R e s e a r c h
In most respects it seems possible to explain this sighting satisfactorily - if not conclusively - as an unusual mirage. Some of the difficulties with the mirage theory - such as those discussed in the Condon Report by THAYER in 1970 - can probably be overcome, as THAYER today agrees . And thanks to the Australian Zanthus case (which occurred on the other side of the world even as the Condon Report was being prepared for publication) and a small number of others we can say that the BOAC phenomenon was after all not quite "so rare that it has never been reported before or since". We have seen that there are other cases, such as the United Airlines sighting of July 4, 1947, that may show at least some of the same signature features, indicating that unusual mirages from aircraft at moderate altitude may occur more often than has previously been assumed .
Despite some superficially attractive features the Sorta Sol or flocking starling theory is quantitatively and behaviourally unsupportable. The theory of a deflating stratosphere balloon caught in a jet stream is also a challenger, but highly improbable. A balloon does not explain the lateral merging of all the little objects, or the sudden dwindling to nothing of the big one (ex hypothesi 18 minutes after any positive buoyancy of the balloon had already leaked away).
In terms of atmospheric optics, however, we could envisage a selection effect analogous to viewing a lateral line through a mask which has two near-vertical converging slits. If the mask is moved up or down, the visible small segments of the line approach or recede from one another, but no real lateral displacement occurs. In the BOAC and smilar cases different parts of a mountain horizon or the tops of cloud towers might be selected sequentially by an undulating duct in this way (or by other changes in relative viewing height) appearing to be small objects that move laterally. We don't have a quantitative model of this process but at least we can understand such an effect qualitatively.
This YouTube video from New Mexico appears to show an undulating mirage duct that causes the miraged images to "move" both laterally and vertically. At times parts of the mountain top on the right seem to detach themselves to become part of the refracted image above. At 23, and again at 44 seconds in the video, a much more unusual phenomenon occurs: images that form in between the two major tops slide sideways to the right to become part of the bigger "blob" (actually this movement is created by a rapid succession of refracted images of ever higher parts of the mountain slope below). [Our thanks to researcher Viktor GOLUBIK for calling our attention to this video.]
Whilst balloon and mirage theories both require some supposition, the balloon theory suffers from the simple fact that there is no natural physical relation between the independent balloon and aircraft motions, necessitating some superadded and quite difficult explanation of how a balloon maintained a constant visual bearing for 18 minutes. In contrast there is a natural geometrical-optical relation between the eye and a remote miraged feature with small visual parallax. We do need to hypothesise the existence of an extensive elevated mirage duct and distant cloud (probably towering cumulus) punching through the duct; but these phenomena are consistent with known or likely weather and geography. The balloon theory requires a less economical set of requirements - statistically very unlikely (although not impossible) jet wind speed, a fortuitous identity of wind and aircraft vectors to explain a constant sighting bearing, a fortuitous onset of neutral buoyancy, and (more importantly) a windspeed shear gradient that is physically very difficult to defend in order for the plane to get close enough to the balloon to even see it.
Another question for the balloon theory is its likely origin. Wim VAN UTRECHT searched archival information for stratospheric balloon launch sites in the area, finding none . But he points out that the major USAF balloon programme Project Genetrix - nicknamed Moby Dick - was in full swing during this time. The officially stated purpose of the Genetrix balloons, some over 80 ft (25 m) across when fully inflated at altitude, was to map high-altitude wind currents , which would obviously include the jet streams. Some US-launched balloons in the early phase of the programme are known to have reached Europe, coming down in Spain and Scotland . By 1954 Genetrix was getting into its stride.
A declassified USAF document dated January 1956  states that Moby Dick launches had been ongoing from the US for two years previously and were now being expanded to other launch sites around the world including Pacific and Far East where launches had also been conducted during 1954-55. They probably 'lost' quite a few balloons during what was a huge project or series of sub-projects spanning some years. We note also the explicit statement in the note  reference that some of the balloons were shaped like an "inverted pear" (exactly the same words were used by Capt. HOWARD to describe the unknown object when it was first spotted).
