On 19 July 1956, people at four locations in Maine reported fast-moving daylight objects with bright flame-like trails. An Air Force investigation brought together accounts from Munsungan Lake, Stockholm, Debsconeag Dam and Great Pond. Its officers favoured a meteor or several meteors. Their report also preserves an important disagreement: the approving officer rejected the suggestion that the witnesses had all described the same size, shape and colour, while agreeing with the probable astronomical explanation.
What is in the surviving file
The twenty-page Project Blue Book file catalogued under Medway contains an early message, summary cards and a fuller ten-page Air Intelligence Information Report, 3G-UFOB-1-56. The latter is dated 1 August 1956 and was prepared by Major John P. Vinik of Flight 3-G, 4602d Air Intelligence Service Squadron. It names Vinik and Technical Sergeant Alonzo A. Small as investigators. Its six sections distinguish four witness accounts, supplementary enquiries and the officers’ assessments.
Medway was an address associated with one source. His actual observation was made while fishing at Debsconeag Dam, described in the report as about fifteen miles west of Millinocket. Another witness lived at Portage but observed from Munsungan Lake; the Great Pond source was associated with Aurora. The catalogue label should not move all four encounters to a single Medway street or imply that they were observed together.
The report’s principal sources are anonymised in the released copy. The descriptions below follow the official investigators’ summaries. They are not four signed civilian questionnaires independently available in this file, and remarks about companions do not supply additional complete interviews.
The Munsungan Lake report
Part One, on PDF page 4, concerns a sporting-camp operator and pilot travelling southward on Munsungan Lake in a boat. Another member of the party drew his attention to something crossing from west to east. He estimated the time as 21:05Z on 19 July and had only two or three seconds to follow it.
He compared its brightness to a distant car headlight and its apparent size to a baseball at arm’s length. The colours seemed to vary through faded red, orange and yellow, with a pale, wavering trail that dispersed more quickly than a jet’s. He explicitly said that he could not judge its solidity, construction material or exact shape. Those reservations are part of the testimony, alongside his stronger impression that it moved much faster than aircraft he knew.
After it disappeared behind a ridge, the party reportedly heard and felt a substantial explosion, compared with a cache of dynamite detonating. His initial thought had been a disabled jet, but the apparent speed led him to doubt that explanation. The file supplies a reported sound after disappearance, not a located crash, recovered meteorite or measured trajectory.
Stockholm: a cone and a long trail
Part Two, PDF page 5, describes two women walking south towards Stockholm village, about half a mile north of it. The interviewed source reported an object moving northwest to southeast in bright daylight around 21:00Z. She described a silver-coloured cone, a yellow flame much longer than the apparent object, and a short-lived pale trail.
She estimated a duration of about fifteen seconds and said that the object disintegrated without passing behind clouds or another obstruction. Her impression of a metallic cone is a witness description, rather than a material identification. The investigator considered her generally reliable but specifically cautioned about her distance estimates. The report therefore does not support treating her estimate of two miles as a surveyed range.
Debsconeag Dam: the longest estimate
Parts Three and its continuation, PDF pages 6–7, concern a papermill repairman and former local guide fishing with another man at Debsconeag Dam. They described an object approaching from the southeast and passing towards the northwest, eventually disappearing behind treetops. They estimated twenty to thirty seconds, longer than the other detailed accounts.
The source believed it was solid and low, estimating three or four hundred feet. He described dark metal colouring, a bright trailing flame that disappeared twice, and a small blue flame apparently coming from its upper part. The men heard no sound and thought that unusual for something moving with such apparent force. They reported no course change or progressive descent, although its straight movement seemed guided to them.
Those observations do not establish an engine, controlled flight or a measured low altitude. The early message and summary material give a much shorter duration and compressed description. The fuller interview’s twenty-to-thirty-second estimate should not be silently replaced by the three seconds printed on a card. Nor should every card detail be assigned indiscriminately to all four locations.
Great Pond: a bright head and uncertain length
Part Four, PDF pages 8–9, records a camp operator changing fishing positions on Great Pond. Around 21:00Z he noticed an object moving west to east, north of his boat. He stopped the motor and watched for at least six seconds. He described a sharply outlined bright head, a blurred blue-grey rear portion and a quickly dispersing trail.
He likened the head to a football and the whole appearance to a tube, estimating about ten feet in length. The report immediately qualifies this: part, or even all, of that apparent length might have been the trail of fire. There was no sound. The witness tried to telephone a Ground Observer Corps post, then went there by car when the call failed. A woman at another camp was also reported to have seen the object, but the file does not provide a corresponding full interview with her.
Checks made by the investigators
The supplementary section, PDF page 10, records a negative aircraft check with Presque Isle Flight Service. The weather information reported scattered clouds at 4,500 feet, twenty-mile visibility and no precipitation. The 765th Aircraft Control and Warning Squadron at Charleston reported no relevant radar pickup. This is a visual case with a negative radar enquiry, not a radar-confirmed object.
The investigators also consulted Dr Maynard Jordan at the University of Maine in Orono. The university observatory and planetarium had been closed during the sightings; Jordan was not supplying his own observation of the event. He explained that more than one meteor could be present in the same general area and that a large meteor could produce an explosion of the sort reported. His comments supported the feasibility of an astronomical explanation without locating an impact or identifying a recovered specimen.
A shared conclusion with an explicit qualification
The preparing officer’s comments on PDF page 13 favour meteors, citing the short durations, speed, bright colours and general appearance. Colonel John W. Meador, the approving commander, then corrects the reasoning: the four sources varied considerably in their size, shape and colour descriptions, so those characteristics could not be generalised in the proposed way. He nevertheless concurred that meteors in the same broad area at approximately the same time were the probable explanation.
The earlier index-card continuation, PDF pages 19–20, also records a provisional astronomical assessment while further information from astronomers was sought. Its use of the word meteorite does not establish that material reached the ground. The fuller report neither presents a recovered object nor calculates one common path tying all observations to a single body.
The outcome is possibly explained. A meteor explanation was supported and explicitly adopted as probable by the investigators; it was not demonstrated by a recovered sample or an independently measured track. Calling the case a confirmed missile would contradict their finding that the aircraft or missile suggestions lacked substantiation. Calling it definitively identified would erase their own qualification.
Location and source trail
Location note: the map uses a rounded US Census reference for Stockholm town, one of the actual reporting localities. It is an approximate area marker for that part of the investigation, not a plotted fireball position, a private residence, or a marker encompassing all four observation sites. This article does not establish present-day access to the fishing locations or private camps.
Read the original accounts on PDF pages 4–9, the supplementary checks and schematic on pages 10–11, the two officers’ comments on page 13, and the earlier message and card material on pages 16–20. The schematic represents the investigators’ summary of reported directions; it is not a modern reconstruction with a proven common trajectory. Minor clerical inconsistencies remain in the file, including references to Friday alongside the repeatedly recorded date 19 July 1956. No alternative incident date has been invented to resolve them.