R e s e a r c h 
" C l o u d   c i g a r "   o v e r   S a i n t - P r o u a n t ,   F r a n c e

A funnel cloud, even one which does not develop to a tornadic touch-down, can be an impressive phenomenon, as many video clips available on the internet demonstrate. We have suggested several ways in which the report details might be reinterpreted consistently with an immature cold-core funnel cloud.

Some features which, if taken at face value, would seem to present problems for the cold-core funnel cloud theory are:

  • long duration around 30 minutes;
  • anchoring in one spot independent of rapid cloud motion;
  • then, instead of slow shrinkage and dissipation, a swift lateral bulk motion of the entire funnel as it "accelerated (...) into the clouds in the distance" or "went away rapidly in the direction towards which it was slanted";
  • a "machine"-like stability of shape during both stasis and discontinuous motion, possibly including rotation of the funnel axis in a vertical plane (i.e., exchange of top and bottom ends);
  • blue-violet luminosity bright enough not only to be visible in daylight but also to illuminate the underside of the cloud deck above;
  • rope-clouds spiralling neatly up and down the exterior of the funnel, appearing to be seeded by some bright emitted object.

Mature, highly energetic and destructive tornadoes spawned from storm cell mesocyclones have been reported to show faint luminosity, encircling rope clouds, and ball lightning and can last for many minutes.

The mean duration of tornado funnels is widely stated as in the order of 10 minutes, but duration is proportional to intensity and the curve is highly skewed by a very few rare cases of extremely violent long-duration tornadoes. According to Encyclopedia Britannica, "Most tornadoes are weak, lasting only about two to three minutes on average. A typical lifetime for strong tornadoes is about 8 minutes, while for violent events it is about 25 minutes" (emphasis added) [39].

A detailed study [40] of almost 150 tornadoes occurring across the midwestern US during the great super-outbreak of April 3-4, 1974 found that the mean duration of 22 weak F-0 tornadoes was 3.9 min, but only 4 examples lasted longer than 5 minutes whilst 14 cases (64%) lasted less than 1 min. Mean durations of 148 tornadoes sorted by F-rating are shown in Table 2.

Mean durations by F-rating of 148 tornadoes

Table 2: Mean durations by F-rating of 148 tornadoes recorded during April 3-4, 1974 American Midwest outbreak (based on data by ABBEY & FUJITA, 1983).

This does not mean that a 30-minute F-0 tornado could not occur. In the above 22 examples, one did. But the probability is low. Evidently the reported 30-minute stable duration of the St-Prouant object is becoming statistically improbable for mature tornadoes of an intensity below a damaging F-3. And we have shown that in the absence of weather evidence for a thunderstorm supercell there is no likelihood of a true tornado at all, rather the weather pattern indicates the likelihood of a cold-core funnel cloud generated by the passage of a weather front in conditions of thermal instability, where rotation begins along the convergence line between oppositely-moving air masses and is translated by vortex-stretching into a vertical funnel. According to the literature on non-mesoscyclone tornadogenesis [41], [42], [43], [44], [45], such a cold-core or cold-air funnel is a much weaker phenomenon, usually more short-lived even than weak tornadoes, with durations typically of a few seconds up to a few minutes. And in a study of about 100 cold-core funnel clouds in Minnesota it was noted that some 70% were "long, slender, rope-like formations" and only 30% resembled the shorter, fatter, columnar funnel with a blunt or rounded extremity [46].

So there are certain features which, on the face of it, may be said to be exceptional for a weak funnel cloud rather than typical. As we saw in the previous Chapter, at least some of these features can be individually explained away, with varying degrees of plausibility. But we have to concede that not all the phenomena described are easy to explain, if reported somewhat accurately.

The next question, then, is: How likely is it that these phenomena really did occur as reported? Perhaps the "blue-violet luminosity", the shiny disc with its "vapour trail", the rapid bulk motions etc., were really exaggerations or misunderstandings introduced by journalists or others in 1954 based on much less exciting features described by the witnesses. After making such allowance the residue of the report might reduce to characteristics of an immature funnel cloud - made remarkable by the unexpected reflectivity of funnel droplets and white bird plumage in direct sunlight against a dark backdrop of storm clouds, and perhaps ringed by some rope-like cloudlets (e.g. Fig. 11) that suggested the notion of a vapour trail.

In support of this sort of idea one can adduce the re-interview of M. FORTIN by two members of the group GEPA which was summarised in a short article in the group's magazine Phénomènes Spatiaux in 1968. It contained testimony from this main witness contradicting various details reported in 1954 and stating that he was unhappy with the treatment his report had received at the time:

We meet the witness on the road of Gabellière [sic], and he initially answers us with a certain reserve. At the sight of our G.E.P.A. membership card, he realizes that we did not come to ridicule him. He then tells us of his observation, quite vivid in his memory after more than ten years.