But such a balloon, hypothetically having a number of radar reflectors or instrument packages attached to explain the "satellite" objects, would not very well explain either the sudden disappearance, the dwindling in line of sight that is so characteristic of mirage, the absence of ground and air radar contact (see below), or the failure of the fighter pilot to see anything in the area. Additionally, Wim VAN UTRECHT concedes that none of the witnesses reported seeing sunlight reflecting off of the objects' surfaces, although these balloons are usually made of a shiny transparent polyethylene and radar-tracking reflectors are wrapped in reflective tinfoil.
Indeed, polyethylene balloons scatter "as much as 20-30% of incident light" both through and off their fabric (LALLY, 1969) and 30% even of sunset light is a lot of light. They are typically described as bright or silvery, especially at twilight, and the flapping, twisting fabric of a deflated balloon would offer changing reflection angles to the scattered sky-brightness (even when the balloon itself is in the Earth's shadow after the Sun completely sets) so some detail should have been seen. It seems highly unlikely that a nearby Genetrix balloon could have provided the necessary optical contrast in this case, whereas the silhouette of a distant opaque cloud dimmed by lossy transmission through a duct might well have done so.
Having said this, a few awkward issues remain imperfectly resolved in the BOAC Centaurus sighting. For example, the implied precision of bearing-constancy from the moving aircraft still puts a strain on any realistic target object - including clouds - as the source of a mirage image.
To keep parallax drift below a possibly-acceptable 10° for a substantially stationary target the distance from the plane needs to be some 450 miles. This tends to locate the responsible clouds beyond the more mountainous part of the interior of Quebec and nearer the lowland region along the coast of Hudson's Bay, which is a problem for clouds depending on orographic lift such as mountain wave clouds (standing-wave lenticularis) or mountain triggered cumulonimbus. On the other hand, towering cumulus congestus or even cumulonimbus might be triggered over coastal water by convective instability due to a hypothetical off-shore evening breeze front, which fits some of the topographical and meteorological information.
Also, this distance seems to be pushing the limit for a high-contrast mirage image with minimal fill-in, due to the optical thickness of this much atmosphere, even at 19,000 ft. Mirage expert Andy YOUNG's calculation suggests that a fairly good visual contrast might be expected in the reported conditions at a distance of perhaps as much as 300 miles (about 500 km) , but this is still a good 30% short of keeping parallax drift down to 10°. And yet, the above allowance of a 10° parallax drift between first and last sighting is already taking a small liberty with Capt. HOWARD's 1954 account, which implies that the objects were "keeping station" within a much smaller tolerance than this. The Navigation Officer, George ALLEN "had his eye on them the whole time" (the aircraft was flying on autopilot) and checked their alignment against the edge of the cockpit window frame. ALLEN was reportedly able to observe only one brief departure from a constant bearing and this was not at the end of the sighting but in the middle, after which the objects "dropped back exactly parallel again".
Given that 10° is a rather large visual angle on the windscreen, corresponding to nearly the span of one's hand at arm's length (and about 20 times the implied angular width of the largest object), the claim of "exact" station-keeping implies a not-insignificant angular error on the part of the Navigator.
On the other hand we do have to remember that 10° is also the likely order of variation in instantaneous heading around an average autopilot heading of 49±5° according to Capt. HOWARD . It is not inconceivable that yaw in the aircraft axis could fortuitously have masked at least some part of a drift in true bearing of this order.
Symmetry & motion
Another issue is the remarkable lateral symmetry of form, and to some extent of motion too, shown in the BOAC case and in the very similar Zanthus case from Australia. In each case a large central shape-changing object is flanked by smaller objects either side. The small objects appear to move laterally in and out, exchanging positions, all finally moving inward to merge with the large central object, which then dwindles in angular width and disappears.
Because significant RI gradients in the free atmosphere  are exclusively vertical, mirage refraction only occurs vertically. True lateral bending of light rays cannot occur, so the reports of sideways motions in these cases (and also in the 1947 United Airlines case) must be explainable as illusions, or perhaps as separate small parts of an angularly-much-larger miraged target object being selected in sequence by fluctuating conditions in the duct - perhaps where the floor of the duct is being undulated by slow wind-driven or gravity waves.
