient
# Laplace coefficients for α = a_Jupiter/a_Saturn ≈ 0.545
alpha_js = 5.2 / 9.5
for j in range(5):
b = laplace_coefficient(0.5, j, alpha_js)
print(f" b_{{1/2}}^{{({j})}}({alpha_js:.3f}) = {b:.6f}") b_{1/2}^{(0)}(0.547) = 2.181903
b_{1/2}^{(1)}(0.547) = 0.623691
b_{1/2}^{(2)}(0.547) = 0.259915
b_{1/2}^{(3)}(0.547) = 0.119477
b_{1/2}^{(4)}(0.547) = 0.057494The \(j=0\) term gives the secular (long-term average) perturbation. The \(j=1\) term drives the dominant periodic perturbation at the synodic frequency — the beat frequency between two planets’ orbital periods.
The package retains a schematic synodic expression for illustrating how mass, orbital spacing, and beat frequency enter a perturbation estimate. It is not a quantitatively validated first-order heliocentric longitude solution: the simple expression omits the indirect disturbing-function term and does not reproduce a controlled REBOUND benchmark in amplitude or phase.
For that reason this page does not overlay the schematic curve with an N-body trajectory or claim agreement. A future quantitative derivation must state its coordinate and phase conventions, include both direct and indirect terms, and pass preregistered amplitude, phase, and waveform-error thresholds before it is published as a validation.
It is tempting to say that by the early 1840s Le Verrier had produced tables for the planets, but he had not. The record is narrower: the secular-variation memoirs of 1839–41, a Theory of the Motion of Mercury presented to the Academy in 1843 — which his own obituarist notes was “subsequently handled more completely, and definitely finished in 1859” — and shorter papers on the mass of Venus, an inequality of Pallas, and the comet of 1843. The general planetary tables came decades later.
What he had built by 1845 was therefore not a body of tables but a habit of mind: perturbation theory carried further than anyone had bothered to carry it, and a practised eye for the small residual that will not go away. When François Arago pointed him toward the Uranus anomaly, that was the right preparation for the job.