Working in a bean field, at the end of the afternoon, under a not very cloudy sky, the witness has his attention drawn by an odd cloud in the shape of a carrot, white-yellow, which descends slowly from the sky, tilted at 45°, pointed end down, and which stops at about fifty meters from the witness and approximately 6 or 8 meters above the ground. The size of this cloud is compared with that of a power transformer station of the EDF [state power company] which the witness indicates at the edge of the road (a concrete building of 3 by 3 meters square base of 10 meters height roughly).

During the descent, some kind of smoke escapes and is coiled up while going up around the cloud, in the manner of a corkscrew. After becoming stationary, the cloud dissipates little by little and it allows a brilliant ball to become visible (or a disc keeping the same face towards the witness) from 35 to 40 cm in diameter. Then this ball carries out an impressive number of fast and jerky operations, stopping and starting a rectilinear trajectory in a new direction, turning from left to right as well as up and down, following a non-flat broken line. Finally, the ball abruptly climbs vertically and is lost in the altitude. The whole thing lasted twenty minutes.

The witness has a negative memory of the investigation at the time, carried out by a journalist "tall, rude and unfriendly," who questioned him only superficially and seemed to know better than him what he had seen. He is furious that the man wanted to make him say things which he did not see, and that his statements were altered. Since then, he has been interviewed only once, by letter, by a journalist from La Rochelle. His sighting is confirmed by other people, who were located very nearby.

Mr. Bonifait, electrician, did not see the disc, but only the odd cloud, which he saw coming from a distance as he was working on the top of a power pole. This cloud made him suspect a sudden storm, but he had been intrigued by the behaviour of this cloud, completely unusual. He only witnessed its descent towards the ground, after which the cloud disappeared from his sight, hidden by trees.

This observation of Saint-Prouant is thus not to be recorded under the type of the large vertical cigar, as in Aimé Michel's classification (...)" [47].

As noted by the editor of Phénomènes Spatiaux, emphasising the thrust of the above report, there are between it and the 1954 account of MICHEL...

...undeniable similarities, but also serious disagreements (...) Whilst in the narrative offered by Michel the 'cloud cigar' remains present until the end - because the disk is housed inside and re-enters through the tip, as it was released - in the deposition of M. Fortin [meaning the new account thereof by JAY and METAYER - authors' note], the cloud begins to dissipate on becoming stationary, and does not reappear, that same cloud dissipation revealing the disc or the ball. Regarding the 'cloud cigar' both descriptions are, except in the very first moments of the appearance of this cloud, totally incompatible... it was only a secondary effect of the presence of the hidden disc, a sort of aura of conical form that was created around the disc and obscured it from the view of witnesses.

The differences between this account and the account(s) originally attributed to M. FORTIN are indeed rather striking, so much so that they scarcely seem like descriptions of the same event. The question of which source should be given the most credence is a ticklish one.

Generally, testimony recorded close to the events is much to be preferred over testimony recorded many years later - a prudent principle regularly proven valid in solutions of historical cases. On the other hand, this depends on knowing that the original source is an accurate and reliable record of the witness's testimony. In the present case it is not entirely clear that we can be confident of this, since the published statements may all originate from the notes of a single anonymous journalist in a local newspaper, and many newspaper reports during this French wave of 1954 are known to have been unreliable, or in some cases even made up.

At the same time, however, the account recorded by JAY and METAYER in 1968, whilst being (one supposes) a conscientious record of an interview with a witness, has the date wrong, contains no relevant direct quotation of the witness (apparently the interview was conducted without the benefit of tape recording, and an uncertain time before the date of the article), and is lacking in analytical perspective. JAY and METAYER do not quote anything that we could claim (even with ordinary journalistic latitude) came verbatim out of the mouth of the witness - except for four words describing a local journalist. This is less than helpful. Moreover, it is ironical that the 1968 version of the story tends to undermine the argument from phenomenology that seemed to point so distinctively to a tornado funnel in the first place.

This relatively small (10 m-long) yellow-white cloud that FORTIN now describes coming near him is nothing like the "gigantic" violet-blue affair in the 1954 accounts, seen from many villages "all over the countryside" and emerging from the clouds "half a mile" above the earth. And the cloud is no longer even the primary phenomenon, merely a transient smokescreen (so to speak) exuded by the real "UFO" within. The duration of the whole incident has also now shrunk from 30 minutes to 20 minutes. And instead of a "storm" brewing he now says that the sky was after all "not very cloudy".