An alternative explanation suggested by Andy YOUNG  depends on parallax. If we imagine that a number of different objects (dense cumulus towers in this case) at different distances are being miraged, then as the viewing geometry changes due to the aircraft's forward motion the relative angles of the lines of sight would vary and might even cross over. Objects nearer than the large one would appear to fall behind it; those farther away would appear to gain on it.
This might explain the appearance of motion, but does not of itself explain the bilateral symmetry of the arrays of moving images in this and in similar cases, which at the moment seems to be down to chance. This is not entirely satisfactory.
Whilst completing this report the author happened to look at radarscope photographs from a little-known USAF incident  and was startled to notice the date: July 2, 1954, or just two days after the BOAC incident. The reason for surprise: The photos were described as showing echoes from a large object and six smaller satellite objects tracked for 19 minutes passing under a B-36 at 14,000 ft near Bermuda.
According to the meagre scraps of information in the original Blue Book file the B-36 crew checked for air or sea traffic in the area and found there was none. The very experienced radar operator was convinced the objects were not ships, as he was familiar with ship echoes on the equipment and had in fact been tracking a ship just before the UFOs appeared. He had never seen anything like it in 10 years and 1500 hrs as a radar observer. But Blue Book "asked the Navy" and the file baldly asserts that the objects were positively identified as the "USS Mindora"  and six escort destroyers sailing from Naples to Norfolk, Virginia.
Considering the way in which Blue Book explained the BOAC sighting as "the planet Mars" with similar and (in that case) reckless bravado one is entitled to be sceptical of this completely undocumented claim; but it could be true . Howsoever, there is no chance of the USS Mindoro and escorts having been anywhere near Hudson Bay if they were at Bermuda when stated. At 19 kt it would take over 6 days to steam >2,800 nmi, even if it were not the case that they had supposedly been coming from Naples. So if the Blue Book identification of the Bermuda objects is accurate the similarity of configuration can only be coincidence.
Nevertheless the coincidence prompted Wim VAN UTRECHT to wonder if a formation of ships in Hudson Bay - perhaps another carrier group - could conceivably have been seen via mirage from the Stratocruiser, a theory which deserves a mention since one can imagine that relative motions among such a group of ships might explain motions of their mirage images.
But from 600 miles away, the angular length of the carrier would be only about 0.01 degree or 36 arcsec, only at best an unresolvable speck even in perfect laboratory conditions, and the angular height would be say 1/10 of this or a few arcsec, thus in practice scarcely perceptible.
So to image this ship and its still smaller escorts we need an axially symmetrical distribution of refractive index gradient around the line of sight, to obtain a 'magnification' in the order 100x as if by a telescopic lens. Such a mirage phenomenon is unknown to science and meteorologically inexplicable. There are some anecdotal stories, in the old literature, of what the cataloguer of anomalies William R. CORLISS has called "telescopic mirages", but modern mirage experts tend to regard such stories as being in the class of "sea serpent" tales .
In our case, the fact that ships at sea level would be 19,000 ft below the elevated duct in which (ex hypothesi) the plane was flying presents a serious problem. To couple into the duct with a <30 arcmin grazing angle light rays would first have to travel through the atmosphere for at least ~450 miles. As we have learned, fill-in due to scattering should grey out even high contrast sources in this sort of situation, never mind inevitable added losses from whatever unknown 100x-magnification "telescopic" miraging is then applied. Yet the UAPs were "black". It's perhaps possible (if highly unlikely) that a duct could lie tilted away from the horizontal by about 20 arcmin and dip towards sea level over a distance of 600 miles (a frontal inversion lying over a wedge of cold air perhaps), but this doesn't help with the contrast problem. Therefore combined with the "telescopic" requirement this means ships are not nearly so useful as towering clouds inside the duct, which can be half the distance from the eye in the first place, with half the distance losses, and higher intrinsic contrast because nearer the sunset longitude and so having much less front-lighting from scattered sky brightness.
