United States.Haynesville · pulsating light and scorched bark

On the night of 30 December 1966, at about 8:30 p.m., Dr. Louie A. Galloway —31 years old, a professor of atomic physics— was driving north on U.S. Highway 79 through wooded country in northern Louisiana near Haynesville, accompanied by his 28-year-old wife and their two children, aged five and seven. They saw a reddish-orange glow among the trees and above them, below treetop level, pulsing regularly between dull red and bright orange at intervals of roughly two seconds. There was no sound. At one point the light turned white and intense, and for a few moments it exceeded, just noticeably, the illumination of the car's headlights; then it returned to its earlier pulsing.

Galloway reported the sighting to Barksdale Air Force Base. The report reached Project Blue Book early in 1967 and also the University of Colorado project, the committee directed by Edward Condon. The witness did his own photometry: he triangulated the bearings on an aerial photograph of the area and placed the source between 1,000 and 1,400 yards away; he judged the intense phase to be just noticeably brighter than his headlights ten meters ahead of the car; his headlamps totalled 175 watts; from that comparison he computed a power of about 500 megawatts, following an earlier estimate of some 800. He himself noted that the total power could be substantially less if the light was concentrated in a beam.

Investigators Saunders and Wadsworth went to the site. Ten weeks after the event they kept a night watch hoping for a reappearance and explored the presumed location on foot. They found, in their words, little but trees, underbrush and oil wells. A burned area registering slightly higher radioactivity than background turned out to be a burned-over oil slick beside a pumping station, and similar radiation anomalies appeared at other oil slicks. Nothing was found that suggested any relation to the unexplained light source. The Condon report (↗) published it as Case 10 —'Pulsating Light Seen from Car; No Explanation'— and the Air Force likewise filed it as unidentified. But the same report is explicit about the figure: without knowing the source's radiative efficiency compared with that of automobile headlamps, or whether it radiated in a beam or in all directions, 'the value of several hundred megawatts is highly dubious'.

The witness kept investigating on his own after the official team withdrew, and from that later work comes the physical evidence the case now rests on: a clearing some nine meters across whose perimeter trees had bark blackened toward the center, attributed to the object's light rather than to a heat source, and bark fragments that Galloway collected and that Vallée has preserved ever since, waiting for a method that does not cut into the wood.

In June 2025 the case was published in Progress in Aerospace Sciences (Elsevier), volume 156, article 101098, by Vallée with Luc Dini (CNES-GEIPAN / 3AF Sigma2), Geoffrey Mestchersky and Nicolas Niasse (3AF Sigma2), in a special issue on UAP. The work builds a numerical thermal-diffusion model of heat transfer in the bark, revises the energy range to 500-900 megawatts, reports that the surface could have gone from 20 °C to over 760 °C within seconds, and describes a gamma spectrometry showing natural radionuclides plus caesium-137 with visible differences against shielded control samples. Vallée returned to the case in his September 2026 interviews.

Here the corpus's two axes come apart, and that separation is what the case contributes. Tier A is justified by the documentation: few files have a professional physicist as primary witness, a report to an Air Force base within days, a Blue Book record, a ten-week field inspection by the Colorado project, publication in the Condon report as unexplained, and sixty years later a peer-reviewed article in an established aerospace journal. The probability of 50 records that the case remains formally unexplained in that same record. And the partition of explanations keeps most of the mass on the prosaic, because the strength of the documentation is not the strength of the anomaly.

Three things hold it down. The first is that the energy figure, which is the case's headline, comes from a naked-eye comparison with the witness's own headlights, and the Condon report called those several hundred megawatts highly dubious in the same pages where it declined to explain the sighting. A 2025 model that refines that number inherits its input: better thermal physics on a bark sample does not repair a photometric baseline that reads 'just noticeably brighter than my headlights'.

The second is that the terrain offers a prosaic candidate nobody tested. The official team reported that the area contained little but trees, underbrush and oil wells, and found a pumping station with a burned-over oil slick. A gas flare at a well —pulsing reddish-orange below treetop level, flaring briefly white, silent at a thousand yards— matches the reported phenomenology closely, and the file has to say so even though no published source proposes it. That the committee stood beside the oil field and did not consider it is a fact about the 1967 investigation, not evidence about the object.

The third is that the chain of custody of the physical evidence runs through no institution. Condon's team looked and found nothing. The clearing and the bark come from the witness's own later work, and the sample travelled from Galloway to Vallée as a private holding. The caesium-137 is the sharpest illustration: it is the signature of atmospheric nuclear testing, present worldwide in organic matter from the 1960s, and the article reports having ruled out fallout and contemporaneous plants, a claim that cannot be checked from the published abstract and that would be checkable only with the sample and the control series in hand.

What the case does earn is the precedent. It is the first UFO case to pass peer review in an established aerospace journal, and the method is the part that travels: a preserved physical sample, a thermal model, a stated energy range and a number one can argue with. That is a different object from a testimony, and the corpus should say so, noting at the same time that peer review certifies the analysis, not the provenance of what was analysed.