The object in this new version is still described as a sort of "cloud", but was simply seen descending, in the clear, coming down almost to the point of making a landing on the ground only metres away from him, not extruding from a high cloud deck and travelling horizontally as originally suggested in 1954. This new impression is reinforced (according to the authors) by the electrician, BONIFAIT, who did not see the little disc but who (they say) did himself describe the main object or "carrot" itself as a free-flying cloud which descended "towards the ground" until it "disappeared from sight, hidden by trees". This again contradicts the 1954 accounts by MICHEL and La Résistance, in which BONIFAIT said that it was the little disc (not the parent cloud) that was obscured by trees. But no interview with BONIFAIT is described by JAY & METAYER and no context or quotation is given for this claim, which is very unsatisfactory.

They then attribute to FORTIN a version of the little glittering "disc" which is quite different (and, by the way, inconsistent with our suggestion above that what he saw was a sunlit bird). The disc is now a "brilliant ball" like some amazing ball-lightning manifestation, whilst the oddest and least funnel-like of the originally-reported "cloud" effects - the spiralling "contrail" left by the disc - is not deprecated in this new 1968 report, but is actually reinforced by way of an unambiguous drawing of a tight "corkscrew" trail (Fig. 10) which one presumes was sanctioned by a specific description from FORTIN if not actually drawn by him.

So the funnel cloud theory is arguably not very well served by this late information. But neither is any other explanation because the authors of the rather cursory article do not make any attempt to probe the meaning of these disparities. Indeed, they show no awareness that such issues might even exist.

For example, what specific statements of the witness lay behind the following paraphrase?

During the descent, some kind of smoke escapes and is coiled up while going up around the cloud, in the manner of a corkscrew. After becoming stationary, the cloud dissipates little by little and it allows a brilliant ball to become visible.

From this version we tend to infer that the cloud itself (i.e., the carrot) apparently dissipates to reveal the bright ball. This is evidently the authors' understanding. But it takes little imagination to realise that the opportunity exists for confusion in the retelling between a dissipation of the "smoke" or "vapour trail" to reveal the disc - as described in the original version - and dissipation of "the cloud" as described here. Unfortunately the authors appear oblivious of the need for readers to be able to calibrate the meaning of such ambiguity. So it is unsatisfactory that we have to rely on the authors' interpretation without the witness's verbatim account.

As noted above, this vital discrepancy in the new account was highlighted by the editor of Phénomènes Spatiaux, but like the authors he does not appear to appreciate the potential for confusion between dissipation of the smoke trail and dissipation of the cloud proper. When neither editor nor authors reflect on these unreasonable-seeming discrepancies, how can we be sure what M. FORTIN was intending to convey? Should we have confidence in their interpretations of FORTIN's words [48]?

JAY and METAYER's article certainly does require us to consider the possibility that FORTIN's "gigantic" carrot of 1954, which does not dissipate but instead re-absorbs the "disc", then rotates back up and moves off horizontally into the distance as a rigid body, could have owed its existence to errors and exaggerations in the original published description. But we note that this description is substantially corroborated by the accounts of other named and quoted witnesses - e.g., Mme PIZOU, who saw the object all the time above the tops of intervening trees, and by M. MERCIER and M. PERROCHEAU and others who saw it (according to very sketchy reports, it must be admitted) above the roofs of nearby buildings from within the farm courtyard. These are situations which appear unlikely to have afforded extended views of a smaller object behaving as latterly claimed by FORTIN, descending near the ground beside him out in the fields. And we should note that the distribution of other witnesses around an area a couple of kilometres across, as shown in Fig. 2, is consistent with the original scenario of a huge object hanging stationary at a high elevation, visible for a long while over a wide area of countryside as reported in 1954, but seems difficult to square with a 10m cloud that came down almost to the ground within metres of M. FORTIN at "La Gabilière" (La Gadbellère) and there dissipated. (Even initally, whilst still high enough to be visible from distant farms a kilometre or more away, the apparent length of this 10m "carrot" would be only millimetres at arm's length, and hardly a striking phenomenon.)

We can speculate that the date error in the article's title was not the fault of Jay and Metayer, but that of the editor; and perhaps the reason why the date of the interview is not given is another editorial lapse. But JAY and METAYER speak of M. FORTIN's recollection having been "more than 10 years" old, suggesting an interview date somewhat after September 1964. That would imply they are writing anything up to 4 years after the interview they are paraphrasing. On the other hand a reference to a later sighting by M. BONIFAIT, dated July 24, 1967, would seem to contradict this, indicating a maximum lag of 17 months between interview and article and suggesting that M. FORTIN's recollection would have been 13 or even 14 years old. However one looks at it this is not entirely coherent.

So we conclude that we are reading a summary of someone's interpretation of what they think M. FORTIN intended to say to them many years after the event, and we should probably have limited confidence in it.