Ground and air radars were involved also in the BOAC incident, and had radar contact been reported this might have made a strong case against the mirage theory. Indeed the case has often been cited as a radar-visual observation, but without any clear evidence. The sources for this appear to be no more than misinterpretation and rumour, such as BOWEN (1982) recounting a clearly confused memory of a conversation with Capt. HOWARD 15 years previously. Capt. HOWARD himself stated several times that he was not aware of any radar contact with the phenomena. It's true that Capt. HOWARD was for the most part in radio contact with Goose GCA (Ground Controlled Approach), the Air Traffic Control radar, and would himself have been beyond the probable ~60 mile range of GCA at least for much of the incident, and he may not have been made aware of any contacts detected by Air Defence radars . A rumour of an ADC radar contact in this case was recently mentioned on an Internet forum for ex-airforce personnel stationed at the Canadian Pinetree Line air defence radar sites . There was nominally no Pinetree Line radar at Goose until November 1954 , when the Melville Radar Station was established attached to Pepperell AFB in nearby St Johns, but Air Defence radar was operated there at the time by 641st Airborne Control & Warning Squadron, and a telex  from the Commander, 641st AC&W timed at 01:50 Z, 27 minutes after the end of the sighting (just after the Centaurus had landed at Goose to be met by a USAF intelligence officer), states:
VECTOR FIGHTER NO INTERCETP [sic] MADE RADAR NONE 
This suggests that the 641st AC&W radar was the fighter's controlling radar, except that the call-signs of that installation apparently were 'Capable' and 'Halfpint' , whereas the pilot of the fighter concerned recalls that the call-sign of his GCI at the time was indeed 'Pinetree'.
Recently, thanks to an Internet lead located by a UK researcher , the author was able to make contact with Lt. Col. Al KRAMER, USAF (ret.), the pilot of the 2-man F-94 jet fighter dispatched to investigate. In 1999, KRAMER had written about his experience on a message board for veterans of the 59th Fighter Interceptor Squadron, remarking:
"Just before I got to them, the BOAC Captain announced that the objects were veering off to the north. Neither Pinetree nor us picked up anything on our radars. The event did cause a certain amount of commotion on the Goose" .
In communication with the author a decade later in 2009, Lt. Col. KRAMER confirmed that Pinetree (located adjacent to Goose Bay; "I don't know if it was a part of the Pinetree Line" ) had both his F-94 and the Stratocruiser on radar, but no unknowns, and more importantly that the F-94 radar operator did pick up the Stratocruiser on the plane's AI radar, but no unknowns.
Lt. Col. KRAMER saw nothing visually, either. He had established radio contact with Capt. HOWARD, who informed him when he was still 30 miles away  that the objects were "veering off". Early on, Capt. HOWARD reported that the F-94 said "I'm heading towards you at a slightly higher altitude" (HOWARD, 1954). Lt. Col. KRAMER volunteered without prompting the information that his own altitude was 20,000 ft . The Stratocruiser is known to have been at the 19,000 ft pressure level. If the objects were mirage images then a 1,000 ft altitude difference may well have been sufficient to keep the F-94 above the duct and beyond the critical angle for refraction, helping to explain why Lt. Col. KRAMER saw nothing.
The F-94 approached the Stratocruiser from the East with guns and gun-camera both readied, and passed it heading West, then turned to the right in the direction the objects had reportedly disappeared. Lt. Col. KRAMER had by this time flipped his armament switch from "guns and camera" to "camera only" with the intention of obtaining 16mm film, but finding nothing visually or on radar the crew gave up and went in for a landing at Goose, where both men were "quarantined" because "the US Air Force did not want us to say anything" .
One other significant point is confirmation that the F-94 was not scrambled from Goose as many accounts suggest, but was already in the air. This answers the argument  that North East Air Command (NEAC) would not have launched a costly and risky interception mission unless they had a radar target to go at. "We were returning from an intercept mission", said Lt. Col. KRAMER, "and were diverted" . Indeed the absence of ground radar contact is implicit in the fact that the F-94 was instructed - very unusually - to make contact with the crew of the Stratocruiser for a final visual vector to the UAPs.
