The Historical Observations

The data that discovered a planet: Le Verrier’s reduction of Uranus observations.

Overview

The discovery of Neptune was not triggered by a single observation, but by the mathematical reconciliation of observations spanning 155 years. Urbain Le Verrier’s breakthrough relied on his decision to trust “ancient” observations (made before Uranus was recognized as a planet in 1781) that his predecessor, Alexis Bouvard, had largely dismissed as unreliable.

The Dataset

Le Verrier utilized two main types of data in his 1846 memoirs, summarized in the figure below and detailed in the sections that follow.

Figure 1: Le Verrier’s evidence: Uranus longitude residuals (observed − theory). The pre-discovery sightings (1690–1771) come from Le Verrier’s 1846 Table I and swing from +110″ to −71″; the post-1781 points are representative of the normal-place trend (see the data README for provenance), climbing to +128″ by 1845. No single Keplerian orbit can flatten both eras — but one unseen planet can.

1. Ancient Observations (1690–1771)

These were pre-discovery sightings where Uranus was recorded as a star. The most famous is John Flamsteed’s 1690 observation (cataloged as 34 Tauri). By re-reducing these observations using modern corrections for refraction and aberration, Le Verrier showed they were far more accurate than previously believed.

The residuals (\(O-C\)) for these observations showed a massive positive discrepancy in the late 17th and early 18th centuries, followed by a sharp negative turn by 1770.

Every sight line Le Verrier trusted, accumulating from Flamsteed’s 1690 sighting to the eve of discovery. Pre-discovery rays render in a cooler hue; the pulse marks Uranus’s 1781 recognition as a planet.

2. Modern Normal Places (1781–1845)

After 1781, observations became abundant. Le Verrier reduced these into “normal places”—weighted averages centered on specific years or oppositions—to smooth out observational noise. By 1845, the discrepancy between the best-fit Newtonian orbit (without Neptune) and these normal places reached over 120 arcseconds.

The same animation restricted to the post-discovery era: just seven points spanning 1781–1845 (representative of the normal-place record). Sparse on its own — which is why the ancient observations mattered.

Why it Matters

Without the ancient observations, the “modern” data from 1781–1845 could still be reasonably fit by an elliptical orbit with slightly modified elements. It was the inability to fit both the ancient and modern data that proved a trans-Uranian perturber was necessary.

Reference Data

The following table summarizes the key residuals Le Verrier analyzed (Table I of his 1846 memoir):

Year	Observer	Description	Residual (\(O-C\))
1690	Flamsteed	Earliest record	+67.1”
1715	Flamsteed	Pre-discovery	+110.0”
1756	Mayer	Star #964	-4.9”
1769	Lemonnier	6-observation series	-55.0”
1781	Herschel	Discovery	~0.0”
1845	Various	Final reduction	~128.0”

Data source: data/leverrier_historical_observations.csv

References

• Le Verrier, U. “Recherches sur les mouvements de la planète Herschel (dite Uranus).” Connaissance des Temps pour 1849 (1846).
• Grosser, M. The Discovery of Neptune. Harvard University Press (1962).
• Standage, T. The Neptune File. Walker & Company (2000).
• Kollerstrom, N. “Neptune’s Discovery: 150 Years of Controversy.” Astronomy Now (1996).