However, let us consider the implications if the GEPA authors' summary does accurately record what M. FORTIN intended. In this case, might he have had a motive for wishing to dissociate himself from the aspects of the 1954 version that suggested a huge funnel-like cloud extruding from storm clouds?

Let us consider the possibility that after initial excitement in the newspapers someone offered an explanation: A tornado funnel. We know the idea occurred to people before what we might call M. FORTIN's "retraction" in 1968. It was raised by MICHEL in 1958, in order to dismiss it, and he apparently sought the opinion of meteorologists. Others no doubt thought of it too. If M. FORTIN was at all attached to the idea that he had seen a true UFO - which would be consistent with his willingness to talk to the GEPA interviewers only after he had been reassured that they were ufologists and likely to be sympathetic - then he could have been motivated in 1968 to deprecate some features of the original giant pendant cloud which had appeared so suggestive of a tornado funnel, and tell instead of this small yellow cloud making a near-landing close to him.

This is very speculative of course, but, whichever way we look at it, this 1968 source provides an uncertain basis for calibrating the original story. It records an interpretation of the distant recollections of one witness, and the reliability of its own reportage is in question. Therefore it is not strong evidence for our proposal that the 1954 accounts were materially distorted. At least those did contain what purported to be quoted witness statements, and even though we would be on strong ground in guessing that they are not really verbatim (knowing how journalism works, especially with pencil-and-paper recording in 1954), multiple people are named and quoted describing the event in somewhat the same terms.


In summary, we have no certain documentary basis for rewriting the original accounts of a large, stable, glowing "carrot", its fast-moving horizontal flight at cloud level punctuated by a 30 minute period of stationarity. The claim that the witnesses did not describe these things rests on contradiction by sources dated years after the event and of uncertain status; therefore taken at face value the reports remain strictly speaking unexplained. There is also scant likelihood of a true tornado. The weather pattern indicates conditions favourable only for a weak cold-core funnel cloud unlikely to exhibit rare and spectacular properties.

However, we have been able to suggest interpretations which, with some reasonable assumptions, might reconcile at least some of the described features with natural effects that might accompany such a funnel cloud. Moreover, in the absence of any reliable gauge by which to calibrate the original sources, it remains reasonable to question if the journalists and flying saucer writers who reported on the events have transmitted to us a reliable record, or a picture distorted by their own imaginations. It is only prudent to note that the incident at Saint-Prouant took place in the midst of a major wave of "flying saucer" reports that swept across France in September and October 1954. Some of these events remain unexplained, but grotesque misidentifications of natural phenomena and common airborne objects were a daily occurrence in this period. In those excitable circumstances, a rare and impressive meteorological phenomenon is likely to stir the imaginations of even well-intentioned journalists, if not witnesses.

In conclusion, without sufficient reason to rule out the possibility of error and exaggeration in the reporting chain, and given the suggestive synoptic weather pattern, a cold-core funnel cloud can't be excluded as a possible cause of the sightings.

 Notes & references


[40] ABBEY, Robert F. & FUJITA, Theodore T., "Tornadoes: The Tornado Outbreak of 3-4 April 1974" in The Thunderstorm in Human Affairs, second edition, edited by Edwin KESSLER, University of Oklahoma Press, Norman, OKlahoma, 1983, pp. 37-66.

[41] ZOLTÁN, Polyánszky, "Non-mesocyclone tornadoes in Hungary", Hungarian Meteorological Service Aviation and Severe Weather Forecasting Division XXX OSTIV Congress 2010, Szeged, Hungary,

[42] DAVIES, Jonathan M., "Tornadoes with Cold Core 500-mb Lows",

[43] WAKIMOTO, R. M. & WILSON, J. W., "Non-supercell tornadoes" in Monthly Weather Review, 117, 1989, pp. 1113-1140.

[44] BRADY, R. H. & SZOKE, E. J., "A case study of nonmesocyclone tornado development in northeast Colorado: Similarities to waterspout formation" in Monthly Weather Review, 117, 1989, pp. 843-856.

[45] COOLEY, Jack R., 1978: "Cold Air Funnel Clouds" in Monthly Weather Review, 106, 1978, pp. 1368-1372.

[46] Ibid.

[47] JAY, Dominique & METAYER, R.P., "Le 'cigare' de Saint-Prouant n'était-il qu'un nuage? Enquête rétrospective et Observation récente", Phénomènes Spatiaux (GEPA), #18, December 1968, pp. 26-29. Amended translation based on a version by Patrick GROSS and scans of the complete original article.

[48] It is true that parts of JAY and METAYER's submission containing such reflections and reasonings could have been cut by the Editor for publication, but there is no evidence of this in the article or in the Editor's lengthy commentary, and we can only deal with what we have.