The only other (implicit) reference to radar in this case occurs in the Blue Book file, in a telex from 6607th Air Base Wing, Thule AFB, Greenland, timed 05:45 Z June 30, 1954, in response to a telex from 641st AC&W at Goose . Thule AFB was the site of a NATO early warning radar station operational from 1953 . The cryptic message is only eight words long but the crucial two words are "negative report". In other words, ADC radar at Goose checked with Thule only to be told that no radar UFOs had been detected there either. Of course, Thule is fully 1,700 miles due North from Goose . It may be that Thule was responding on behalf of other northern tier radars also, but even so this negative report is only indirect circumstantial evidence.
In conclusion, a couple of awkward issues remain that keep the mirage theory from being completely resolved; nevertheless several significant objections have been overcome and so many features are suggestive of mirage that it seems by far the least implausible explanation, bearing in mind the limitations of the available data. There is evidence that the observation is one of a hitherto unrecognised class of very similar mirage observations from aircraft which would repay further focused study.
A final note concerns Capt. HOWARD's recollection (1967; 1982) that soon after departing New York the Centaurus was kept in an unusual - indeed, unique in his experience - 10-minute holding pattern by Boston ATC before being allowed to proceed to Goose by way of a diversion to the East of Boston. No explanation was given. However as a result of publicity after the sighting Capt. HOWARD received a letter from an American doctor who had been holidaying in a Massachussetts lakeside cabin on that evening. They had heard a "roaring" noise and looked up to see seven dark objects - one large and six small - heading NE across the lake. Capt. HOWARD wondered if these might be the same seven "strange things" he had seen, and if so, whether they were responsible for Boston ATC radar re-routing his aircraft away from the area.
The latter could be the case if we interpret these ill-described "roaring" objects as jets.
Perhaps a sighting that had not seemed so strange to these witnesses at the time was reevaluated after Capt. HOWARD's story appeared in the press and on the newsreels. If so their description may have been influenced by it. Perhaps a military exercise of some sort or a midair refuelling operation was underway west of Boston that evening and had strayed close to a commercial airway? It is never desirable to invoke mere coincidence in a scientific treatment, but given the weight of circumstantial evidence favouring a mirage over Labrador this seems the most likely explanation.
Notes & references
 David THAYER writes:
"Based on Howard's first-hand account not long after the sighting near Goose bay, I am convinced now that what they saw was an unusually persistent superior mirage. In particular, HOWARD's description of the 'exit' of the assumed UFO forms an almost eerie parallel to my description of the behavior of a superior mirage on page 140 of the Condon Report: 'As the mirage-producing layer weakens (...) or the viewing angle increases (...) the mirage appears to dwindle to a point and disappears'. Compare that to HOWARD's description (...) If the observer thought he was observing a material object, it would be natural if to assume that the drastic shrinking effect represented a rapid recession in space rather than simply the change in size of an image. In retrospect, my 'rare event' may have been a bit ill-considered, although the material I had at hand with which to evaluate the incident was apparently somewhat slanted and included a few details that seemed to contradict the mirage theory, details that your investigation reveals may have been [witness] memory artifacts (...) I certainly would no longer tend to doubt the mirage theory" (personal communication, August 13, 2009).
 Personal communication, July 28, 2009.
 In fact they carried powerful cameras and like the US Navy Skyhook balloons were part of a programme of surveillance of Soviet military assets which also employed high-altitude jet overflights. The effort was continuous with later black projects including the U2, which was deployed with a similar weather-mapping cover story.
 AIR 2/17903, Project Genetrix, USAF press-release instructions.
 Personal communication, March 2, 2009.
 Letter from R. H. B. WINDER to Robert J. LOW, August 30, 1967, forwarding Capt. HOWARD's answers to questions posed by LOW. Colorado University UFO Project files; Library of the Am. Phil. Society, Philadelphia.
 19,000 ft above Newfoundland would be considered the "free atmosphere". Significant lateral gradients might occur in regions of extreme convective or radiative heat transfer, such as across the heat plume above a fire or adjacent to a vertical sun-heated wall.
 Personal communication, March 4, 2009.
 Two photos from the series of 32 appeared in the photo section of a book by a scientific consultant to the USAF's Project Blue Book (HYNEK, 1978) which gives the date as July 3, 1954. This is Zulu or GMT. Local time would have been on the evening of July 2. HYNEK incorrectly recorded Blue Book's evaluation as a "battleship" - not a carrier - and six destroyers. The complete Blue Book case file is available at www.footnote.com.
