On 7 September 1820, people in Embrun reported unusual lights near the partly eclipsed Sun. The surviving printed account describes bright globes appearing along different paths, some apparently returning towards the solar disc. It is a striking historical observation, but its original wording differs substantially from the familiar story of a formation manoeuvring with military precision.
The central source is a letter by J.-H. Serres, the town’s sub-prefect, addressed to the president of the Académie des sciences and dated 5 October 1820. It appears in the 1825 volume of the Annales de chimie et de physique, within a discussion of luminous phenomena seen by day. An English translation followed in the Edinburgh New Philosophical Journal in 1826. Arago’s later collected works reproduce an abridged version. These are successive publications of the same underlying testimony, rather than three independent sightings.
A walk interrupted
Serres says he had already watched the solar eclipse and was crossing the town on his way to walk in the fields. In a public square, he noticed a substantial group of people of different ages looking towards the Sun. Still preoccupied with the eclipse, he initially continued without examining what had caught their attention.
A second group prompted him to stop. They were in a shaded street, in a position from which the Sun itself could not be seen. He therefore realised they must be looking at something else. In the fuller journal version, he identifies a veterinary practitioner, Thommé, as the person he questioned. Thommé described what the group regarded as stars detaching themselves from the Sun and invited Serres to look.
Serres then reported seeing globes of fire, with an apparent diameter comparable to that of the largest stars. He placed their apparent origin around the Sun’s upper half. This is a description of an observer’s visual impression and angular appearance. It does not measure the objects’ physical size or establish that anything was actually being expelled from the Sun.
What the paths looked like
The journal account describes short, unequal intervals between appearances. Sometimes several globes were visible at once, but they diverged from one another. Some seemed to follow straight paths and fade; others followed curved, approximately parabolic paths. A further category appeared to travel outwards and then retrace the same line towards the solar disc while still luminous.
The wording does not describe a set of equally spaced objects maintaining a formation. Nor does Serres give a measured speed. His impression of extremely rapid motion is not accompanied by timed tracks, a known distance or an instrument record. The report preserves an unusual appearance, with considerable variety in its trajectories, rather than the geometrical flight plan often attributed to it.
Serres also specifies his viewing position. A house blocked the Sun, while his line of sight passed the roof edge towards a point close to the solar limb. The eclipse was declining. He says the lights gradually became invisible as ordinary sunlight returned, and reports a bystander’s impression that more had been visible when the Sun was paler. That changing background illumination is part of the observation itself and should not disappear from a case summary.
Who else was involved?
The fuller account contains more named context than the later abridgement. Serres had noticed a law student, Cézanne, in the first group. After observing the phenomenon, he asked Cézanne and Thommé how they had first learned of it. Thommé said a woman had called him from his stable to see it; Cézanne said children had drawn attention to the display and attracted a gathering.
In a postscript, Serres adds that Fouré, an engineer with the Ponts et Chaussées, had told him he too had seen the phenomenon and would attest to it if asked. This is useful evidence of a reported wider observation, but the document reviewed here remains Serres’s letter. It does not include separate signed depositions from everyone in the crowds, and a repeated publication of the letter cannot supply those missing independent records.
Serres was not simply asserting that his interpretation must be accepted. His introduction allows the academy’s president to consider whether he had been deceived by an appearance. He says he had arranged for the phenomenon to be drawn so that astronomers and physicists could examine comparable observations during future eclipses. This article is based on the printed text reviewed, rather than a claimed examination of that drawing or of the original manuscript letter.
Arago’s scientific setting
The surrounding discussion matters. It begins with Hansteen’s report of a luminous point passing through a telescope’s field while he observed the Pole Star in daylight in August 1825. Thomas Dick proposed a distant bird reflecting sunlight. Arago accepts that reflected light from feathers could sometimes produce a similar appearance, while arguing against treating that explanation as universal.
Arago also describes his own observations of bright points while using a telescope to observe the Sun. He wonders about the occasions when airborne spider silk was abundant, and discusses whether some daytime phenomena might resemble shooting stars. He introduces Serres’s letter by asking readers to consider a possible analogy with these other observations.
This is exploratory comparison, not an identification of the Embrun lights. Hansteen’s telescope observation, Arago’s repeated experiences and Serres’s account happened in different circumstances. The discussion neither proves that Embrun involved birds or spider silk nor establishes an extraordinary craft. Its value is that it preserves both the observations and the uncertainties being considered by an early scientific reader.
How the later account changed
In Anatomy of a Phenomenon, Jacques Vallée presents an Embrun passage as Arago’s account of objects maintaining equal spacing and turning with military precision during a lunar eclipse. The original journal text reviewed here specifies a solar eclipse and divergent globes at unequal intervals. That discrepancy makes the earlier printed letter especially important when assessing the case.
There is a smaller discrepancy between historical publications too. The French account gives about 1:45 in the afternoon, while the 1826 English translation gives about 4:15. The observation date and solar-eclipse setting agree. No exact clock time is adopted here as a settled measurement, and the disagreement is not resolved by silently selecting the more convenient version.
Assessment and location
The classification is Insufficient evidence. Serres supplies an identifiable, dated account and reports multiple observers, but the reviewed material lacks independent measurements of distance, speed or physical dimensions, and does not establish a cause. A later inaccurate formation summary cannot strengthen the original evidence. Equally, a proposed natural analogy in Arago’s discussion is insufficient to declare the entire Embrun observation explained.
The UFO / UAP / USO grouping records the history of an unidentified aerial observation and its later treatment. It does not assert an extraterrestrial object. The sphere and multiple-light tags reflect the reported globes and simultaneous appearances. The sources do not justify adding a disc shape, radar detection, occupants or a landing.
The approximate map pin is 44.564, 6.496, a rounded municipal-centre reference for Embrun from the French government geographical API. It does not identify Serres’s house corner, either crowd’s exact position or a flight path. The account’s square and shaded street are not securely located by this review, so no particular building is presented as an authenticated encounter venue or visitor attraction.
The lead photograph is a modern view of Embrun from Saint-André-d’Embrun, taken in July 2020. It provides geographical context and shows neither the 1820 display nor an eclipse reconstruction. For further investigation, the 1825 journal passage is the best starting point among the sources reviewed here: it retains the named local contacts, the observation’s qualifications and the postscript that were shortened in the later collected version.
Sources and further reading
Benoît Prieur, 15July2020: exact town photograph, provenance and CC0 dedication
French government geographical API, Embrun05046: name, municipal-centre coordinates and region