 Repeated mis-spelling of Bermuda as "Burmuda" does not increase one's confidence in ATIC's thoroughness. It's hard to work out what's going on from the poor images. We can just see from the data plate that it's a K-3A bombing/navigation system, which was the state of the art system on the B-36D variant flying at that time. The K-3A incorporated an upgraded Western Electric K-band tunable AN/APS-23 radar with a 60 inch flushmounted steerable antenna which could scan 360 degrees in PPI mode at up to 60 RPM, or in sector scans of 40°-180°, with a range of 5-200 miles. According to home.att.net "At 30,000 feet, large cities could be detected at a range of up to 200 miles and shipping could be detected at ranges of 50 to 100 miles". Usually the 'scope cameras on radars of this type expose one frame per antenna rotation. Unfortunately the selected scan rate and range scale are not shown, and even the counter numbers are hard to read. The display can usually be north-up or heading-up and there will be a heading marker strobe. If the white line is a heading marker on a north-up presentation, then the B-36 was flying roughly West. If the ships were heading West also (as reported) then the plane's true ground speed would be the relative target rate plus the ships' headway. If the range ring interval is 5 nmi then the objects fell behind underneath at a relative speed (if I read the numbers correctly) of about 180 kts, which, added to the 19 kt cruise speed of the USS Mindoro (Note ) is exactly the normal B-36 cruise speed (200 kts, neglecting winds).
 Andy YOUNG (personal communication, May 3, 2009) is very sceptical about many of CORLISS's entries. I tend to agree. They are often anecdotal and lack probative detail. One such "telescopic mirage"record is as follows:
"X9. No date given. Near Port Danger, on the South African coast. Passengers on a vessel saw a mirage of a well-known English man-of-war, which displayed great detail. Expecting to find the warship at Port Danger,they were surprised to learn that it was some 300 miles away at the time of the mirage" (CORLISS, 1984. p. 145).
CORLISS's source in this case is LIDDEL (1953). (Urner LIDDEL, employed by Bendix Corporation and the Office of Naval Research, was a close associate and vociferous supporter of Donald MENZEL and Harlow SHAPLEY and generally not very critical where optical "explanations" of UFO cases were concerned.) LIDDEL was himself quoting a Victorian book (BASSETT, F. S., Legends & Superstitions of the Sea & Sailors, Chicago, 1885) which in turn was paraphrasing a tale in another "modern book of travels" published half a century earlier. I am indebted to Roberto LABANTI (personal communications, October 3, 2009) for locating this book (Narrative of Voyages to Explore the Shores of Africa, Arabia & Madagascar, Bentley, London, 1833, Vol. 1, pp. 241-242).
According to page iv of its introduction it was put together in the form of a continuous narrative "from the journals of Capt. [W. F. W] OWEN [of HMS Leven] and the officers engaged under him" by a London editor, Mr. HEATON BOWSTEAD ROBINSON, who puts the above tale into the mouths of some anonymous spokesmen for the crew of HMS Leven (the tale is told in the second person plural; the only crewmember excluded as an author is in fact Capt. OWEN himself, referred to therein in the third-person singular). ROBINSON's account tells us that it was on 6th April 1823 when "we" were surprised to see, not two miles off, the Leven's companion ship, the sloop HMS Barracouta, believed to have been "above three hundred miles from us". The story says that "many well-known faces could be observed on deck" which along with "the peculiarity of her rigging" convinced the observers that it was really the Barracouta, even though "she made no effort to join us, but on the contrary, stood away" beyond the above-mentioned two miles. The writer concludes that such a "strange and at present unaccountable fact" may be explainable "by natural and probably simple causes" such as "refraction". I am further indebted to Roberto LABANTi for discovering the account of the corresponding period from the Journal of the 1st Lt. of HMS Barracouta (BOTELER, Thomas, Narrative of a voyage of discovery to Africa and Arabia, performed in His Majesty's ships, Leven and Barracouta, from 1821 to 1826, under the command of Capt. W.F.W. OWEN, R. N. Volume I, Bentley, London, 1835, pp. 220-21) describing the eventual arrival of HMS Barracouta on 14th April at Simon's Bay, where HMS Leven had been awaiting them anxiously for a week, after "a tedious and rough passage of ten days" from Algoa. This seems to confirm the view of the storyteller(s) that Barracouta must indeed have been far away when apparently seen off Port Danger; but the rough seas and uncomfortable cold complained of by 1st Lt. BOTELER during the journey do not on the face of it suggest the sort of stable anticyclonic high-pressure regime that typically produces the widespread sea-level temperature inversions most likely to give rise to exceptional long-range superior mirages. "Rough passage" suggests that the strong winds characteristic of the Cape summer were persisting into autumn (a week earlier HMS Leven had encountered "variable" weather on the same route "with two or three severe gales and some dead calms" (OWEN, op. cit., pp. 240-241). The April climate of the Cape is typically temperate and very humid, 12°C to 22°C and RH 60%-90% the mean circadian ranges, daily probability of rain ~20%, suggesting less than ideal atmospheric transparency (www.bbc.co.uk/weather/...).
Consider first the difficulty of identifying "many well-known faces" even at two miles distance "in the evening (...) at sunset", as it reportedly appeared, even in the clearest air; and then consider the difficulty when those faces are in fact 300 miles East of the observers, 5° of longitude nearer to Algoa. The extraordinary fidelity of this supposed mirage of "well known faces" over such a distance requires not only axially-symmetric telescopic refraction of an inexplicable kind, it also requires an almost lossless optical pathway and at least a well-lit scene to project. This distance corresponds to about 20 minutes of solar time. When it was sunset at Simon's Bay - as the crew of HMS Leven were watching their supposed counterparts on the Barracouta launching a small boat to retrieve a man overboard - at the location of the real Barracouta 300 miles to the East dusk was already well advanced, the sun's disk some 10 solar diameters below the local horizon, and the ship would be illuminated only by remnant scattered sky brightness. So should we accept this as evidence of a truly extraordinary "mirage"? Ought we not to require better and more circumstantial evidence than an anonymous anecdote before ruling out the arguably more probable theory that some unnamed observers were mistaken as to the identity of the ship, and that to explain their mistake someone said "it must have been a mirage"?
 In his "Voyage Report" (see Appendix B) he did say that he had spoken with "Fighter Control" but this may be a confusion. Elsewhere he describes how he spoke with the fighter pilot - although it's true to say that if he was given the fighter's ground control frequency he may have been able to hear both pilot and controller.
 Personal communication from Jan ALDRICH, Project 1947, August 7, 2009.
 Blue Book file documents (Telex from Commander 641st Aircraft Control and Warning Squadron, Goose Bay, to various commands).
 The first telex from 641st AC&W gives the time of BOAC's radio report (to Goose GCA) as 01:09 Z, with the message being relayed to 641st radar three minutes later at 01:12 Z. Presumably radar then began to keep a watch, but the BOAC itself was not picked up until almost 01:23. The next follow-up telex from 641st states "when we and fighter got contact with the aircraft objects faded [visually]". The visual bearing to the UAPs would place them necessarily at greater range than BOAC from both ground and AI radars, so strictly speaking absence of radar contact is not positive evidence for the mirage theory.
 Eric RUSH, personal communication, August 12, 2009
 Personal communication, August 25, 2009. The 641st AC&W radar located adjacent to Goose Bay was on Mount Dome 6 miles NW of the airfield) and was operational from August 1953 with CPS-6B and/ FPS-6 radars.
 Early accounts by Capt. HOWARD give the range as 20 miles.
 Personal communication, August 29, 2009
 Personal communication, August 25, 2009.
 Brad SPARKS, personal communications, August 13, 2009.
 Personal communication, August 25, 2009
 Blue Book file documents (telex from Commander 6607th Air Base Wing, Thule AFB, Greenland, to Commander 641st AC&W Squadron, Goose AFB).
 The Thule surveillance radar was an FPS-3, a modified L-band CPS-5 with no more than 200-mile range.